{
  "status": "completed",
  "completed_count": 20,
  "applicable_count": 20,
  "failed_count": 0,
  "schema_version": 13,
  "config_hash": "1d4fed2e9ff4e107",
  "data": {
    "scenes": [
      {
        "scene_index": 1,
        "heading": "INT. 산악 접촉소 — 황혼",
        "beat_title": "경고 문구 점멸과 격리",
        "representative_moment": "통신판 디스플레이 중앙에 붉게 점멸하는 'LEGACY AUTHORITY DETECTED' 텍스트.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "Perpendicular extreme close-up on the digital display surface with cool low-key contrast",
            "t2i_prompt": "Photorealistic cinematic still. Seen from a perpendicular extreme close-up directly aligned with the lens axis, the active surface of the communication panel display occupies the entire frame. Centered in sharp focus, the digital warning text \"LEGACY AUTHORITY DETECTED\" dominates the composition, caught at the brightest instant of its pulse. The surrounding cool, muted low-key ambient is pierced by the intense, glowing warning-red accent of the letters. A subtle digital pixel grid texture is visible across the illuminated text, capturing the austere utilitarian interface without showing the physical device borders.",
            "outfit_assignments": [],
            "source_facts": [
              "S1 action: 통신판의 글자가 붉게 맥박친다. LEGACY AUTHORITY DETECTED",
              "shot description: 붉은빛이 맥박치듯 가장 밝아진 순간의 통신판 스크린 중앙에 뜬 텍스트 클로즈업"
            ],
            "visual_inferences": [
              "Display interface has a digital pixel grid texture consistent with legacy technology",
              "The screen surface itself acts as the primary focal plane"
            ],
            "creative_decisions": [
              "Omitted physical screen bezels to align with the through_device rule, focusing purely on the digital text and its red pulse",
              "Applied cool, muted tones to the unlit portions of the screen to enhance the low-key contrast with the warning red"
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [],
            "reference_phrase_kinds": [],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "a106da035b6c123f",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "03194940ff958fae",
              "owned_usage_hash": "8de771497838ce3d",
              "validator": "scene_detail_owned_objects.v1.close_skip",
              "violations": []
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": -1,
        "dependency_reason": "No dependent visual context required for this isolated screen insert shot.",
        "_shot_index": 1,
        "visible_entities": [],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 1,
            "shot_index": 1
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "close",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "At display height, the camera faces the communication panel perpendicularly at extreme close range, with the warning text held precisely on the lens axis and the screen edges retained as a narrow spatial reference. The static composition catches “LEGACY AUTHORITY DETECTED” at the brightest instant of its pulse, concentrating the restrained warning-red accent in the center of the frame.",
            "lighting_mood": "Cool, muted low-key rendering is punctuated by the warning text at the peak of its red pulse.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 0명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "primary_subject_only_fully_visible": "the named primary subject (hand / face / object) is the only fully visible entity in the frame",
                  "other_entities_appearance_options": [
                    "partial visibility — shoulder slice / arm slice / sleeve / fingertips entering the frame edge",
                    "soft background — out-of-focus silhouette via shallow depth of field",
                    "absent — not rendered at all even if listed in visible_entities"
                  ],
                  "forbidden_combinations": [
                    "same character's full body and body-part close-up in the same frame (view mixing)",
                    "two characters both rendered as sharp close-up faces in the same frame"
                  ],
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "wide_medium_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=close — wide/medium 규칙은 이 샷에 해당하지 않는다"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": null,
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [],
            "allowed_outlook_pairs": [],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": true,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "skipped_close_framing",
            "bg_id": "L11B02",
            "reference_usage": "skipped_close_framing",
            "owned_objects": [],
            "camera_reference": null,
            "close_framing_skips_background_ref": true,
            "constraints": [
              "this shot uses close framing — the background reference is intentionally omitted at composition time",
              "describe the close subject and its immediate surroundings without relying on the absent background reference",
              "use [L##: ...] block for frame-edge surface description"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [],
            "previous_shot_refs": [],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [],
            "forbidden_refs": [
              {
                "kind": "background",
                "reason": "close framing skips chain_bg reference"
              }
            ],
            "readiness_policy": "skipped_by_policy",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": true,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "4637666592fbb169"
      },
      {
        "scene_index": 1,
        "heading": "S1. INT. 산악 접촉소 — 황혼",
        "beat_title": "토니를 향한 앨런의 무력 위협",
        "representative_moment": "화면 왼쪽에서 들어온 P14의 총구가 가슴 정중앙에 닿자 굳은 표정으로 시선을 돌리는 C01.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "chest-height oblique three-quarter close view with low-key red emergency illumination",
            "t2i_prompt": "Photorealistic cinematic still. C01O01 in dark rescue workwear, a Caucasian man in his 30s, occupies the middle-right midground, his torso held rigid. Seen from chest height in an oblique three-quarter close view, his upper body isolates the right side of the frame. The long barrel of P14 enters obliquely from the left edge, the muzzle touching the exact center of his chest with firm, decisive contact. Angled in a three-quarter profile from the left, his eyes are fixed beyond the left edge on the unseen threat, ignoring the camera lens. Low-key red emergency illumination floods the confined space, casting stark, muted red highlights across his tense jawline and the metallic weapon. Immediately behind the subjects, [: rough rocky wall surfaces submerged in heavy shadow] dissolve into a soft blur, emphasizing the tight, claustrophobic atmosphere.",
            "outfit_assignments": [
              {
                "character_id": "C01",
                "outlook_id": "O01"
              }
            ],
            "source_facts": [
              "앨런이 장총을 토니의 가슴에 댄다.",
              "바위문이 닫히고 내부 조명이 붉게 변한다.",
              "Shot 9: 장총의 총구가 토니의 가슴 정중앙에 닿은 찰나"
            ],
            "visual_inferences": [
              "fixed_element body_full_pose: Tony's torso braced rigidly against the muzzle pressure",
              "softly blurred rocky background surfaces due to the close framing scale",
              "red highlights accentuating the metallic texture of the rifle and Tony's facial features"
            ],
            "creative_decisions": [
              "chest-height close framing isolated on the contact point to maximize spatial tension between the weapon and the subject",
              "restrained low-key contrast to maintain the tense, dark techno-thriller mood"
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [
              "fsc_001"
            ],
            "reference_phrase_kinds": [
              "character"
            ],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "7ba535503c371111",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "1fc536df21e8599c",
              "owned_usage_hash": "8de771497838ce3d",
              "validator": "scene_detail_owned_objects.v1.close_skip",
              "violations": []
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 1,
        "dependency_reason": "접촉소 내부의 붉은 비상 조명 및 거친 바위 질감의 공간 연속성 유지 (close framing으로 인해 아웃포커싱 처리됨)",
        "_shot_index": 9,
        "visible_entities": [
          "C01",
          "P14"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 1,
            "shot_index": 9
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "close",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "At Tony's chest height beside the Alan–Tony axis, the dolly-in resolves into an oblique three-quarter close view as the rifle enters from the left and meets the exact center of Tony's chest. Tony occupies the right side of the frame with his torso held rigid against the pressure, while his eyes remain fixed beyond the left edge on Alan rather than on the lens.",
            "lighting_mood": "The contact-post interior remains under the stated red lighting, rendered with low-key contrast and restrained highlights.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 1명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "primary_subject_only_fully_visible": "the named primary subject (hand / face / object) is the only fully visible entity in the frame",
                  "other_entities_appearance_options": [
                    "partial visibility — shoulder slice / arm slice / sleeve / fingertips entering the frame edge",
                    "soft background — out-of-focus silhouette via shallow depth of field",
                    "absent — not rendered at all even if listed in visible_entities"
                  ],
                  "forbidden_combinations": [
                    "same character's full body and body-part close-up in the same frame (view mixing)",
                    "two characters both rendered as sharp close-up faces in the same frame"
                  ],
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "wide_medium_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=close — wide/medium 규칙은 이 샷에 해당하지 않는다"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": {
              "reason": "primary_subject_isolation",
              "constraints": [
                {
                  "constraint_id": "fsc_001",
                  "target_kind": "character",
                  "target_id": "C01",
                  "label": "토니(앤서니 로저스)",
                  "screen_zone": "middle_right",
                  "depth_plane": "midground",
                  "gesture_action": "none",
                  "gesture_target_label": ""
                }
              ]
            },
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "C01",
              "P14"
            ],
            "allowed_outlook_pairs": [
              {
                "character_id": "C01",
                "outlook_id": "O01"
              }
            ],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": false,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "skipped_close_framing",
            "bg_id": "L11B02",
            "reference_usage": "skipped_close_framing",
            "owned_objects": [],
            "camera_reference": null,
            "close_framing_skips_background_ref": true,
            "constraints": [
              "this shot uses close framing — the background reference is intentionally omitted at composition time",
              "describe the close subject and its immediate surroundings without relying on the absent background reference",
              "use [L##: ...] block for frame-edge surface description"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "contact_post_red_interior",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The enclosed interior of the mountain contact post features rough, rocky wall surfaces submerged in heavy shadows. Low-key red emergency illumination floods the space, casting stark, muted red highlights across the uneven stone textures and creating a highly confined, tense atmosphere.",
                "element_scope": "full",
                "applies_to_shots": [
                  9,
                  12
                ],
                "source_facts": [
                  "The scene takes place in a mountain contact post with a rock door.",
                  "The interior lighting turns red after Alan hits the switch and the rock door closes.",
                  "Shots 9 and 12 explicitly describe the interior remaining under low-key red lighting with restrained highlights."
                ],
                "visual_inferences": [
                  "The walls retain their natural, unrefined rock texture, which catches the red emergency light only on the protruding edges while leaving the crevices in deep shadow."
                ],
                "creative_decisions": [
                  "Combined the rocky cavern architecture and the red emergency lighting into a single, cohesive environmental state to ensure background continuity for all shots following the lockdown."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 1,
                "shot_index": 1,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "character_outlook",
                "id": "C01O01",
                "policy": "required"
              }
            ],
            "forbidden_refs": [
              {
                "kind": "background",
                "reason": "close framing skips chain_bg reference"
              }
            ],
            "readiness_policy": "skipped_by_policy",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": true,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "879e00f6d2e50ba7"
      },
      {
        "scene_index": 1,
        "heading": "S1. INT. 산악 접촉소 — 황혼",
        "beat_title": "포위망 접근",
        "representative_moment": "지도 스크린 위로 붉은 점들이 좁혀오는 화면의 클로즈업.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "Steep overhead close-up focusing purely on the digital map surface, bypassing the physical bezel, with digital texture and low-key red reflections.",
            "t2i_prompt": "Photorealistic cinematic still. Seen from a steep overhead angle, a digital map interface fills the frame. Concentrated warning-red points form a tight circular encirclement in the center of the dark topographical layout. Faint digital texture and subtle pixel grids overlay the graphics. A dim reflection of a rough, rocky ceiling bathed in low-key red emergency illumination catches the glossy surface, maintaining the confined, tense atmosphere without softening the sharp graphic severity of the digital markers.",
            "outfit_assignments": [],
            "source_facts": [
              "The scene takes place in a mountain contact post with a rock door.",
              "The interior lighting turns red after Alan hits the switch and the rock door closes.",
              "지도 위 붉은 점들이 포위망을 좁힌다."
            ],
            "visual_inferences": [
              "fixed_element environment_state: contact_post_red_interior",
              "The through_device perception mode requires rendering the screen contents directly, skipping the physical frame.",
              "The screen surface reflects the red-lit rocky ceiling, preserving environmental continuity."
            ],
            "creative_decisions": [
              "Omitted the physical device bezel completely to fulfill the 'through_device' constraint.",
              "Applied a subtle screen reflection to carry over the 'contact_post_red_interior' fixed element without breaking the digital focus.",
              "Described the tightening dots as a frozen tight circular encirclement to respect the single-moment shutter rule."
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [],
            "reference_phrase_kinds": [],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "9de56681d9d8fd7f",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "b41d622f74b37edc",
              "owned_usage_hash": "8de771497838ce3d",
              "validator": "scene_detail_owned_objects.v1.close_skip",
              "violations": []
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 1,
        "dependency_reason": "Relies on the established environment state (contact_post_red_interior) from the earlier shots in the same scene to maintain lighting and texture continuity.",
        "_shot_index": 12,
        "visible_entities": [],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 1,
            "shot_index": 12
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "close",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "The completed crane rise places the camera above and close to the map screen, looking steeply downward as the move settles into near stillness. The map field occupies the central composition with a slim perimeter of its display retained, and the tightening ring of red points becomes the sole visual center.",
            "lighting_mood": "Low-key red interior illumination supports the concentrated warning-red points on the map without softening their graphic severity.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 0명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "primary_subject_only_fully_visible": "the named primary subject (hand / face / object) is the only fully visible entity in the frame",
                  "other_entities_appearance_options": [
                    "partial visibility — shoulder slice / arm slice / sleeve / fingertips entering the frame edge",
                    "soft background — out-of-focus silhouette via shallow depth of field",
                    "absent — not rendered at all even if listed in visible_entities"
                  ],
                  "forbidden_combinations": [
                    "same character's full body and body-part close-up in the same frame (view mixing)",
                    "two characters both rendered as sharp close-up faces in the same frame"
                  ],
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "wide_medium_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=close — wide/medium 규칙은 이 샷에 해당하지 않는다"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": null,
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [],
            "allowed_outlook_pairs": [],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": true,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "skipped_close_framing",
            "bg_id": "L11B02",
            "reference_usage": "skipped_close_framing",
            "owned_objects": [],
            "camera_reference": null,
            "close_framing_skips_background_ref": true,
            "constraints": [
              "this shot uses close framing — the background reference is intentionally omitted at composition time",
              "describe the close subject and its immediate surroundings without relying on the absent background reference",
              "use [L##: ...] block for frame-edge surface description"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "contact_post_red_interior",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The enclosed interior of the mountain contact post features rough, rocky wall surfaces submerged in heavy shadows. Low-key red emergency illumination floods the space, casting stark, muted red highlights across the uneven stone textures and creating a highly confined, tense atmosphere.",
                "element_scope": "full",
                "applies_to_shots": [
                  9,
                  12
                ],
                "source_facts": [
                  "The scene takes place in a mountain contact post with a rock door.",
                  "The interior lighting turns red after Alan hits the switch and the rock door closes.",
                  "Shots 9 and 12 explicitly describe the interior remaining under low-key red lighting with restrained highlights."
                ],
                "visual_inferences": [
                  "The walls retain their natural, unrefined rock texture, which catches the red emergency light only on the protruding edges while leaving the crevices in deep shadow."
                ],
                "creative_decisions": [
                  "Combined the rocky cavern architecture and the red emergency lighting into a single, cohesive environmental state to ensure background continuity for all shots following the lockdown."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 1,
                "shot_index": 1,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [],
            "forbidden_refs": [
              {
                "kind": "background",
                "reason": "close framing skips chain_bg reference"
              }
            ],
            "readiness_policy": "skipped_by_policy",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": true,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "cdb496ecc11db10b"
      },
      {
        "scene_index": 2,
        "heading": "S2. INT. 산악 접촉소 — 황혼 (00:45–01:30)",
        "beat_title": "무질량 잔류 인지",
        "representative_moment": "P17 휴대용 스캐너 화면에 무질량 잔류 경고 텍스트가 선명하게 떠 있다.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "downward-oblique close view from chest height",
            "t2i_prompt": "Photorealistic cinematic still. Seen from just above chest height and angled downward in a steep oblique close view, the active front face of the P17 portable scanner dominates the frame. The device is angled slightly upward toward the camera. Within the central field of its glowing digital interface, the bright text \"MASS DEFICIT / UNKNOWN RESIDUE\" is clearly displayed, rendered with subtle digital pixelation and a stark, austere layout. A narrow portion of the rugged physical casing of the scanner body frames the screen, preserving its scale and orientation. Neutral, low-key ambient illumination from the [: dim, twilight-lit enclosed workspace] catches the edges of the bezel, casting soft shadows that contrast sharply with the crisp diagnostic readout.",
            "outfit_assignments": [],
            "source_facts": [
              "S2 action: 스캐너 화면: MASS DEFICIT / UNKNOWN RESIDUE",
              "shot description: 스캐너 화면에 뜬 MASS DEFICIT / UNKNOWN RESIDUE 텍스트 클로즈업"
            ],
            "visual_inferences": [
              "fixed_element workspace_ambient_lighting: Neutral, low-key ambient illumination, subdued twilight atmosphere",
              "The interface has a stark layout with slight digital pixelation to represent the future's functional technology"
            ],
            "creative_decisions": [
              "Used a steep downward oblique angle to isolate the scanner face and exclude any character elements, adhering to the 0 persons visibility constraint",
              "Applied a high-contrast lighting balance between the glowing text and the dim physical casing to enhance the tense, claustrophobic mood"
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [],
            "reference_phrase_kinds": [],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "e8b1b09c78017207",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "73d0da9163b06d47",
              "owned_usage_hash": "8de771497838ce3d",
              "validator": "scene_detail_owned_objects.v1.close_skip",
              "violations": []
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 1,
        "dependency_reason": "분위기/조명 연속성 유지 (workspace_ambient_lighting)",
        "_shot_index": 3,
        "visible_entities": [
          "P17"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 2,
            "shot_index": 3
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "close",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "From just above Wilma's hand position at chest height, the dolly-in ends on a downward-oblique close view of the scanner display. “MASS DEFICIT / UNKNOWN RESIDUE” sits within the central field, while a narrow portion of the scanner body preserves its scale and orientation.",
            "lighting_mood": "Neutral low-key ambient light appropriate to the enclosed work area keeps the austere display text crisp.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 0명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "primary_subject_only_fully_visible": "the named primary subject (hand / face / object) is the only fully visible entity in the frame",
                  "other_entities_appearance_options": [
                    "partial visibility — shoulder slice / arm slice / sleeve / fingertips entering the frame edge",
                    "soft background — out-of-focus silhouette via shallow depth of field",
                    "absent — not rendered at all even if listed in visible_entities"
                  ],
                  "forbidden_combinations": [
                    "same character's full body and body-part close-up in the same frame (view mixing)",
                    "two characters both rendered as sharp close-up faces in the same frame"
                  ],
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "wide_medium_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=close — wide/medium 규칙은 이 샷에 해당하지 않는다"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": null,
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "P17"
            ],
            "allowed_outlook_pairs": [],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": true,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "skipped_close_framing",
            "bg_id": "L11B02",
            "reference_usage": "skipped_close_framing",
            "owned_objects": [],
            "camera_reference": null,
            "close_framing_skips_background_ref": true,
            "constraints": [
              "this shot uses close framing — the background reference is intentionally omitted at composition time",
              "describe the close subject and its immediate surroundings without relying on the absent background reference",
              "use [L##: ...] block for frame-edge surface description"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "workspace_ambient_lighting",
                "element_type": "environment_state",
                "character_name": "",
                "description": "Neutral, low-key ambient illumination fills the enclosed workspace. The subdued twilight atmosphere casts soft shadows and ensures that the bright text of electronic displays stands out sharply against the dim surroundings without any unmotivated color sources.",
                "element_scope": "full",
                "applies_to_shots": [
                  3,
                  8,
                  11
                ],
                "source_facts": [
                  "Neutral low-key ambient light appropriate to the enclosed work area",
                  "Neutral, subdued ambient illumination preserves the monitor's austere interface",
                  "Neutral low-key ambient light shapes Wilma's tense features without introducing an unmotivated color source"
                ],
                "visual_inferences": [
                  "The primary light sources are practical dim overheads combined with the glow spilling from the operational screens."
                ],
                "creative_decisions": [
                  "The lighting design enhances the feeling of a secluded, twilight-lit outpost, drawing focus primarily to the bright data readouts and the tension on the character's face."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 1,
                "shot_index": 12,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [],
            "forbidden_refs": [
              {
                "kind": "background",
                "reason": "close framing skips chain_bg reference"
              }
            ],
            "readiness_policy": "skipped_by_policy",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": true,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "8e16b656f76cc0e8"
      },
      {
        "scene_index": 2,
        "heading": "INT. 산악 접촉소 — 황혼",
        "beat_title": "모니터에 복제된 배터리 경고",
        "representative_moment": "외부 모니터 화면에 복제된 토니의 휴대폰 배터리 2% 경고가 떠 있다.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "close framing on the external monitor screen, neutral low-key ambient lighting",
            "t2i_prompt": "Photorealistic cinematic still. Seen at monitor height and slightly off its central axis, a functional external monitor unit with a solid, austere casing occupies the center of the composition, resting on a dark workspace surface. The monitor screen displays a duplicated phone interface with the text 'BATTERY 2%' sharply legible, exhibiting a faint digital texture. At the frame edges, the blurred boundary of a transparent isolation box and a thin wired connection provide restrained context. Neutral, low-key ambient illumination fills the close space, the subdued twilight atmosphere casting soft shadows while the bright glow of the electronic display stands out sharply against the dim surroundings.",
            "outfit_assignments": [],
            "source_facts": [
              "외부 모니터에 휴대폰과 동일한 BATTERY 2% 화면이 뜬 상태",
              "무선은 차단되고 화면만 외부 모니터에 복제된다."
            ],
            "visual_inferences": [
              "fixed_element external_monitor_prop: functional external monitor unit with a solid, austere casing resting on the workspace surface",
              "fixed_element workspace_ambient_lighting: neutral, low-key ambient illumination fills the enclosed workspace",
              "transparent isolation box and wire visible at the periphery as restrained edge context"
            ],
            "creative_decisions": [
              "close framing slightly off-axis to emphasize the urgency of the battery text while maintaining the physical presence of the monitor casing",
              "faint digital texture applied to the screen content to align with the through_device perception mode",
              "soft focus on the frame-edge elements to prevent them from competing with the primary readout"
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [],
            "reference_phrase_kinds": [],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "cf14f95fc94f2421",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "0fc3d0a4585ce68f",
              "owned_usage_hash": "8de771497838ce3d",
              "validator": "scene_detail_owned_objects.v1.close_skip",
              "violations": []
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 1,
        "dependency_reason": "S1_Shot12 background reference (atmosphere continuity maintained despite close framing skip)",
        "_shot_index": 8,
        "visible_entities": [],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 2,
            "shot_index": 8
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "close",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "At monitor height and slightly off its central axis, the dolly-in pauses with the duplicated phone interface held in the middle of the composition and the monitor casing still visible around it. “BATTERY 2%” remains sharply legible, while the wired isolation setup stays as restrained edge context rather than competing with the screen.",
            "lighting_mood": "Neutral, subdued ambient illumination preserves the monitor's austere interface and the urgency of the low-battery reading.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 0명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "primary_subject_only_fully_visible": "the named primary subject (hand / face / object) is the only fully visible entity in the frame",
                  "other_entities_appearance_options": [
                    "partial visibility — shoulder slice / arm slice / sleeve / fingertips entering the frame edge",
                    "soft background — out-of-focus silhouette via shallow depth of field",
                    "absent — not rendered at all even if listed in visible_entities"
                  ],
                  "forbidden_combinations": [
                    "same character's full body and body-part close-up in the same frame (view mixing)",
                    "two characters both rendered as sharp close-up faces in the same frame"
                  ],
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "wide_medium_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=close — wide/medium 규칙은 이 샷에 해당하지 않는다"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": null,
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [],
            "allowed_outlook_pairs": [],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": true,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "skipped_close_framing",
            "bg_id": "L11B02",
            "reference_usage": "skipped_close_framing",
            "owned_objects": [],
            "camera_reference": null,
            "close_framing_skips_background_ref": true,
            "constraints": [
              "this shot uses close framing — the background reference is intentionally omitted at composition time",
              "describe the close subject and its immediate surroundings without relying on the absent background reference",
              "use [L##: ...] block for frame-edge surface description"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "external_monitor_prop",
                "element_type": "persistent_prop",
                "character_name": "",
                "description": "A functional external monitor unit with a solid, austere casing resting on the workspace surface. It serves as the primary readout for the isolated device, maintaining a consistent physical presence in the room whether viewed directly from the front or from an angled side view near a character.",
                "element_scope": "full",
                "applies_to_shots": [
                  8,
                  11
                ],
                "source_facts": [
                  "monitor casing still visible around it",
                  "monitor's side edge soft in the near foreground"
                ],
                "visual_inferences": [
                  "The monitor is securely positioned on a desk or console within the contact post to allow easy viewing."
                ],
                "creative_decisions": [
                  "The monitor's physical design looks utilitarian and slightly rugged to fit the overall aesthetic of the mountain outpost setting."
                ],
                "confidence": "high"
              },
              {
                "element_id": "workspace_ambient_lighting",
                "element_type": "environment_state",
                "character_name": "",
                "description": "Neutral, low-key ambient illumination fills the enclosed workspace. The subdued twilight atmosphere casts soft shadows and ensures that the bright text of electronic displays stands out sharply against the dim surroundings without any unmotivated color sources.",
                "element_scope": "full",
                "applies_to_shots": [
                  3,
                  8,
                  11
                ],
                "source_facts": [
                  "Neutral low-key ambient light appropriate to the enclosed work area",
                  "Neutral, subdued ambient illumination preserves the monitor's austere interface",
                  "Neutral low-key ambient light shapes Wilma's tense features without introducing an unmotivated color source"
                ],
                "visual_inferences": [
                  "The primary light sources are practical dim overheads combined with the glow spilling from the operational screens."
                ],
                "creative_decisions": [
                  "The lighting design enhances the feeling of a secluded, twilight-lit outpost, drawing focus primarily to the bright data readouts and the tension on the character's face."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 1,
                "shot_index": 12,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [],
            "forbidden_refs": [
              {
                "kind": "background",
                "reason": "close framing skips chain_bg reference"
              }
            ],
            "readiness_policy": "skipped_by_policy",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": true,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "e317604af4482054"
      },
      {
        "scene_index": 2,
        "heading": "S2. INT. 산악 접촉소 — 황혼 (00:45–01:30)",
        "beat_title": "미래의 비상 규격과 호환된 과거의 잔재",
        "representative_moment": "Wilma staring at the monitor with a furrowed brow, realizing the implications of the ancient phone's sync request.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "tight three-quarter facial close-up from slightly above eye level, neutral low-key twilight ambient",
            "t2i_prompt": "Photorealistic cinematic still. Seen from slightly above eye level beside the external monitor, tightened into a facial close-up, C02O02 in a green tactical uniform, a Caucasian woman in her late 20s, occupies the middle-right of the frame. She appears at three-quarter angle from the left, leaning subtly toward the left, her furrowed brow catching the neutral, low-key ambient illumination. Her eyes are fixed off-screen to the left, shaping tense features in the subdued twilight atmosphere that casts soft shadows. In the soft near-left foreground, the solid, austere casing of an external monitor rests out of focus, its near side edge visible while the content face is angled away. The space behind her dissolves into [: a dim enclosed workspace with neutral shadows], preserving the isolated room's depth without unmotivated color sources.",
            "outfit_assignments": [
              {
                "character_id": "C02",
                "outlook_id": "O02"
              }
            ],
            "source_facts": [
              "shot description: Wilma looking at the monitor with a furrowed brow",
              "beat information: Wilma realizes the ancient phone received an external request compatible with current emergency specs"
            ],
            "visual_inferences": [
              "fixed_element external_monitor_prop: monitor side edge placed in the soft near foreground to anchor her off-screen eyeline",
              "fixed_element workspace_ambient_lighting: neutral low-key ambient light shaping tense features without unmotivated color sources"
            ],
            "creative_decisions": [
              "tight framing on the face emphasizing the realization and tension",
              "monitor blurred in foreground to establish the object of focus while keeping the character as the primary subject",
              "omitted background reference as per close framing rule, substituting with a generic frame-edge depth descriptor"
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [],
            "reference_phrase_kinds": [
              "character"
            ],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "35ade37292c1794d",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "ffd5298173817340",
              "owned_usage_hash": "8de771497838ce3d",
              "validator": "scene_detail_owned_objects.v1.close_skip",
              "violations": []
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 2,
        "dependency_reason": "previous_shot_refs requires maintaining the same workspace_ambient_lighting and external_monitor_prop states established in shot 8",
        "_shot_index": 11,
        "visible_entities": [
          "C02"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 2,
            "shot_index": 11
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "close",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "The lateral track finishes beside the external monitor at slightly above Wilma's eye level, tightening into a three-quarter facial close-up with the monitor's side edge soft in the near foreground. Wilma occupies the middle-right of frame, leaning subtly toward the unseen display at left as her furrowed brow and fixed off-screen eyeline register the implication of the data.",
            "lighting_mood": "Neutral low-key ambient light shapes Wilma's tense features without introducing an unmotivated color source.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 1명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "primary_subject_only_fully_visible": "the named primary subject (hand / face / object) is the only fully visible entity in the frame",
                  "other_entities_appearance_options": [
                    "partial visibility — shoulder slice / arm slice / sleeve / fingertips entering the frame edge",
                    "soft background — out-of-focus silhouette via shallow depth of field",
                    "absent — not rendered at all even if listed in visible_entities"
                  ],
                  "forbidden_combinations": [
                    "same character's full body and body-part close-up in the same frame (view mixing)",
                    "two characters both rendered as sharp close-up faces in the same frame"
                  ],
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "wide_medium_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=close — wide/medium 규칙은 이 샷에 해당하지 않는다"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": null,
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "C02"
            ],
            "allowed_outlook_pairs": [
              {
                "character_id": "C02",
                "outlook_id": "O02"
              }
            ],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": false,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "skipped_close_framing",
            "bg_id": "L11B02",
            "reference_usage": "skipped_close_framing",
            "owned_objects": [],
            "camera_reference": null,
            "close_framing_skips_background_ref": true,
            "constraints": [
              "this shot uses close framing — the background reference is intentionally omitted at composition time",
              "describe the close subject and its immediate surroundings without relying on the absent background reference",
              "use [L##: ...] block for frame-edge surface description"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "external_monitor_prop",
                "element_type": "persistent_prop",
                "character_name": "",
                "description": "A functional external monitor unit with a solid, austere casing resting on the workspace surface. It serves as the primary readout for the isolated device, maintaining a consistent physical presence in the room whether viewed directly from the front or from an angled side view near a character.",
                "element_scope": "full",
                "applies_to_shots": [
                  8,
                  11
                ],
                "source_facts": [
                  "monitor casing still visible around it",
                  "monitor's side edge soft in the near foreground"
                ],
                "visual_inferences": [
                  "The monitor is securely positioned on a desk or console within the contact post to allow easy viewing."
                ],
                "creative_decisions": [
                  "The monitor's physical design looks utilitarian and slightly rugged to fit the overall aesthetic of the mountain outpost setting."
                ],
                "confidence": "high"
              },
              {
                "element_id": "workspace_ambient_lighting",
                "element_type": "environment_state",
                "character_name": "",
                "description": "Neutral, low-key ambient illumination fills the enclosed workspace. The subdued twilight atmosphere casts soft shadows and ensures that the bright text of electronic displays stands out sharply against the dim surroundings without any unmotivated color sources.",
                "element_scope": "full",
                "applies_to_shots": [
                  3,
                  8,
                  11
                ],
                "source_facts": [
                  "Neutral low-key ambient light appropriate to the enclosed work area",
                  "Neutral, subdued ambient illumination preserves the monitor's austere interface",
                  "Neutral low-key ambient light shapes Wilma's tense features without introducing an unmotivated color source"
                ],
                "visual_inferences": [
                  "The primary light sources are practical dim overheads combined with the glow spilling from the operational screens."
                ],
                "creative_decisions": [
                  "The lighting design enhances the feeling of a secluded, twilight-lit outpost, drawing focus primarily to the bright data readouts and the tension on the character's face."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 2,
                "shot_index": 8,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "character_outlook",
                "id": "C02O02",
                "policy": "required"
              }
            ],
            "forbidden_refs": [
              {
                "kind": "background",
                "reason": "close framing skips chain_bg reference"
              }
            ],
            "readiness_policy": "skipped_by_policy",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": true,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "23ede57e7baa792c"
      },
      {
        "scene_index": 3,
        "heading": "INT. 산악 접촉소 — 황혼 (01:30–02:20)",
        "beat_title": "2026년 작업자의 증명",
        "representative_moment": "C01 holds up P21 toward the hologram to prove his legacy worker status.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "oblique medium composition from slightly below shoulder height, low-key lighting",
            "t2i_prompt": "Photorealistic cinematic still. C01O01 in rugged workwear, a Caucasian male in his 30s with natural facial features, stands in the middle-left foreground, looking toward a monochrome holographic figure, while the ultraphone disc rests in the lower-right midground. Seen from slightly below shoulder height on the left, the oblique medium composition frames C01O01 in right profile at frame left, his chest and upper torso maintaining a rigid, braced posture that conveys physical tension. He holds P21 forward near the lower center; the damaged identification face of P21 is angled partly toward the hologram and partly toward the camera, keeping its cracked condition readable. The monochrome holographic figure of a man stands at three-quarter angle from the left at frame right, leaning into the address with eyes fixed on C01O01. The operating projection disc sits beneath the hologram, its upper face angled diagonally away from the camera. The interior is illuminated by a neutral, muted low-key ambient light reflecting the dusk setting, where deep, restrained shadows softly highlight a background workstation and a cabinet without introducing bright sources.",
            "outfit_assignments": [
              {
                "character_id": "C01",
                "outlook_id": "O01"
              }
            ],
            "source_facts": [
              "Tony holds out his cracked identification card (S3 action)",
              "Tony looks at Bill Hern's hologram while holding the card (shot description)",
              "Bill Hern leans into the address as a monochrome hologram (staging)"
            ],
            "visual_inferences": [
              "fixed_element tony_braced_chest: C01's chest and upper torso maintain a rigid, braced posture.",
              "fixed_element outpost_dusk_lighting: neutral, muted low-key ambient light reflecting the dusk setting.",
              "Since Bill Hern's ID is not an allowed base entity, he is represented generically as 'a monochrome holographic figure of a man'."
            ],
            "creative_decisions": [
              "Medium framing with oblique composition from below shoulder height to emphasize the spatial relationship and tension between the physical worker and the digital projection.",
              "Restrained lighting ensures the monochrome holographic outline stands out as the primary technological accent."
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [
              "fsc_001",
              "fsc_002"
            ],
            "reference_phrase_kinds": [
              "character"
            ],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "anchor",
                "source_phrase": "and a cabinet"
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "anchor",
                "source_phrase": "The operating projection disc sits beneath the hologram"
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "anchor",
                "source_phrase": "highlight a background workstation"
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "310e08c12c376946",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "5436b756bee00f2a",
              "owned_usage_hash": "7a8f5d5dcd1b48e9",
              "validator": "scene_detail_owned_objects.v1",
              "violations": [
                {
                  "owned_object": "cabinet",
                  "violating_phrase": "and a cabinet",
                  "reason": "The cabinet is referenced solely as an existing background element within the ambient lighting description, matching the producer anchor declaration.",
                  "verdict": "anchor_reference"
                },
                {
                  "owned_object": "projection disc",
                  "violating_phrase": "The operating projection disc sits beneath the hologram, its upper face angled diagonally away from the camera.",
                  "reason": "The projection disc is referenced as an existing spatial anchor for the hologram position and orientation, matching the producer anchor declaration.",
                  "verdict": "anchor_reference"
                },
                {
                  "owned_object": "workstation",
                  "violating_phrase": "highlight a background workstation",
                  "reason": "The workstation is mentioned only as a background anchor receiving ambient dusk illumination, matching the producer anchor declaration.",
                  "verdict": "anchor_reference"
                }
              ]
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 2,
        "dependency_reason": "Background consistency and continuous dusk lighting from the prior shot (S2_Shot8) must be maintained.",
        "_shot_index": 8,
        "visible_entities": [
          "C01",
          "P21"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 3,
            "shot_index": 8
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "medium",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "From Tony's left side slightly below shoulder height, the dolly-in settles on his right-facing profile at frame left, with the cracked identification card held near the lower center and Bill Hern's projected figure above the disc at frame right. Tony's gaze remains locked on Bill while Bill addresses him across the shared projection space, allowing the card, skeptical exchange, and spatial hierarchy to read in one oblique composition.",
            "lighting_mood": "Muted low-key ambient rendering preserves the monochrome holographic outline as the shot's restrained technological accent.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 1명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=medium — close 규칙은 이 샷에 해당하지 않는다"
                },
                "wide_medium_rules": {
                  "primary_and_secondary_both_visible": "primary subject + secondary entities all renderable",
                  "body_part_close_up_forbidden": "body-part close-up of any character is forbidden in wide/medium framing — close-up belongs in a separate shot",
                  "fg_bg_separation_requires_shared_anchor": "if two characters are rendered fg/bg separated, also apply fg_bg_shared_anchor_rule",
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": {
              "reason": "shared_space_relation",
              "constraints": [
                {
                  "constraint_id": "fsc_001",
                  "target_kind": "background",
                  "target_id": "",
                  "label": "ultraphone disc",
                  "screen_zone": "lower_right",
                  "depth_plane": "midground",
                  "gesture_action": "none",
                  "gesture_target_label": ""
                },
                {
                  "constraint_id": "fsc_002",
                  "target_kind": "character",
                  "target_id": "C01",
                  "label": "토니(앤서니 로저스)",
                  "screen_zone": "middle_left",
                  "depth_plane": "foreground",
                  "gesture_action": "looks_toward",
                  "gesture_target_label": "Bill Hern hologram"
                }
              ]
            },
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "C01",
              "P21"
            ],
            "allowed_outlook_pairs": [
              {
                "character_id": "C01",
                "outlook_id": "O01"
              }
            ],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": true,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "background_ref_attached",
            "bg_id": "L11B02",
            "reference_usage": "exact_background",
            "owned_objects": [
              "breach ring",
              "cabinet",
              "control panel",
              "debris",
              "door",
              "dust",
              "hatch",
              "isolation unit",
              "lever",
              "light fixture",
              "long gun",
              "map",
              "monitor",
              "projection disc",
              "scanner",
              "smartphone",
              "ventilation duct",
              "workstation"
            ],
            "camera_reference": null,
            "close_framing_skips_background_ref": false,
            "constraints": [
              "owned_objects lists all environment objects already drawn in the background reference — do not create new ones",
              "if focusing on an owned object, declare its reference-kind via the reference_phrase_kinds sidecar (producer contract)"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "tony_braced_chest",
                "element_type": "character_state",
                "character_name": "Tony",
                "description": "Tony's chest and upper torso maintain a rigid, braced posture. He is dressed in functional, rugged clothing typical of a 2026 facility worker, and his upper body conveys physical tension.",
                "element_scope": "close",
                "applies_to_shots": [
                  8,
                  12
                ],
                "source_facts": [
                  "chest braced against muzzle",
                  "braced upper torso"
                ],
                "visual_inferences": [
                  "His attire consists of practical, durable fabrics suitable for a legacy 2026 worker in a harsh mountain environment, showing subtle signs of wear."
                ],
                "creative_decisions": [
                  "Focusing on his chest and shoulder tension to visually represent his resilient stance across different shot framings without relying on facial expressions."
                ],
                "confidence": "high"
              },
              {
                "element_id": "outpost_dusk_lighting",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The interior of the mountain outpost is illuminated by a neutral, muted low-key ambient light that reflects the dusk setting. Deep, restrained shadows dominate the space, softly highlighting shapes without introducing bright or unsupported light sources.",
                "element_scope": "full",
                "applies_to_shots": [
                  8,
                  12,
                  13
                ],
                "source_facts": [
                  "황혼 (01:30–02:20)",
                  "Muted low-key ambient rendering",
                  "Neutral low-key ambient light",
                  "Neutral low-key ambient illumination"
                ],
                "visual_inferences": [
                  "The ambient light serves as a subtle fill from distant or low-power sources, keeping the overall environment shrouded in heavy shadows typical of a remote interior."
                ],
                "creative_decisions": [
                  "Maintaining a consistent low-key exposure to enhance the tense atmosphere and keep the visual focus entirely on the physical interactions and minimal technological interfaces."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 2,
                "shot_index": 8,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "character_outlook",
                "id": "C01O01",
                "policy": "required"
              },
              {
                "kind": "background",
                "id": "L11B02",
                "policy": "required"
              }
            ],
            "forbidden_refs": [],
            "readiness_policy": "block_if_missing",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": false,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "a02bcaa3b4ee03b1"
      },
      {
        "scene_index": 3,
        "heading": "INT. 산악 접촉소 — 황혼 (01:30–02:20)",
        "beat_title": "위협의 완화와 지속",
        "representative_moment": "The muzzle of P14 dipping slightly while resting against the braced chest of C01 in an extreme close-up.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "chest-height insert framing with a descending tilt, isolating the precise contact point",
            "t2i_prompt": "Photorealistic cinematic still. Seen from extremely close to the near side at chest height with a descending tilt, the frame isolates a precise physical contact point. C01 in a rugged functional worker uniform, an American man in his 30s, upper torso occupying the right frame at three-quarter angle from the right, his chest maintaining a rigid, braced posture. P14, a heavy dark metallic long gun barrel, crosses the frame side-on, nearly perpendicular to the camera and clear of its firing axis. The muzzle dips slightly, resting firmly against the fabric of the braced chest. The face and lower body of C01, as well as the weapon holder, remain completely outside the crop. Neutral low-key ambient light reflects softly off the cold mechanical texture of P14. Deep, restrained shadows dominate the out-of-focus interior background, keeping the surrounding outpost shrouded in darkness.",
            "outfit_assignments": [
              {
                "character_id": "C01",
                "outlook_id": "O01"
              }
            ],
            "source_facts": [
              "S3 action: '앨런이 총구를 몇 센티미터 내린다.'",
              "shot description: precise contact point between the rifle muzzle and his chest",
              "camera_direction: At Tony's chest height and extremely close to his near side, descending tilt, muzzle crosses the crop side-on"
            ],
            "visual_inferences": [
              "fixed_element tony_braced_chest: Tony's chest and upper torso maintain a rigid, braced posture in functional, rugged clothing.",
              "fixed_element long_rifle_barrel: heavy, dark metallic barrel with a cold, mechanical texture.",
              "fixed_element outpost_dusk_lighting: neutral, muted low-key ambient light with deep, restrained shadows."
            ],
            "creative_decisions": [
              "Utilizing an insert scale to strictly enforce the absence of faces, directing all visual tension to the physical contact between the weapon and the fabric.",
              "Matching the 'long gun' owned object token directly with P14 to ensure continuity with the background and prop assets."
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [],
            "reference_phrase_kinds": [
              "character",
              "background"
            ],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "redraw",
                "source_phrase": "P14, a heavy dark metallic long gun barrel"
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "1fcfbde9b419c490",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "a1cdacfce6952f00",
              "owned_usage_hash": "71a6b3361a7c20fe",
              "validator": "scene_detail_owned_objects.v1",
              "violations": [
                {
                  "owned_object": "long gun",
                  "violating_phrase": "P14, a heavy dark metallic long gun barrel, crosses the frame side-on, nearly perpendicular to the camera and clear of its firing axis. The muzzle dips slightly, resting firmly against the fabric of the braced chest.",
                  "reason": "The prompt describes the physical contact geometry and barrel orientation of the held long gun resting against the character's chest without inventing a new object or redesigning its fundamental appearance, fulfilling all conditions of narrow exception (2).",
                  "verdict": "allowed_prop_contact"
                }
              ]
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 1,
        "dependency_reason": "exact_background rule applies to maintain the structural continuity of the mountain outpost interior from the prior shot.",
        "_shot_index": 12,
        "visible_entities": [
          "C01",
          "P14"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 3,
            "shot_index": 12
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "insert",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "At Tony's chest height and extremely close to his near side, the descending tilt isolates the precise contact point between the rifle muzzle and his chest while remaining nearly perpendicular to the barrel and clear of its firing axis. Tony's face and the rifle holder stay outside the crop; only enough of his braced upper torso and the barrel remain to make the sustained threat physically legible.",
            "lighting_mood": "Neutral low-key ambient light emphasizes the pressure point and restrained tactile contrast without adding an unsupported light source.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 1명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "primary_subject_only_fully_visible": "the named primary subject (hand / face / object) is the only fully visible entity in the frame",
                  "other_entities_appearance_options": [
                    "partial visibility — shoulder slice / arm slice / sleeve / fingertips entering the frame edge",
                    "soft background — out-of-focus silhouette via shallow depth of field",
                    "absent — not rendered at all even if listed in visible_entities"
                  ],
                  "forbidden_combinations": [
                    "same character's full body and body-part close-up in the same frame (view mixing)",
                    "two characters both rendered as sharp close-up faces in the same frame"
                  ],
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "wide_medium_rules": {
                  "primary_and_secondary_both_visible": "primary subject + secondary entities all renderable",
                  "body_part_close_up_forbidden": "body-part close-up of any character is forbidden in wide/medium framing — close-up belongs in a separate shot",
                  "fg_bg_separation_requires_shared_anchor": "if two characters are rendered fg/bg separated, also apply fg_bg_shared_anchor_rule",
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": null,
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "C01",
              "P14"
            ],
            "allowed_outlook_pairs": [
              {
                "character_id": "C01",
                "outlook_id": "O01"
              }
            ],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": false,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [
              {
                "subject_id": "C01",
                "policy_type": "identity_reference",
                "policy": "base_id_required",
                "reason": "The insert excludes Tony's face and isolates only his chest at the rifle contact point, so an outlook reference cannot align reliably."
              }
            ],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "background_ref_attached",
            "bg_id": "L11B02",
            "reference_usage": "exact_background",
            "owned_objects": [
              "breach ring",
              "cabinet",
              "control panel",
              "debris",
              "door",
              "dust",
              "hatch",
              "isolation unit",
              "lever",
              "light fixture",
              "long gun",
              "map",
              "monitor",
              "projection disc",
              "scanner",
              "smartphone",
              "ventilation duct",
              "workstation"
            ],
            "camera_reference": null,
            "close_framing_skips_background_ref": false,
            "constraints": [
              "owned_objects lists all environment objects already drawn in the background reference — do not create new ones",
              "if focusing on an owned object, declare its reference-kind via the reference_phrase_kinds sidecar (producer contract)"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "tony_braced_chest",
                "element_type": "character_state",
                "character_name": "Tony",
                "description": "Tony's chest and upper torso maintain a rigid, braced posture. He is dressed in functional, rugged clothing typical of a 2026 facility worker, and his upper body conveys physical tension.",
                "element_scope": "close",
                "applies_to_shots": [
                  8,
                  12
                ],
                "source_facts": [
                  "chest braced against muzzle",
                  "braced upper torso"
                ],
                "visual_inferences": [
                  "His attire consists of practical, durable fabrics suitable for a legacy 2026 worker in a harsh mountain environment, showing subtle signs of wear."
                ],
                "creative_decisions": [
                  "Focusing on his chest and shoulder tension to visually represent his resilient stance across different shot framings without relying on facial expressions."
                ],
                "confidence": "high"
              },
              {
                "element_id": "outpost_dusk_lighting",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The interior of the mountain outpost is illuminated by a neutral, muted low-key ambient light that reflects the dusk setting. Deep, restrained shadows dominate the space, softly highlighting shapes without introducing bright or unsupported light sources.",
                "element_scope": "full",
                "applies_to_shots": [
                  8,
                  12,
                  13
                ],
                "source_facts": [
                  "황혼 (01:30–02:20)",
                  "Muted low-key ambient rendering",
                  "Neutral low-key ambient light",
                  "Neutral low-key ambient illumination"
                ],
                "visual_inferences": [
                  "The ambient light serves as a subtle fill from distant or low-power sources, keeping the overall environment shrouded in heavy shadows typical of a remote interior."
                ],
                "creative_decisions": [
                  "Maintaining a consistent low-key exposure to enhance the tense atmosphere and keep the visual focus entirely on the physical interactions and minimal technological interfaces."
                ],
                "confidence": "high"
              },
              {
                "element_id": "long_rifle_barrel",
                "element_type": "persistent_prop",
                "character_name": "",
                "description": "A long rifle with a heavy, dark metallic barrel and muzzle. Its utilitarian and rugged physical design remains a constant visual anchor, retaining a cold, mechanical texture.",
                "element_scope": "close",
                "applies_to_shots": [
                  12,
                  13
                ],
                "source_facts": [
                  "장총",
                  "precise contact point between the rifle muzzle and his chest",
                  "two-handed rifle grip",
                  "muzzle dips slightly"
                ],
                "visual_inferences": [
                  "The weapon features a matte finish that absorbs rather than reflects the low-key ambient light, fitting the practical, non-ornamental armory of a rugged mountain outpost."
                ],
                "creative_decisions": [
                  "Highlighting the harsh, metallic texture of the barrel in close proximity to the subjects to visually enforce the immediate threat and mechanical presence."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 1,
                "shot_index": 9,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "character",
                "id": "C01",
                "policy": "required"
              },
              {
                "kind": "background",
                "id": "L11B02",
                "policy": "required"
              }
            ],
            "forbidden_refs": [],
            "readiness_policy": "block_if_missing",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": false,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "815ea774001cbbaf"
      },
      {
        "scene_index": 3,
        "heading": "S3. INT. 산악 접촉소 — 황혼 (01:30–02:20)",
        "beat_title": "1차적 살해 위협의 완화",
        "representative_moment": "C07 lowers the muzzle of P14 slightly upon seeing the manual seal instruction on the display.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "close high-angle over-the-shoulder, looking down at hands and display, neutral low-key contrast",
            "t2i_prompt": "Photorealistic cinematic still. Seen from close above and three-quarters behind the working side, looking down past the right shoulder. C07O04 in a gray mountain combat uniform, a Caucasian adult man, occupies the near frame from the chest up, his shoulders pitched over a console. His bowed eyeline, visible in partial right profile, is fixed on the active content face below. His hands maintain a two-handed grip on P14, a long rifle with a heavy, dark metallic barrel and muzzle, its cold mechanical texture catching faint highlights as it angles away and dips slightly toward the lower edge. Just beyond the muzzle, [: a glowing records display surface] lies directly beneath his downward eyeline, clearly showing a 2026 facility drawing and a manual seal instruction through digital noise. The composition is illuminated by neutral, muted low-key ambient light reflecting the dusk setting, with deep, restrained shadows holding the hands and weapon in subtle contrast against the readable display.",
            "outfit_assignments": [
              {
                "character_id": "C07",
                "outlook_id": "O04"
              }
            ],
            "source_facts": [
              "S3 text: Alan looks at the facility drawing reading 'MANUAL DUAL-ACTION SEAL' and lowers the muzzle by a few centimeters.",
              "shot description: Alan's two hands grip the long rifle, the muzzle dipping slightly toward the lower frame.",
              "camera staging: over_shoulder, Alan's shoulders pitched over console, gaze=down."
            ],
            "visual_inferences": [
              "fixed_element outpost_dusk_lighting: neutral low-key ambient light reflecting dusk, deep restrained shadows.",
              "fixed_element long_rifle_barrel: heavy dark metallic barrel and muzzle, cold mechanical texture.",
              "Alan's face is rendered in partial profile to satisfy the face identifiability rule for the C## ID while maintaining the required over-the-shoulder camera direction."
            ],
            "creative_decisions": [
              "Utilized a close framing scale to skip the wider room background, tightly anchoring the spatial relationship between Alan's yielding grip on the rifle and the verification display at the bottom frame edge.",
              "Mapped the neutral low-key illumination to create a stark contrast between the rugged, shadowed weapon and the luminous text of the legacy digital interface."
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [],
            "reference_phrase_kinds": [
              "character"
            ],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "642b17879e2fdd15",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "e317de18c744d24b",
              "owned_usage_hash": "8de771497838ce3d",
              "validator": "scene_detail_owned_objects.v1.close_skip",
              "violations": []
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": -1,
        "dependency_reason": "",
        "_shot_index": 13,
        "visible_entities": [
          "C07",
          "P14"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 3,
            "shot_index": 13
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "close",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "After the logical cut, the camera looks down from close above and three-quarters behind Alan's working side, beginning on the facility drawing before the tilt reaches his two-handed rifle grip. The frame locks at the instant his hands yield and the muzzle dips slightly toward the lower edge, while his bowed eyeline remains directed toward the verification display beyond the rifle.",
            "lighting_mood": "Neutral low-key ambient illumination keeps the verification display readable while holding Alan's hands and weapon in restrained contrast.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 1명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "primary_subject_only_fully_visible": "the named primary subject (hand / face / object) is the only fully visible entity in the frame",
                  "other_entities_appearance_options": [
                    "partial visibility — shoulder slice / arm slice / sleeve / fingertips entering the frame edge",
                    "soft background — out-of-focus silhouette via shallow depth of field",
                    "absent — not rendered at all even if listed in visible_entities"
                  ],
                  "forbidden_combinations": [
                    "same character's full body and body-part close-up in the same frame (view mixing)",
                    "two characters both rendered as sharp close-up faces in the same frame"
                  ],
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "wide_medium_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=close — wide/medium 규칙은 이 샷에 해당하지 않는다"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": null,
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "C07",
              "P14"
            ],
            "allowed_outlook_pairs": [
              {
                "character_id": "C07",
                "outlook_id": "O04"
              }
            ],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": true,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "skipped_close_framing",
            "bg_id": "L11B02",
            "reference_usage": "skipped_close_framing",
            "owned_objects": [],
            "camera_reference": null,
            "close_framing_skips_background_ref": true,
            "constraints": [
              "this shot uses close framing — the background reference is intentionally omitted at composition time",
              "describe the close subject and its immediate surroundings without relying on the absent background reference",
              "use [L##: ...] block for frame-edge surface description"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "outpost_dusk_lighting",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The interior of the mountain outpost is illuminated by a neutral, muted low-key ambient light that reflects the dusk setting. Deep, restrained shadows dominate the space, softly highlighting shapes without introducing bright or unsupported light sources.",
                "element_scope": "full",
                "applies_to_shots": [
                  8,
                  12,
                  13
                ],
                "source_facts": [
                  "황혼 (01:30–02:20)",
                  "Muted low-key ambient rendering",
                  "Neutral low-key ambient light",
                  "Neutral low-key ambient illumination"
                ],
                "visual_inferences": [
                  "The ambient light serves as a subtle fill from distant or low-power sources, keeping the overall environment shrouded in heavy shadows typical of a remote interior."
                ],
                "creative_decisions": [
                  "Maintaining a consistent low-key exposure to enhance the tense atmosphere and keep the visual focus entirely on the physical interactions and minimal technological interfaces."
                ],
                "confidence": "high"
              },
              {
                "element_id": "long_rifle_barrel",
                "element_type": "persistent_prop",
                "character_name": "",
                "description": "A long rifle with a heavy, dark metallic barrel and muzzle. Its utilitarian and rugged physical design remains a constant visual anchor, retaining a cold, mechanical texture.",
                "element_scope": "close",
                "applies_to_shots": [
                  12,
                  13
                ],
                "source_facts": [
                  "장총",
                  "precise contact point between the rifle muzzle and his chest",
                  "two-handed rifle grip",
                  "muzzle dips slightly"
                ],
                "visual_inferences": [
                  "The weapon features a matte finish that absorbs rather than reflects the low-key ambient light, fitting the practical, non-ornamental armory of a rugged mountain outpost."
                ],
                "creative_decisions": [
                  "Highlighting the harsh, metallic texture of the barrel in close proximity to the subjects to visually enforce the immediate threat and mechanical presence."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 2,
                "shot_index": 8,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "character_outlook",
                "id": "C07O04",
                "policy": "required"
              }
            ],
            "forbidden_refs": [
              {
                "kind": "background",
                "reason": "close framing skips chain_bg reference"
              }
            ],
            "readiness_policy": "skipped_by_policy",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": true,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "86a01a461161eb59"
      },
      {
        "scene_index": 4,
        "heading": "INT. 산악 접촉소 — 황혼",
        "beat_title": "H.A.N.의 기원과 변질 확인",
        "representative_moment": "A blue monochrome hologram of Hart projects from the P20 ultraphone disc.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "Low angle diagonal observer view",
            "t2i_prompt": "Photorealistic cinematic still. Seen from slightly below and looking diagonally upward, the P20 smartphone's ultraphone disc in the lower-center foreground anchors the shot. A blue monochrome holographic Caucasian middle-aged male figure resolves in the upper center at a three-quarter angle from the right. The figure's upper body and the disc below share the frame. The figure is leaning forward mid-address with eyes looking off-screen. The upper face of the projection disc is visible along the bottom edge, receding diagonally beneath the hologram. The explicitly blue holographic emission casts a restrained cool accent across the figure's volume, standing out softly against the muted, low-key twilight illumination of the shadowy interior.",
            "outfit_assignments": [],
            "source_facts": [
              "ultraphone disc on and projecting Hart's hologram",
              "Hart's explicitly blue holographic emission",
              "holographic figure leaning forward mid-address, looking off-screen"
            ],
            "visual_inferences": [
              "fixed_element interior_twilight_lighting: muted, low-key twilight illumination of the shadowy interior",
              "the projection disc provides a physical anchor at the bottom edge for the holographic volume"
            ],
            "creative_decisions": [
              "low camera angle looking diagonally upward emphasizes the spatial relationship between the physical disc and the projection",
              "the blue holographic emission provides the primary light source for the upper frame against the low-key background"
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [
              "fsc_001"
            ],
            "reference_phrase_kinds": [
              "prop",
              "background"
            ],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "redraw",
                "source_phrase": "The upper face of the projection disc is visible along the bottom edge"
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "anchor",
                "source_phrase": "the P20 smartphone's ultraphone disc"
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "1d33a91147072933",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "14049799ead3dbec",
              "owned_usage_hash": "2440ac0587ea02e1",
              "validator": "scene_detail_owned_objects.v1",
              "violations": [
                {
                  "owned_object": "projection disc",
                  "violating_phrase": "The upper face of the projection disc is visible along the bottom edge, receding diagonally beneath the hologram.",
                  "reason": "The prompt references the projection disc strictly for spatial framing and compositional alignment beneath the hologram rather than generating or redesigning a new object, making it an anchor reference contrary to the producer's redraw declaration.",
                  "verdict": "anchor_reference"
                },
                {
                  "owned_object": "smartphone",
                  "violating_phrase": "the P20 smartphone's ultraphone disc in the lower-center foreground anchors the shot",
                  "reason": "The smartphone is mentioned purely as an anchor and spatial descriptor for scene framing, matching the producer's anchor declaration.",
                  "verdict": "anchor_reference"
                }
              ]
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 2,
        "dependency_reason": "exact_background from S2_Shot8 is preserved for the interior space",
        "_shot_index": 7,
        "visible_entities": [
          "P20"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 4,
            "shot_index": 7
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "medium",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "Slightly below the top of the ultraphone disc at close range, the dolly-in looks diagonally upward with the disc anchored along the lower frame and open projection space above it. Hart's blue monochrome holographic figure resolves in the upper center at a three-quarter angle, leaning into the address and directing attention toward the people outside the frame rather than toward the lens.",
            "lighting_mood": "Hart's explicitly blue holographic emission supplies the restrained cool accent within an otherwise muted, low-key composition.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 0명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=medium — close 규칙은 이 샷에 해당하지 않는다"
                },
                "wide_medium_rules": {
                  "primary_and_secondary_both_visible": "primary subject + secondary entities all renderable",
                  "body_part_close_up_forbidden": "body-part close-up of any character is forbidden in wide/medium framing — close-up belongs in a separate shot",
                  "fg_bg_separation_requires_shared_anchor": "if two characters are rendered fg/bg separated, also apply fg_bg_shared_anchor_rule",
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": {
              "reason": "required_background_position",
              "constraints": [
                {
                  "constraint_id": "fsc_001",
                  "target_kind": "background",
                  "target_id": "",
                  "label": "ultraphone disc",
                  "screen_zone": "lower_center",
                  "depth_plane": "foreground",
                  "gesture_action": "none",
                  "gesture_target_label": ""
                }
              ]
            },
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "P20"
            ],
            "allowed_outlook_pairs": [],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": false,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "background_ref_attached",
            "bg_id": "L11B02",
            "reference_usage": "exact_background",
            "owned_objects": [
              "breach ring",
              "cabinet",
              "control panel",
              "debris",
              "door",
              "dust",
              "hatch",
              "isolation unit",
              "lever",
              "light fixture",
              "long gun",
              "map",
              "monitor",
              "projection disc",
              "scanner",
              "smartphone",
              "ventilation duct",
              "workstation"
            ],
            "camera_reference": null,
            "close_framing_skips_background_ref": false,
            "constraints": [
              "owned_objects lists all environment objects already drawn in the background reference — do not create new ones",
              "if focusing on an owned object, declare its reference-kind via the reference_phrase_kinds sidecar (producer contract)"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "interior_twilight_lighting",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The interior environment features low-key illumination reflecting a twilight setting. The space maintains a muted, cool-toned color palette with restrained contrast, creating a consistently dim and subdued atmosphere where localized light sources can stand out softly.",
                "element_scope": "full",
                "applies_to_shots": [
                  7,
                  10
                ],
                "source_facts": [
                  "muted, low-key composition",
                  "Low-key illumination appropriate to the twilight interior",
                  "preserves a cool, muted palette and restrained contrast"
                ],
                "visual_inferences": [
                  "The ambient darkness of the basement is uniform and shadowy, naturally framing the glowing technology such as holograms or monitors."
                ],
                "creative_decisions": [
                  "Utilize a consistent cool, shadowy baseline across all angles to emphasize the underground or enclosed nature of the contact point at twilight."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 2,
                "shot_index": 8,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "background",
                "id": "L11B02",
                "policy": "required"
              },
              {
                "kind": "prop",
                "id": "P20",
                "policy": "required"
              }
            ],
            "forbidden_refs": [],
            "readiness_policy": "block_if_missing",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [
            {
              "contract_id": "rc_001",
              "scope": {
                "scene_index": 4,
                "shot_indices": [
                  7
                ],
                "duration": "single_shot"
              },
              "targets": [
                {
                  "entity_id": "P20",
                  "role": "visual_target"
                }
              ],
              "requirements": [
                {
                  "dimension": "visual_identity",
                  "operation": "preserve",
                  "strength": "required",
                  "reference_policy": "use_entity_reference"
                }
              ]
            }
          ],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": false,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "35e9d3387d5d2c5a"
      },
      {
        "scene_index": 4,
        "heading": "S4. INT. 산악 접촉소 — 황혼",
        "beat_title": "과거 시스템과의 연관성 추론",
        "representative_moment": "C01 points to the enlarged authority text on the monitor behind him while locking eyes with the other figure.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "over-the-shoulder upper-chest height framing with low-key twilight lighting",
            "t2i_prompt": "Photorealistic cinematic still. Seen from upper-chest height, C01O01 in a dark work jacket, an American man in his 30s, occupies the center-left frame from the chest up at a three-quarter angle from the right. His torso is turning and his arm is extended, his pointing finger connected compositionally to the monitor screen behind him. His eyes are fixed on a Caucasian middle-aged man whose near shoulder and partial profile slice the right edge of the foreground in an over-the-shoulder composition. The man in the foreground leans into the gesture, his gaze returning to C01O01. [L11: A glowing monitor sits in the dim background, its content-bearing face oblique but readable to the camera, the enlarged authority line visible just behind the extended hand, surrounded by the shadowy interior surfaces.] Low-key illumination appropriate to the twilight interior preserves a cool, muted palette and restrained contrast around the localized screen glow.",
            "outfit_assignments": [
              {
                "character_id": "C01",
                "outlook_id": "O01"
              }
            ],
            "source_facts": [
              "Tony enlarges the authority text on the screen.",
              "Tony points at the text while looking at Alan.",
              "Alan leans into the exchange."
            ],
            "visual_inferences": [
              "Chest-up framing to capture Tony's upper body and hand gesture together.",
              "Alan is positioned as a foreground shoulder slice to establish the over-the-shoulder dynamic.",
              "The monitor's text is placed in the background right behind Tony's pointing hand to link the gesture and the evidence."
            ],
            "creative_decisions": [
              "Utilized a low-key twilight illumination to emphasize the glowing monitor text while keeping the rest of the palette cool and muted.",
              "Maintained a tight spatial relationship between the pointing hand and the screen to ensure the connection is readable in a single frame."
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [],
            "reference_phrase_kinds": [
              "character"
            ],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "546a8bebdafd3989",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "974e986411de48c8",
              "owned_usage_hash": "8de771497838ce3d",
              "validator": "scene_detail_owned_objects.v1.close_skip",
              "violations": []
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 3,
        "dependency_reason": "Shares the same interior environment and twlight lighting tone as the previous location.",
        "_shot_index": 10,
        "visible_entities": [
          "C01",
          "P01",
          "L11"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 4,
            "shot_index": 10
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "close",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "Settle at Tony's upper-chest height on his monitor-side three-quarter angle, with his upper body occupying the center-left and Alan's near shoulder and partial profile contained at the right edge. Tony's extended finger remains connected compositionally to the enlarged authority text behind him while his eyes lock onto Alan; Alan leans into the exchange and returns Tony's look, making the tightened camera distance the shot's principal change.",
            "lighting_mood": "Low-key illumination appropriate to the twilight interior preserves a cool, muted palette and restrained contrast around the readable monitor text.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 1명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "primary_subject_only_fully_visible": "the named primary subject (hand / face / object) is the only fully visible entity in the frame",
                  "other_entities_appearance_options": [
                    "partial visibility — shoulder slice / arm slice / sleeve / fingertips entering the frame edge",
                    "soft background — out-of-focus silhouette via shallow depth of field",
                    "absent — not rendered at all even if listed in visible_entities"
                  ],
                  "forbidden_combinations": [
                    "same character's full body and body-part close-up in the same frame (view mixing)",
                    "two characters both rendered as sharp close-up faces in the same frame"
                  ],
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "wide_medium_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=close — wide/medium 규칙은 이 샷에 해당하지 않는다"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": null,
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "C01",
              "P01",
              "L11"
            ],
            "allowed_outlook_pairs": [
              {
                "character_id": "C01",
                "outlook_id": "O01"
              }
            ],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": true,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "skipped_close_framing",
            "bg_id": "L11B02",
            "reference_usage": "skipped_close_framing",
            "owned_objects": [],
            "camera_reference": null,
            "close_framing_skips_background_ref": true,
            "constraints": [
              "this shot uses close framing — the background reference is intentionally omitted at composition time",
              "describe the close subject and its immediate surroundings without relying on the absent background reference",
              "use [L##: ...] block for frame-edge surface description"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "interior_twilight_lighting",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The interior environment features low-key illumination reflecting a twilight setting. The space maintains a muted, cool-toned color palette with restrained contrast, creating a consistently dim and subdued atmosphere where localized light sources can stand out softly.",
                "element_scope": "full",
                "applies_to_shots": [
                  7,
                  10
                ],
                "source_facts": [
                  "muted, low-key composition",
                  "Low-key illumination appropriate to the twilight interior",
                  "preserves a cool, muted palette and restrained contrast"
                ],
                "visual_inferences": [
                  "The ambient darkness of the basement is uniform and shadowy, naturally framing the glowing technology such as holograms or monitors."
                ],
                "creative_decisions": [
                  "Utilize a consistent cool, shadowy baseline across all angles to emphasize the underground or enclosed nature of the contact point at twilight."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 3,
                "shot_index": 13,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "character_outlook",
                "id": "C01O01",
                "policy": "required"
              }
            ],
            "forbidden_refs": [
              {
                "kind": "background",
                "reason": "close framing skips chain_bg reference"
              }
            ],
            "readiness_policy": "skipped_by_policy",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": true,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "bacf9a17f73e4694"
      },
      {
        "scene_index": 5,
        "heading": "S5. INT. 산악 접촉소 — 황혼 (03:10–04:15)",
        "beat_title": "decision",
        "representative_moment": "A close-up of the UNSENT — 393 YEARS AGO popup on the monitor screen.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "Insert oblique close view at monitor height, cool low-key lighting.",
            "t2i_prompt": "Photorealistic cinematic still. Seen from monitor height in an oblique close view beside one screen corner, the UNSENT notification area is displayed in the middle-center midground. The glowing digital popup with the text UNSENT — 393 YEARS AGO occupies most of the central field without losing the physical monitor edge, which angles away as a narrow side margin. Cool, low-key interior illumination from the screen casts a faint light against the heavy shadows of the room, keeping the stark, minimal interface legible with restrained contrast and no other emphasized light source.",
            "outfit_assignments": [],
            "source_facts": [
              "UNSENT — 393 YEARS AGO notification occupies most of the central field without losing the monitor edge",
              "Cool, low-key interior illumination keeps the notification legible with restrained contrast"
            ],
            "visual_inferences": [
              "fixed_element unsent_message_popup: The text is clearly legible against a stark, minimal computer interface.",
              "fixed_element outpost_low_key_lighting: The space is heavily shadowed, relying on dim practical illumination."
            ],
            "creative_decisions": [
              "Oblique framing emphasizes the digital notification while keeping the physical hardware edge visible for scale and depth.",
              "The screen acts as the sole practical light source within the frame to isolate the critical narrative information."
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [
              "fsc_001"
            ],
            "reference_phrase_kinds": [
              "background"
            ],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "redraw",
                "source_phrase": "physical monitor edge"
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "124dca4dac93b3de",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "d5a818caac5fe46f",
              "owned_usage_hash": "c11ae30b45f52025",
              "validator": "scene_detail_owned_objects.v1",
              "violations": [
                {
                  "owned_object": "monitor",
                  "violating_phrase": "UNSENT notification area is displayed in the middle-center midground. The glowing digital popup with the text UNSENT — 393 YEARS AGO occupies most of the central field without losing the physical monitor edge",
                  "reason": "The prompt depicts the glowing notification popup content on the monitor screen while framing its physical edge, which is explicitly mandated by shot_intent ('A close-up of the UNSENT — 393 YEARS AGO popup on the monitor screen.') and meets all criteria for allowed visual state change.",
                  "verdict": "allowed_visual_state_change"
                }
              ]
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 2,
        "dependency_reason": "Background L11B02 continuity (S2_Shot8 exact_background).",
        "_shot_index": 7,
        "visible_entities": [],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 5,
            "shot_index": 7
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "insert",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "At monitor height, hold an oblique close view beside one screen corner so the UNSENT — 393 YEARS AGO notification occupies most of the central field without losing the monitor edge. The completed dolly-in isolates the age of the message while leaving a narrow side margin for the following lateral move toward Tony.",
            "lighting_mood": "Cool, low-key interior illumination keeps the notification legible with restrained contrast and no additional emphasized source.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 0명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "primary_subject_only_fully_visible": "the named primary subject (hand / face / object) is the only fully visible entity in the frame",
                  "other_entities_appearance_options": [
                    "partial visibility — shoulder slice / arm slice / sleeve / fingertips entering the frame edge",
                    "soft background — out-of-focus silhouette via shallow depth of field",
                    "absent — not rendered at all even if listed in visible_entities"
                  ],
                  "forbidden_combinations": [
                    "same character's full body and body-part close-up in the same frame (view mixing)",
                    "two characters both rendered as sharp close-up faces in the same frame"
                  ],
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "wide_medium_rules": {
                  "primary_and_secondary_both_visible": "primary subject + secondary entities all renderable",
                  "body_part_close_up_forbidden": "body-part close-up of any character is forbidden in wide/medium framing — close-up belongs in a separate shot",
                  "fg_bg_separation_requires_shared_anchor": "if two characters are rendered fg/bg separated, also apply fg_bg_shared_anchor_rule",
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": {
              "reason": "required_background_position",
              "constraints": [
                {
                  "constraint_id": "fsc_001",
                  "target_kind": "background",
                  "target_id": "",
                  "label": "UNSENT notification area",
                  "screen_zone": "middle_center",
                  "depth_plane": "midground",
                  "gesture_action": "none",
                  "gesture_target_label": ""
                }
              ]
            },
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [],
            "allowed_outlook_pairs": [],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": true,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "background_ref_attached",
            "bg_id": "L11B02",
            "reference_usage": "exact_background",
            "owned_objects": [
              "breach ring",
              "cabinet",
              "control panel",
              "debris",
              "door",
              "dust",
              "hatch",
              "isolation unit",
              "lever",
              "light fixture",
              "long gun",
              "map",
              "monitor",
              "projection disc",
              "scanner",
              "smartphone",
              "ventilation duct",
              "workstation"
            ],
            "camera_reference": null,
            "close_framing_skips_background_ref": false,
            "constraints": [
              "owned_objects lists all environment objects already drawn in the background reference — do not create new ones",
              "if focusing on an owned object, declare its reference-kind via the reference_phrase_kinds sidecar (producer contract)"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "unsent_message_popup",
                "element_type": "persistent_prop",
                "character_name": "",
                "description": "A digital monitor screen displays a distinct notification popup on its corner with the text 'UNSENT — 393 YEARS AGO'. The glowing popup is positioned near the edge of the screen and casts a visible, cool light into the dark surroundings.",
                "element_scope": "close",
                "applies_to_shots": [
                  7,
                  8
                ],
                "source_facts": [
                  "UNSENT — 393 YEARS AGO notification occupies most of the central field without losing the monitor edge",
                  "a sliver of the notification remains outside his direct eyeline",
                  "The notification's light passes across Tony's conflicted face"
                ],
                "visual_inferences": [
                  "The notification emits a strong enough glow to act as a localized practical light source.",
                  "The text is clearly legible against a stark, minimal computer interface."
                ],
                "creative_decisions": [
                  "The popup is designed with utilitarian typography fitting a centuries-old legacy computer system."
                ],
                "confidence": "high"
              },
              {
                "element_id": "outpost_low_key_lighting",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The interior environment is characterized by cool, muted, and low-key ambient lighting. The space is heavily shadowed with restrained contrast, lacking any overwhelming main light source and relying on dim practical illumination.",
                "element_scope": "full",
                "applies_to_shots": [
                  7,
                  8,
                  11
                ],
                "source_facts": [
                  "Cool, low-key interior illumination",
                  "cool, muted, low-key interior",
                  "Low-key ambient interior illumination"
                ],
                "visual_inferences": [
                  "The space feels confined and shadowy, typical of an underground or sealed bunker setting.",
                  "Overall color temperature leans toward cool, clinical blues or grays."
                ],
                "creative_decisions": [
                  "The dim ambient lighting enhances the sense of isolation and emphasizes localized light sources like the monitor screen."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 2,
                "shot_index": 8,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "background",
                "id": "L11B02",
                "policy": "required"
              }
            ],
            "forbidden_refs": [],
            "readiness_policy": "block_if_missing",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": false,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "b21d17d9e9e54f93"
      },
      {
        "scene_index": 5,
        "heading": "S5. INT. 산악 접촉소 — 황혼 (03:10–04:15)",
        "beat_title": "decision",
        "representative_moment": "팝업 창의 불빛을 받으며 입을 꾹 다문 채 고뇌하는 C01의 얼굴",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "Tight close-up from slightly below the eye line with low-key, localized screen illumination.",
            "t2i_prompt": "Photorealistic cinematic still. C01O01, a Caucasian man in his 30s in a dark disaster relief uniform representing Tony (Anthony Rogers), is positioned in the middle-center foreground, looking toward the off-frame message notification. Seen from slightly below his eye line on the monitor side, his face at three-quarter left angle occupies most of the frame. His jaw is firmly set, lips compressed, and shoulders braced. A partial digital monitor screen is visible at the frame edge, displaying a distinct glowing popup with the text 'UNSENT — 393 YEARS AGO'. The localized cool light from this popup illuminates his face against [: a cool, muted, and heavily shadowed low-key interior] that forms a soft, out-of-focus backdrop.",
            "outfit_assignments": [
              {
                "character_id": "C01",
                "outlook_id": "O01"
              }
            ],
            "source_facts": [
              "The notification's light passes across Tony's conflicted face within an otherwise cool, muted, low-key interior.",
              "Tony is in three-quarter left angle, jaw set, shoulders braced, looking off-screen.",
              "Track beside Tony from slightly below his eye line on the monitor side, tightening until his three-quarter face occupies most of the frame.",
              "A sliver of the notification remains outside his direct eyeline."
            ],
            "visual_inferences": [
              "fixed_element unsent_message_popup: glowing popup on monitor edge casting cool light",
              "fixed_element outpost_low_key_lighting: cool, muted, heavily shadowed space",
              "The tight framing isolates Tony's face to emphasize his internal decision while keeping the source of light at the edge."
            ],
            "creative_decisions": [
              "Utilized a soft depth of field for the background to firmly anchor the primary subject in the foreground.",
              "Translated the contract label into 'Tony (Anthony Rogers)' to integrate the specific scene context naturally."
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [
              "fsc_001"
            ],
            "reference_phrase_kinds": [
              "character"
            ],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "c2d3558ae684042e",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "3897222b17f8f354",
              "owned_usage_hash": "8de771497838ce3d",
              "validator": "scene_detail_owned_objects.v1.close_skip",
              "violations": []
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": -1,
        "dependency_reason": "",
        "_shot_index": 8,
        "visible_entities": [
          "C01"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 5,
            "shot_index": 8
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "close",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "Track beside Tony from slightly below his eye line on the monitor side, tightening until his three-quarter face occupies most of the frame while a sliver of the notification remains outside his direct eyeline. His jaw is set and his lips remain compressed as he stares toward the off-frame popup, holding the decision in stillness rather than addressing the lens.",
            "lighting_mood": "The notification's light passes across Tony's conflicted face within an otherwise cool, muted, low-key interior.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 1명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "primary_subject_only_fully_visible": "the named primary subject (hand / face / object) is the only fully visible entity in the frame",
                  "other_entities_appearance_options": [
                    "partial visibility — shoulder slice / arm slice / sleeve / fingertips entering the frame edge",
                    "soft background — out-of-focus silhouette via shallow depth of field",
                    "absent — not rendered at all even if listed in visible_entities"
                  ],
                  "forbidden_combinations": [
                    "same character's full body and body-part close-up in the same frame (view mixing)",
                    "two characters both rendered as sharp close-up faces in the same frame"
                  ],
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "wide_medium_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=close — wide/medium 규칙은 이 샷에 해당하지 않는다"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": {
              "reason": "primary_subject_isolation",
              "constraints": [
                {
                  "constraint_id": "fsc_001",
                  "target_kind": "character",
                  "target_id": "C01",
                  "label": "토니(앤서니 로저스)",
                  "screen_zone": "middle_center",
                  "depth_plane": "foreground",
                  "gesture_action": "looks_toward",
                  "gesture_target_label": "off-frame message notification"
                }
              ]
            },
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "C01"
            ],
            "allowed_outlook_pairs": [
              {
                "character_id": "C01",
                "outlook_id": "O01"
              }
            ],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": true,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "skipped_close_framing",
            "bg_id": "L11B02",
            "reference_usage": "skipped_close_framing",
            "owned_objects": [],
            "camera_reference": null,
            "close_framing_skips_background_ref": true,
            "constraints": [
              "this shot uses close framing — the background reference is intentionally omitted at composition time",
              "describe the close subject and its immediate surroundings without relying on the absent background reference",
              "use [L##: ...] block for frame-edge surface description"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "unsent_message_popup",
                "element_type": "persistent_prop",
                "character_name": "",
                "description": "A digital monitor screen displays a distinct notification popup on its corner with the text 'UNSENT — 393 YEARS AGO'. The glowing popup is positioned near the edge of the screen and casts a visible, cool light into the dark surroundings.",
                "element_scope": "close",
                "applies_to_shots": [
                  7,
                  8
                ],
                "source_facts": [
                  "UNSENT — 393 YEARS AGO notification occupies most of the central field without losing the monitor edge",
                  "a sliver of the notification remains outside his direct eyeline",
                  "The notification's light passes across Tony's conflicted face"
                ],
                "visual_inferences": [
                  "The notification emits a strong enough glow to act as a localized practical light source.",
                  "The text is clearly legible against a stark, minimal computer interface."
                ],
                "creative_decisions": [
                  "The popup is designed with utilitarian typography fitting a centuries-old legacy computer system."
                ],
                "confidence": "high"
              },
              {
                "element_id": "outpost_low_key_lighting",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The interior environment is characterized by cool, muted, and low-key ambient lighting. The space is heavily shadowed with restrained contrast, lacking any overwhelming main light source and relying on dim practical illumination.",
                "element_scope": "full",
                "applies_to_shots": [
                  7,
                  8,
                  11
                ],
                "source_facts": [
                  "Cool, low-key interior illumination",
                  "cool, muted, low-key interior",
                  "Low-key ambient interior illumination"
                ],
                "visual_inferences": [
                  "The space feels confined and shadowy, typical of an underground or sealed bunker setting.",
                  "Overall color temperature leans toward cool, clinical blues or grays."
                ],
                "creative_decisions": [
                  "The dim ambient lighting enhances the sense of isolation and emphasizes localized light sources like the monitor screen."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 5,
                "shot_index": 7,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "character_outlook",
                "id": "C01O01",
                "policy": "required"
              }
            ],
            "forbidden_refs": [
              {
                "kind": "background",
                "reason": "close framing skips chain_bg reference"
              }
            ],
            "readiness_policy": "skipped_by_policy",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": true,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "c21cb4450de02ef9"
      },
      {
        "scene_index": 5,
        "heading": "S5. INT. 산악 접촉소 — 황혼 (03:10–04:15)",
        "beat_title": "천장의 충격과 낙하하는 돌가루",
        "representative_moment": "천장의 틈새에서 굵은 돌가루가 쏟아져 내리는 순간.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "steep upward angle looking at the ceiling, low-key ambient illumination focusing on the falling dust",
            "t2i_prompt": "Photorealistic cinematic still. Seen from a steep upward angle looking at the [L11: ceiling of the underground interior], the underside of the ceiling faces the camera, revealing a fresh cracked gap occupying the upper third of the frame. The heaviest stream of falling stone dust descends from the crack, extending into the lower frame and preserving the vertical force of the structural impact. The cool, muted, low-key ambient lighting casts deep shadows across the ceiling surface, providing enough separation for the falling dust to register clearly without introducing a new visible light source, keeping the space confined and shadowed.",
            "outfit_assignments": [],
            "source_facts": [
              "밖에서 둔탁한 충격.",
              "천장에서 돌가루가 떨어진다.",
              "Cool, low-key interior illumination"
            ],
            "visual_inferences": [
              "The ceiling crack marks the point of structural failure.",
              "The falling dust represents the immediate physical danger following the external impact."
            ],
            "creative_decisions": [
              "Used a steep upward angle to emphasize the vulnerability of the underground space.",
              "Focused entirely on the ceiling crack and the falling dust as the primary subjects, omitting any characters to isolate the structural threat."
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [],
            "reference_phrase_kinds": [],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "6d5777f205f18d64",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "7b8bf491022a522e",
              "owned_usage_hash": "8de771497838ce3d",
              "validator": "scene_detail_owned_objects.v1.close_skip",
              "violations": []
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 5,
        "dependency_reason": "Shot 7의 조명 및 배경 분위기(low-key ambient, cool muted tones)를 연속성 차원에서 유지",
        "_shot_index": 11,
        "visible_entities": [
          "L11"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 5,
            "shot_index": 11
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "close",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "Complete the sharp tilt from Tony's position into a steep upward angle, locking on the cracked ceiling gap as the heaviest stream of stone dust breaks downward through the center. The gap sits across the upper third while the falling particles extend into the lower frame, preserving the vertical force of the impact.",
            "lighting_mood": "Low-key ambient interior illumination gives the falling stone dust enough separation to register against the ceiling without introducing a new visible source.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 0명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "primary_subject_only_fully_visible": "the named primary subject (hand / face / object) is the only fully visible entity in the frame",
                  "other_entities_appearance_options": [
                    "partial visibility — shoulder slice / arm slice / sleeve / fingertips entering the frame edge",
                    "soft background — out-of-focus silhouette via shallow depth of field",
                    "absent — not rendered at all even if listed in visible_entities"
                  ],
                  "forbidden_combinations": [
                    "same character's full body and body-part close-up in the same frame (view mixing)",
                    "two characters both rendered as sharp close-up faces in the same frame"
                  ],
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "wide_medium_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=close — wide/medium 규칙은 이 샷에 해당하지 않는다"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": null,
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "L11"
            ],
            "allowed_outlook_pairs": [],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": false,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "skipped_close_framing",
            "bg_id": "L11B02",
            "reference_usage": "skipped_close_framing",
            "owned_objects": [],
            "camera_reference": null,
            "close_framing_skips_background_ref": true,
            "constraints": [
              "this shot uses close framing — the background reference is intentionally omitted at composition time",
              "describe the close subject and its immediate surroundings without relying on the absent background reference",
              "use [L##: ...] block for frame-edge surface description"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "outpost_low_key_lighting",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The interior environment is characterized by cool, muted, and low-key ambient lighting. The space is heavily shadowed with restrained contrast, lacking any overwhelming main light source and relying on dim practical illumination.",
                "element_scope": "full",
                "applies_to_shots": [
                  7,
                  8,
                  11
                ],
                "source_facts": [
                  "Cool, low-key interior illumination",
                  "cool, muted, low-key interior",
                  "Low-key ambient interior illumination"
                ],
                "visual_inferences": [
                  "The space feels confined and shadowy, typical of an underground or sealed bunker setting.",
                  "Overall color temperature leans toward cool, clinical blues or grays."
                ],
                "creative_decisions": [
                  "The dim ambient lighting enhances the sense of isolation and emphasizes localized light sources like the monitor screen."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 5,
                "shot_index": 7,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [],
            "forbidden_refs": [
              {
                "kind": "background",
                "reason": "close framing skips chain_bg reference"
              }
            ],
            "readiness_policy": "skipped_by_policy",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": true,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "dc33c3c7da09bcfe"
      },
      {
        "scene_index": 6,
        "heading": "S6. INT. 산악 접촉소 — 황혼 (04:15–05:10)",
        "beat_title": "적의 물리적 침투 개시",
        "representative_moment": "A brilliant circular line is suddenly drawn on the center of the closed rock door as the enemy breach begins.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "Low knee-height observer POV with severe local highlight contrast",
            "t2i_prompt": "Photorealistic cinematic still. The narrow tunnel entrance edge occupies the middle-left foreground, while the closed rock door and circular line stand in the middle-center midground. Seen from knee height beside the opening, the door's oblique front face is visible several paces away inside the L11B02 space. A brilliant glowing breach ring supplies a severe local highlight at the center of the door, cutting through the otherwise cool, muted, and low-key illumination of the rocky underground facility. Deep, heavy shadows dominate the unlit corners of the utilitarian environment. The rough, unpolished natural rock boundaries of the escape passage press closely around the left edge, defining the claustrophobic structure.",
            "outfit_assignments": [],
            "source_facts": [
              "Second impact. A glowing circle appears on the rock door.",
              "Narrow tunnel entrance open and low to the ground.",
              "cool, muted, low-key illumination"
            ],
            "visual_inferences": [
              "contact_station_illumination: Deep, heavy shadows dominating the unlit corners of the rocky interior space",
              "escape_tunnel_structure: Rough, unpolished natural rock textures forming the boundaries of the passage"
            ],
            "creative_decisions": [
              "Low knee-height framing to emphasize the confined physical space of the crawlway against the looming threat of the breached door",
              "High contrast lighting driven by the glowing breach ring, serving as the sole bright focus in the otherwise cold, dim room"
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [
              "fsc_001",
              "fsc_002"
            ],
            "reference_phrase_kinds": [
              "background"
            ],
            "owned_object_usage": [
              {
                "owned_token": "breach ring",
                "usage_kind": "redraw",
                "source_phrase": "A brilliant glowing breach ring"
              },
              {
                "owned_token": "cabinet",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "control panel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "debris",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "door",
                "usage_kind": "redraw",
                "source_phrase": "the closed rock door and circular line"
              },
              {
                "owned_token": "dust",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "isolation unit",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "light fixture",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "long gun",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "scanner",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "smartphone",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "cb4bd3e7df57c4df",
              "owned_hash": "e004a82b71728f88",
              "camera_direction_hash": "4e4c8899965c8663",
              "owned_usage_hash": "965028a24cbe3bb2",
              "validator": "scene_detail_owned_objects.v1",
              "violations": [
                {
                  "owned_object": "breach ring",
                  "violating_phrase": "A brilliant glowing breach ring supplies a severe local highlight at the center of the door",
                  "reason": "The glowing breach ring on the door is explicitly and structurally required by the shot_intent ('A brilliant circular line is suddenly drawn on the center of the closed rock door as the enemy breach begins'), satisfying narrow exception (1) for visual state change.",
                  "verdict": "allowed_visual_state_change"
                },
                {
                  "owned_object": "door",
                  "violating_phrase": "the closed rock door and circular line stand in the middle-center midground",
                  "reason": "The door is referenced purely as an existing spatial anchor and positional reference in the scene rather than being newly created, contrary to the producer's redraw declaration.",
                  "verdict": "anchor_reference"
                }
              ]
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 5,
        "dependency_reason": "Exact background reference from previous shot S5_Shot11 (location L11)",
        "_shot_index": 11,
        "visible_entities": [
          "L11"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 6,
            "shot_index": 11
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "medium",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "From knee height beside the tunnel entrance, hold a low, static three-quarter view of the closed rock door several paces away. The door remains centered in the midground with its oblique face readable, and the brilliant circular line sits at its center while the open tunnel edge lightly brackets one side of the frame.",
            "lighting_mood": "The brilliant circular line supplies a severe local highlight against the scene's otherwise cool, muted, low-key illumination.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 0명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=medium — close 규칙은 이 샷에 해당하지 않는다"
                },
                "wide_medium_rules": {
                  "primary_and_secondary_both_visible": "primary subject + secondary entities all renderable",
                  "body_part_close_up_forbidden": "body-part close-up of any character is forbidden in wide/medium framing — close-up belongs in a separate shot",
                  "fg_bg_separation_requires_shared_anchor": "if two characters are rendered fg/bg separated, also apply fg_bg_shared_anchor_rule",
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": {
              "reason": "required_background_position",
              "constraints": [
                {
                  "constraint_id": "fsc_001",
                  "target_kind": "background",
                  "target_id": "",
                  "label": "closed rock door and circular line",
                  "screen_zone": "middle_center",
                  "depth_plane": "midground",
                  "gesture_action": "none",
                  "gesture_target_label": ""
                },
                {
                  "constraint_id": "fsc_002",
                  "target_kind": "background",
                  "target_id": "",
                  "label": "narrow tunnel entrance edge",
                  "screen_zone": "middle_left",
                  "depth_plane": "foreground",
                  "gesture_action": "none",
                  "gesture_target_label": ""
                }
              ]
            },
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "L11"
            ],
            "allowed_outlook_pairs": [],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": false,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "background_ref_attached",
            "bg_id": "L11B02",
            "reference_usage": "exact_background",
            "owned_objects": [
              "breach ring",
              "cabinet",
              "control panel",
              "debris",
              "door",
              "dust",
              "hatch",
              "isolation unit",
              "lever",
              "light fixture",
              "long gun",
              "map",
              "monitor",
              "projection disc",
              "scanner",
              "smartphone",
              "ventilation duct",
              "workstation"
            ],
            "camera_reference": null,
            "close_framing_skips_background_ref": false,
            "constraints": [
              "owned_objects lists all environment objects already drawn in the background reference — do not create new ones",
              "if focusing on an owned object, declare its reference-kind via the reference_phrase_kinds sidecar (producer contract)"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "contact_station_illumination",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The interior environment features a consistent cool, muted, and low-key illumination. Deep shadows dominate the space, establishing a cold, secretive, and confined atmosphere throughout the rocky underground facility.",
                "element_scope": "full",
                "applies_to_shots": [
                  11,
                  13,
                  17
                ],
                "source_facts": [
                  "cool, muted, low-key illumination in Shot 11",
                  "Restrained, low-key illumination appropriate to the tunnel in Shot 13",
                  "cool, muted, low-key contact-station interior in Shot 17"
                ],
                "visual_inferences": [
                  "Deep, heavy shadows dominating the unlit corners of the rocky interior space",
                  "Utilitarian and unpolished underground facility aesthetic"
                ],
                "creative_decisions": [
                  "Maintaining a strictly cold color palette to emphasize the bleak and urgent escape situation"
                ],
                "confidence": "high"
              },
              {
                "element_id": "escape_tunnel_structure",
                "element_type": "environment_state",
                "character_name": "",
                "description": "A narrow, confined escape passage surrounded by rough rock boundaries. The tunnel opening is small and low to the ground, physically requiring a person to crawl, with unpolished natural rock walls pressing closely around the space.",
                "element_scope": "full",
                "applies_to_shots": [
                  11,
                  13
                ],
                "source_facts": [
                  "open tunnel edge lightly brackets one side of the frame",
                  "narrow tunnel requiring crawling on hands and knees",
                  "tunnel boundaries press close around the subject",
                  "confined passage"
                ],
                "visual_inferences": [
                  "Rough, unpolished natural rock textures forming the boundaries of the passage",
                  "The tunnel floor is uneven and rugged, typical of a raw mountain cave"
                ],
                "creative_decisions": [
                  "Emphasizing the claustrophobic tightness of the tunnel to heighten the physical tension of the escape sequence"
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 5,
                "shot_index": 11,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "background",
                "id": "L11B02",
                "policy": "required"
              }
            ],
            "forbidden_refs": [],
            "readiness_policy": "block_if_missing",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": false,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "fdcdcd5aff02c1fc"
      },
      {
        "scene_index": 6,
        "heading": "INT. 산악 접촉소 — 황혼",
        "beat_title": "좁은 환기축 터널 진입",
        "representative_moment": "C02가 어둡고 좁은 터널 안으로 몸을 굽혀 기어 들어가는 뒷모습.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "knee-height observer tracking shot from behind, medium framing with claustrophobic tension",
            "t2i_prompt": "Photorealistic cinematic still. C02O02 in a green aerial reconnaissance combat uniform, a Caucasian woman in her late 20s, with back to camera, crawls on hands and knees in the middle-center midground, moving toward the deeper tunnel passage in the upper-center background. Seen from behind at knee height, the figure's bent back and alternating hands anchor the composition, her face completely hidden from view as she gazes off-screen into the path ahead. The narrow, confined crawl tunnel of the L11B01 space, surrounded by rough rock boundaries, presses closely around the subject. The unpolished natural rock walls recede directly away from the camera, bathed in restrained, low-key illumination with deep, heavy shadows dominating the cool, muted atmosphere.",
            "outfit_assignments": [
              {
                "character_id": "C02",
                "outlook_id": "O02"
              }
            ],
            "source_facts": [
              "narrow tunnel requiring crawling on hands and knees",
              "tunnel boundaries press close around the subject",
              "Restrained, low-key illumination appropriate to the tunnel maintains a cool, muted image",
              "Wilma crawling away from the lower center toward the confined passage"
            ],
            "visual_inferences": [
              "Rough, unpolished natural rock textures forming the boundaries of the passage",
              "View from behind at knee height capturing the tight squeeze of the tunnel",
              "Deep shadows emphasizing the secretive and confined atmosphere"
            ],
            "creative_decisions": [
              "Emphasizing the claustrophobic tightness of the tunnel to heighten the physical tension of the escape sequence",
              "Maintaining a strictly cold, low-key color palette without specifying an additional light source to align with the prior shot's mood"
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [
              "fsc_001",
              "fsc_002"
            ],
            "reference_phrase_kinds": [
              "character",
              "background"
            ],
            "owned_object_usage": [
              {
                "owned_token": "crawl tunnel",
                "usage_kind": "redraw",
                "source_phrase": "The narrow, confined crawl tunnel of the L11B01 space, surrounded by rough rock boundaries, presses closely around the subject."
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map display",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "operations room",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "5dc3c73bb24ac866",
              "owned_hash": "8a424665b182caf5",
              "camera_direction_hash": "3aa105eb22076981",
              "owned_usage_hash": "625a9761f00a4a4c",
              "validator": "scene_detail_owned_objects.v1",
              "violations": [
                {
                  "owned_object": "crawl tunnel",
                  "violating_phrase": "The narrow, confined crawl tunnel of the L11B01 space, surrounded by rough rock boundaries, presses closely around the subject.",
                  "reason": "The crawl tunnel is referenced purely as existing environmental space and composition framing around the subject rather than a newly generated or modified asset, making it an anchor reference despite the producer's redraw declaration.",
                  "verdict": "anchor_reference"
                }
              ]
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 6,
        "dependency_reason": "Shot 11과 Shot 13 간의 산악 접촉소 조명 상태(contact_station_illumination) 및 터널 형태(escape_tunnel_structure) 연속성 유지",
        "_shot_index": 13,
        "visible_entities": [
          "C02",
          "L11"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 6,
            "shot_index": 13
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "medium",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "Track at knee height from one body length behind Wilma, aligned with the tunnel as she crawls away from the lower center toward the confined passage ahead. Her bent back and alternating hands define the midground while the tunnel boundaries press close around her, keeping her destination directly along her off-screen gaze.",
            "lighting_mood": "Restrained, low-key illumination appropriate to the tunnel maintains a cool, muted image without specifying an additional source.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 1명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=medium — close 규칙은 이 샷에 해당하지 않는다"
                },
                "wide_medium_rules": {
                  "primary_and_secondary_both_visible": "primary subject + secondary entities all renderable",
                  "body_part_close_up_forbidden": "body-part close-up of any character is forbidden in wide/medium framing — close-up belongs in a separate shot",
                  "fg_bg_separation_requires_shared_anchor": "if two characters are rendered fg/bg separated, also apply fg_bg_shared_anchor_rule",
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": {
              "reason": "movement_direction",
              "constraints": [
                {
                  "constraint_id": "fsc_001",
                  "target_kind": "background",
                  "target_id": "",
                  "label": "deeper tunnel passage",
                  "screen_zone": "upper_center",
                  "depth_plane": "background",
                  "gesture_action": "none",
                  "gesture_target_label": ""
                },
                {
                  "constraint_id": "fsc_002",
                  "target_kind": "character",
                  "target_id": "C02",
                  "label": "윌마 디어링",
                  "screen_zone": "middle_center",
                  "depth_plane": "midground",
                  "gesture_action": "moves_toward",
                  "gesture_target_label": "deeper tunnel passage"
                }
              ]
            },
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "C02",
              "L11"
            ],
            "allowed_outlook_pairs": [
              {
                "character_id": "C02",
                "outlook_id": "O02"
              }
            ],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": false,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "background_ref_attached",
            "bg_id": "L11B01",
            "reference_usage": "exact_background",
            "owned_objects": [
              "crawl tunnel",
              "hatch",
              "lever",
              "map display",
              "monitor",
              "operations room",
              "projection disc",
              "ventilation duct",
              "workstation"
            ],
            "camera_reference": null,
            "close_framing_skips_background_ref": false,
            "constraints": [
              "owned_objects lists all environment objects already drawn in the background reference — do not create new ones",
              "if focusing on an owned object, declare its reference-kind via the reference_phrase_kinds sidecar (producer contract)"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "contact_station_illumination",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The interior environment features a consistent cool, muted, and low-key illumination. Deep shadows dominate the space, establishing a cold, secretive, and confined atmosphere throughout the rocky underground facility.",
                "element_scope": "full",
                "applies_to_shots": [
                  11,
                  13,
                  17
                ],
                "source_facts": [
                  "cool, muted, low-key illumination in Shot 11",
                  "Restrained, low-key illumination appropriate to the tunnel in Shot 13",
                  "cool, muted, low-key contact-station interior in Shot 17"
                ],
                "visual_inferences": [
                  "Deep, heavy shadows dominating the unlit corners of the rocky interior space",
                  "Utilitarian and unpolished underground facility aesthetic"
                ],
                "creative_decisions": [
                  "Maintaining a strictly cold color palette to emphasize the bleak and urgent escape situation"
                ],
                "confidence": "high"
              },
              {
                "element_id": "escape_tunnel_structure",
                "element_type": "environment_state",
                "character_name": "",
                "description": "A narrow, confined escape passage surrounded by rough rock boundaries. The tunnel opening is small and low to the ground, physically requiring a person to crawl, with unpolished natural rock walls pressing closely around the space.",
                "element_scope": "full",
                "applies_to_shots": [
                  11,
                  13
                ],
                "source_facts": [
                  "open tunnel edge lightly brackets one side of the frame",
                  "narrow tunnel requiring crawling on hands and knees",
                  "tunnel boundaries press close around the subject",
                  "confined passage"
                ],
                "visual_inferences": [
                  "Rough, unpolished natural rock textures forming the boundaries of the passage",
                  "The tunnel floor is uneven and rugged, typical of a raw mountain cave"
                ],
                "creative_decisions": [
                  "Emphasizing the claustrophobic tightness of the tunnel to heighten the physical tension of the escape sequence"
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 6,
                "shot_index": 11,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "character_outlook",
                "id": "C02O02",
                "policy": "required"
              },
              {
                "kind": "background",
                "id": "L11B01",
                "policy": "required"
              }
            ],
            "forbidden_refs": [],
            "readiness_policy": "block_if_missing",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": false,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "079bab3d7e2ea8ec"
      },
      {
        "scene_index": 6,
        "heading": "S6. INT. 산악 접촉소 — 황혼 (04:15–05:10)",
        "beat_title": "기록 삭제 완료와 탈출",
        "representative_moment": "모니터의 백색 빛을 받으며 터널 안쪽을 향해 고개를 돌린 C07의 얼굴",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "low-angle close composition emphasizing stark white monitor light against deep shadows",
            "t2i_prompt": "Photorealistic cinematic still. Seen from slightly below chest height, C07O04 in a gray combat uniform, a Caucasian adult man, occupies the center of the frame from the chest up. He is positioned at a three-quarter angle from the right, twisting his shoulders away from a pulled mechanical lever visible near the lower frame edge, looking off-screen. Stark white light from erased monitor screens behind his outer shoulder shapes his face and jawline, keeping the monitors outside his direct eyeline. The background is a soft-focus rocky interior under cool, muted, low-key illumination, with deep heavy shadows dominating the unlit corners.",
            "outfit_assignments": [
              {
                "character_id": "C07",
                "outlook_id": "O04"
              }
            ],
            "source_facts": [
              "앨런이 마지막으로 접촉소의 기록 삭제 레버를 당긴다. 모든 화면이 백색으로 지워진다.",
              "모니터의 백색 빛을 받으며 터널 안쪽을 향해 고개를 돌린 앨런의 얼굴",
              "cool, muted, low-key contact-station interior"
            ],
            "visual_inferences": [
              "Deep, heavy shadows dominating the unlit corners of the rocky interior space",
              "The pulled lever remains visible at the bottom edge as a physical anchor for his action"
            ],
            "creative_decisions": [
              "Translated the dolly-in DP instruction into a close framing from slightly below chest height to firmly anchor the subject without wide-shot elements",
              "Kept the monitors soft-focused behind his shoulder to prioritize his facial expression under the stark white light"
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [],
            "reference_phrase_kinds": [
              "character"
            ],
            "owned_object_usage": [
              {
                "owned_token": "crawl tunnel",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "hatch",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "lever",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "map display",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "monitor",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "operations room",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "projection disc",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "ventilation duct",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "workstation",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "0919fa7eb29c2ecc",
              "owned_hash": "8a424665b182caf5",
              "camera_direction_hash": "9b89639cf31fb3ac",
              "owned_usage_hash": "9a2be5efede07cb2",
              "validator": "scene_detail_owned_objects.v1.close_skip",
              "violations": []
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 5,
        "dependency_reason": "S5_Shot7로부터 이어지는 contact_station_illumination(cool, muted, low-key illumination) 환경 상태 유지",
        "_shot_index": 17,
        "visible_entities": [
          "C07"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 6,
            "shot_index": 17
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "close",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "Complete the dolly-in beside the console from slightly below Alan's chest height, holding his three-quarter face near center as he turns away from the pulled deletion lever toward the off-frame tunnel. The erased white monitors remain behind his outer shoulder rather than on his eyeline, separating what he has sacrificed from the route he now intends to follow.",
            "lighting_mood": "White monitor light shapes Alan's face within the cool, muted, low-key contact-station interior.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 1명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "primary_subject_only_fully_visible": "the named primary subject (hand / face / object) is the only fully visible entity in the frame",
                  "other_entities_appearance_options": [
                    "partial visibility — shoulder slice / arm slice / sleeve / fingertips entering the frame edge",
                    "soft background — out-of-focus silhouette via shallow depth of field",
                    "absent — not rendered at all even if listed in visible_entities"
                  ],
                  "forbidden_combinations": [
                    "same character's full body and body-part close-up in the same frame (view mixing)",
                    "two characters both rendered as sharp close-up faces in the same frame"
                  ],
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "wide_medium_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=close — wide/medium 규칙은 이 샷에 해당하지 않는다"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": null,
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "C07"
            ],
            "allowed_outlook_pairs": [
              {
                "character_id": "C07",
                "outlook_id": "O04"
              }
            ],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": true,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "skipped_close_framing",
            "bg_id": "L11B01",
            "reference_usage": "skipped_close_framing",
            "owned_objects": [],
            "camera_reference": null,
            "close_framing_skips_background_ref": true,
            "constraints": [
              "this shot uses close framing — the background reference is intentionally omitted at composition time",
              "describe the close subject and its immediate surroundings without relying on the absent background reference",
              "use [L##: ...] block for frame-edge surface description"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "contact_station_illumination",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The interior environment features a consistent cool, muted, and low-key illumination. Deep shadows dominate the space, establishing a cold, secretive, and confined atmosphere throughout the rocky underground facility.",
                "element_scope": "full",
                "applies_to_shots": [
                  11,
                  13,
                  17
                ],
                "source_facts": [
                  "cool, muted, low-key illumination in Shot 11",
                  "Restrained, low-key illumination appropriate to the tunnel in Shot 13",
                  "cool, muted, low-key contact-station interior in Shot 17"
                ],
                "visual_inferences": [
                  "Deep, heavy shadows dominating the unlit corners of the rocky interior space",
                  "Utilitarian and unpolished underground facility aesthetic"
                ],
                "creative_decisions": [
                  "Maintaining a strictly cold color palette to emphasize the bleak and urgent escape situation"
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 5,
                "shot_index": 7,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "character_outlook",
                "id": "C07O04",
                "policy": "required"
              }
            ],
            "forbidden_refs": [
              {
                "kind": "background",
                "reason": "close framing skips chain_bg reference"
              }
            ],
            "readiness_policy": "skipped_by_policy",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": true,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "d500d2ee02309018"
      },
      {
        "scene_index": 7,
        "heading": "EXT. 산비탈 아래 비밀 출구 — 밤",
        "beat_title": "적의 추격을 따돌린 능선",
        "representative_moment": "어두운 산 정상의 능선에 수많은 붉은 불빛들이 밀집해 있는 원경.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "Elevated distant uphill view with cool, muted night darkness and stark red contrast",
            "t2i_prompt": "Photorealistic cinematic still. Seen from an elevated distant vantage, an empty hillside recedes upward toward the crest. In the upper-center background, the mountain ridge with converging red lights extends laterally across the frame, set against the existing pursuit lights in the distance. Focus is fixed along the glowing ridge, emphasizing the vast unoccupied distance below. The L12B01 environment is enveloped in deep night darkness with a cool, muted lighting tone, keeping the lower slope shrouded in heavy shadow while the stark, restrained red illumination provides the primary focal pattern against the dark sky.",
            "outfit_assignments": [],
            "source_facts": [
              "Shot description: 어두운 산 정상의 능선을 따라 수많은 붉은 불빛들이 밀집한 원경",
              "S07 action: 붉은 점들은 빈 산 정상으로 몰려간다."
            ],
            "visual_inferences": [
              "fixed_element ambient_night_lighting: The overall environment is enveloped in deep night darkness with a cool, muted lighting tone.",
              "Elevated distant vantage emphasizing the empty separation between the escapees' true location and the lured enemy forces."
            ],
            "creative_decisions": [
              "Wide uphill composition to establish spatial scale and isolation.",
              "Restrained contrast highlighting the warning-red lights against the deep, cool shadows of the empty hillside."
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [
              "fsc_001"
            ],
            "reference_phrase_kinds": [
              "background"
            ],
            "owned_object_usage": [
              {
                "owned_token": "pursuit light",
                "usage_kind": "anchor",
                "source_phrase": "set against the existing pursuit lights in the distance"
              },
              {
                "owned_token": "rock outcrop",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "rockfall",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "signal lamp",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "transmitter",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "tunnel entrance",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "6e673f0c6e101fed",
              "owned_hash": "aaee8f021b8749a3",
              "camera_direction_hash": "2d7f3da512cb42c6",
              "owned_usage_hash": "47f6d6cd4786368a",
              "validator": "scene_detail_owned_objects.v1",
              "violations": [
                {
                  "owned_object": "pursuit light",
                  "violating_phrase": "set against the existing pursuit lights in the distance",
                  "reason": "The prompt references the existing pursuit lights as a background spatial anchor without instructing any new generation or modification, consistent with the producer's anchor declaration.",
                  "verdict": "anchor_reference"
                }
              ]
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": -1,
        "dependency_reason": "",
        "_shot_index": 3,
        "visible_entities": [
          "L12"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 7,
            "shot_index": 3
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "wide",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "From the elevated distant vantage, finish the pan and lock into a long uphill view across the empty separation toward the mountain ridge. The ridge runs through the upper portion of the frame, where numerous red lights cluster densely along its contour while the unoccupied distance below preserves their deceptive concentration.",
            "lighting_mood": "Night darkness and the concentrated warning-red lights create a cool, muted long view with stark but restrained points of contrast.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 0명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=wide — close 규칙은 이 샷에 해당하지 않는다"
                },
                "wide_medium_rules": {
                  "primary_and_secondary_both_visible": "primary subject + secondary entities all renderable",
                  "body_part_close_up_forbidden": "body-part close-up of any character is forbidden in wide/medium framing — close-up belongs in a separate shot",
                  "fg_bg_separation_requires_shared_anchor": "if two characters are rendered fg/bg separated, also apply fg_bg_shared_anchor_rule",
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": {
              "reason": "required_background_position",
              "constraints": [
                {
                  "constraint_id": "fsc_001",
                  "target_kind": "background",
                  "target_id": "",
                  "label": "ridge with converging red lights",
                  "screen_zone": "upper_center",
                  "depth_plane": "background",
                  "gesture_action": "none",
                  "gesture_target_label": ""
                }
              ]
            },
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "L12"
            ],
            "allowed_outlook_pairs": [],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": false,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "background_ref_attached",
            "bg_id": "L12B01",
            "reference_usage": "exact_background",
            "owned_objects": [
              "pursuit light",
              "rock outcrop",
              "rockfall",
              "signal lamp",
              "transmitter",
              "tunnel entrance"
            ],
            "camera_reference": null,
            "close_framing_skips_background_ref": false,
            "constraints": [
              "owned_objects lists all environment objects already drawn in the background reference — do not create new ones",
              "if focusing on an owned object, declare its reference-kind via the reference_phrase_kinds sidecar (producer contract)"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "ambient_night_lighting",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The overall environment is enveloped in deep night darkness with a cool, muted lighting tone. This low-key illumination provides stark but restrained contrast across the scene, deliberately keeping the background shrouded in shadow while emphasizing specific points of interest or isolated light sources.",
                "element_scope": "full",
                "applies_to_shots": [
                  3,
                  10,
                  13
                ],
                "source_facts": [
                  "Night darkness and the concentrated warning-red lights create a cool, muted long view",
                  "Night-appropriate low-key illumination renders the engraved relief with cool, muted contrast",
                  "The dark night forest remains cool and muted"
                ],
                "visual_inferences": [
                  "The absence of a strong moon or general ambient light unifies the vast outdoor landscapes and the extreme macro close-ups under the same heavy, oppressive shadow."
                ],
                "creative_decisions": [
                  "Applying a consistent cool color palette and restrained lighting model across both wide shots and macro shots to maintain a tense, unified nocturnal atmosphere."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "background",
                "id": "L12B01",
                "policy": "required"
              }
            ],
            "forbidden_refs": [],
            "readiness_policy": "block_if_missing",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": false,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "add0b514e33f287f"
      },
      {
        "scene_index": 7,
        "heading": "EXT. 산비탈 아래 비밀 출구 — 밤",
        "beat_title": "내부 첩자의 존재 확인",
        "representative_moment": "초근접 앵글 속에서 P22 송신기 표면에 새겨진 세 갈래 나무 문장이 차가운 조명 아래 뚜렷하게 드러난다.",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "extreme macro insert shot with steeply oblique angle and cool, muted low-key lighting",
            "t2i_prompt": "Photorealistic cinematic still. Seen from a steeply oblique angle on an extreme macro view, the tiny P22 metal transmitter fills the frame. The emblem-bearing face is angled steeply across the camera, with the surrounding transmitter surface retained at the edges so the mark reads as a physical engraving. The three-branched tree emblem occupies the central portion of the image, allowing the three branches and engraved depth to remain legible. The overall environment is enveloped in deep night darkness, keeping the out-of-focus background completely shrouded in shadow. Night-appropriate low-key illumination renders the engraved relief with cool, muted contrast, emphasizing the isolated physical detail.",
            "outfit_assignments": [],
            "source_facts": [
              "S7 action: 윌마가 송신기를 손바닥에 올린다. 송신기 표면에 Wyoming 공동체의 세 갈래 나무 문장이 새겨져 있다.",
              "S7 action: 안에서 쌀알만 한 송신기가 나온다.",
              "shot description: 송신기 표면에 음각으로 새겨진 세 갈래 나무 문장의 초근접 클로즈업"
            ],
            "visual_inferences": [
              "fixed_element three_branched_tree_emblem: The precise structural design and geometric proportions of this emblem remain absolutely identical, showing clear engraved depth.",
              "fixed_element ambient_night_lighting: The overall environment is enveloped in deep night darkness with a cool, muted lighting tone."
            ],
            "creative_decisions": [
              "Utilized an extreme macro framing ('fills the frame') to capture only the P22 transmitter, completely excluding any human body parts in accordance with the camera direction.",
              "Applied low-key illumination against a deeply shadowed background to highlight the recessed depth of the three-branched tree emblem."
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [],
            "reference_phrase_kinds": [
              "prop",
              "background"
            ],
            "owned_object_usage": [
              {
                "owned_token": "pursuit light",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "rock outcrop",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "rockfall",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "signal lamp",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "transmitter",
                "usage_kind": "redraw",
                "source_phrase": "the tiny P22 metal transmitter fills the frame"
              },
              {
                "owned_token": "tunnel entrance",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "c150798f4bbb38b7",
              "owned_hash": "aaee8f021b8749a3",
              "camera_direction_hash": "dcd6343f85edb5a3",
              "owned_usage_hash": "38f66ca8d2758d32",
              "validator": "scene_detail_owned_objects.v1",
              "violations": [
                {
                  "owned_object": "transmitter",
                  "violating_phrase": "the tiny P22 metal transmitter fills the frame",
                  "reason": "producer는 redraw로 선언했으나, shot_intent에서 'P22 송신기 표면에 새겨진 세 갈래 나무 문장이 차가운 조명 아래 뚜렷하게 드러난다'라고 명시적으로 요구하여 송신기 표면의 각인 묘사는 narrow exception (1)에 부합하므로 allowed_visual_state_change에 해당합니다.",
                  "verdict": "allowed_visual_state_change"
                }
              ]
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 7,
        "dependency_reason": "동일 씬 내 이전 샷(Shot 3)에서 설정된 깊은 밤의 어두운 환경과 조명 분위기를 그대로 유지하여 연속성 확보",
        "_shot_index": 10,
        "visible_entities": [
          "P22"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 7,
            "shot_index": 10
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "insert",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "End the steeply oblique dolly-in just above Wilma's palm on an extreme macro view of the rice-grain-sized transmitter's engraved surface, excluding the rest of her hand as specified. The three-branched tree emblem occupies the central portion of the image, with surrounding transmitter surface retained at the edges so the mark reads as an engraving rather than a floating graphic.",
            "lighting_mood": "Night-appropriate low-key illumination renders the engraved relief with cool, muted contrast and no newly specified source.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 0명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "primary_subject_only_fully_visible": "the named primary subject (hand / face / object) is the only fully visible entity in the frame",
                  "other_entities_appearance_options": [
                    "partial visibility — shoulder slice / arm slice / sleeve / fingertips entering the frame edge",
                    "soft background — out-of-focus silhouette via shallow depth of field",
                    "absent — not rendered at all even if listed in visible_entities"
                  ],
                  "forbidden_combinations": [
                    "same character's full body and body-part close-up in the same frame (view mixing)",
                    "two characters both rendered as sharp close-up faces in the same frame"
                  ],
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "wide_medium_rules": {
                  "primary_and_secondary_both_visible": "primary subject + secondary entities all renderable",
                  "body_part_close_up_forbidden": "body-part close-up of any character is forbidden in wide/medium framing — close-up belongs in a separate shot",
                  "fg_bg_separation_requires_shared_anchor": "if two characters are rendered fg/bg separated, also apply fg_bg_shared_anchor_rule",
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": null,
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "P22"
            ],
            "allowed_outlook_pairs": [],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": true,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "background_ref_attached",
            "bg_id": "L12B01",
            "reference_usage": "exact_background",
            "owned_objects": [
              "pursuit light",
              "rock outcrop",
              "rockfall",
              "signal lamp",
              "transmitter",
              "tunnel entrance"
            ],
            "camera_reference": null,
            "close_framing_skips_background_ref": false,
            "constraints": [
              "owned_objects lists all environment objects already drawn in the background reference — do not create new ones",
              "if focusing on an owned object, declare its reference-kind via the reference_phrase_kinds sidecar (producer contract)"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "ambient_night_lighting",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The overall environment is enveloped in deep night darkness with a cool, muted lighting tone. This low-key illumination provides stark but restrained contrast across the scene, deliberately keeping the background shrouded in shadow while emphasizing specific points of interest or isolated light sources.",
                "element_scope": "full",
                "applies_to_shots": [
                  3,
                  10,
                  13
                ],
                "source_facts": [
                  "Night darkness and the concentrated warning-red lights create a cool, muted long view",
                  "Night-appropriate low-key illumination renders the engraved relief with cool, muted contrast",
                  "The dark night forest remains cool and muted"
                ],
                "visual_inferences": [
                  "The absence of a strong moon or general ambient light unifies the vast outdoor landscapes and the extreme macro close-ups under the same heavy, oppressive shadow."
                ],
                "creative_decisions": [
                  "Applying a consistent cool color palette and restrained lighting model across both wide shots and macro shots to maintain a tense, unified nocturnal atmosphere."
                ],
                "confidence": "high"
              },
              {
                "element_id": "three_branched_tree_emblem",
                "element_type": "environment_state",
                "character_name": "",
                "description": "A distinct symbol depicting a three-branched tree serves as a recurring visual mark. The precise structural design and geometric proportions of this emblem remain absolutely identical, whether it appears as an engraved physical relief on a tiny metal surface or as a brightly illuminated signal pattern in the distance.",
                "element_scope": "close",
                "applies_to_shots": [
                  10,
                  13
                ],
                "source_facts": [
                  "The three-branched tree emblem occupies the central portion of the image",
                  "its fully illuminated pattern forms the same three-branched tree emblem",
                  "송신기 표면에 Wyoming 공동체의 세 갈래 나무 문장이 새겨져 있다",
                  "같은 문장의 신호등을 한 번 켰다가 끈다"
                ],
                "visual_inferences": [
                  "The emblem is composed of sharp, well-defined lines that allow its shape to be instantly identifiable and perfectly matched regardless of the medium (metal engraving vs. light source)."
                ],
                "creative_decisions": [
                  "Ensuring the exact silhouette of the three branches perfectly matches between the macro physical prop and the distant light signal to visually confirm the narrative link of an internal traitor."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 7,
                "shot_index": 3,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "background",
                "id": "L12B01",
                "policy": "required"
              },
              {
                "kind": "prop",
                "id": "P22",
                "policy": "required"
              }
            ],
            "forbidden_refs": [],
            "readiness_policy": "block_if_missing",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [
            {
              "contract_id": "rc_001",
              "scope": {
                "scene_index": 7,
                "shot_indices": [
                  10
                ],
                "duration": "single_shot"
              },
              "targets": [
                {
                  "entity_id": "P22",
                  "role": "visual_target"
                }
              ],
              "requirements": [
                {
                  "dimension": "visual_identity",
                  "operation": "preserve",
                  "strength": "required",
                  "reference_policy": "use_entity_reference"
                }
              ]
            }
          ],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": false,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "bc55563e9f5fd903"
      },
      {
        "scene_index": 7,
        "heading": "EXT. 산비탈 아래 비밀 출구 — 밤",
        "beat_title": "적성 세력의 신호 확인",
        "representative_moment": "어두운 숲 너머로 세 갈래 나무 문양의 신호등 불빛이 선명하게 켜진 찰나",
        "t2i_variations": [
          {
            "variant_label": "var_1",
            "camera_effect": "Eye-level wide shot across a dark forest, focusing on a single distant point of contrast in cool, muted night lighting.",
            "t2i_prompt": "Photorealistic cinematic still. [L12: A vast outdoor mountain forest enveloped in deep night darkness, cool and muted, with heavy shadows shrouding the intervening trees.] Seen from observer eye level, the wide view passes between the dark tree trunks. In the middle-center background, a three-branched signal light is fully illuminated, directed precisely toward the camera. Focus is on the sharp, well-defined three-branched tree emblem formed by the signal lamp, providing a stark point of contrast against the low-key illumination of the silent woods.",
            "outfit_assignments": [],
            "source_facts": [
              "shot description: Distant three-branched signal light: Fully illuminated for a single instant before switching off.",
              "shot description: Forest between group and signal: Dark, with the signal visible beyond it.",
              "shot description: Its signal-bearing face is directed toward the group's position."
            ],
            "visual_inferences": [
              "fixed_element ambient_night_lighting: The overall environment is enveloped in deep night darkness with a cool, muted lighting tone.",
              "fixed_element three_branched_tree_emblem: The precise structural design and geometric proportions remain identical as a brightly illuminated signal pattern."
            ],
            "creative_decisions": [
              "Using silhouetted tree trunks in the foreground to establish distance and concealment, maintaining the exact_background consistency.",
              "Applying a stark, sharp focal point on the distant emblem to highlight the specific point of contrast against the low-key environment."
            ],
            "confidence": "high",
            "applied_frame_spatial_constraint_ids": [
              "fsc_001"
            ],
            "reference_phrase_kinds": [
              "background"
            ],
            "owned_object_usage": [
              {
                "owned_token": "pursuit light",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "rock outcrop",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "rockfall",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "signal lamp",
                "usage_kind": "redraw",
                "source_phrase": "Focus is on the sharp, well-defined three-branched tree emblem formed by the signal lamp"
              },
              {
                "owned_token": "transmitter",
                "usage_kind": "absent",
                "source_phrase": ""
              },
              {
                "owned_token": "tunnel entrance",
                "usage_kind": "absent",
                "source_phrase": ""
              }
            ],
            "owned_validation": {
              "schema_version": 2,
              "t2i_prompt_hash": "2a70fb4cc6bbc826",
              "owned_hash": "aaee8f021b8749a3",
              "camera_direction_hash": "c71ab405f523af66",
              "owned_usage_hash": "348e1efb6aac0c34",
              "validator": "scene_detail_owned_objects.v1",
              "violations": [
                {
                  "owned_object": "signal lamp",
                  "violating_phrase": "Focus is on the sharp, well-defined three-branched tree emblem formed by the signal lamp",
                  "reason": "The prompt describes the illuminated visual pattern of the signal lamp, which satisfies narrow exception (1) as the illuminated three-branched tree emblem state is explicitly and structurally required by the shot_intent ('세 갈래 나무 문양의 신호등 불빛이 선명하게 켜진 찰나').",
                  "verdict": "allowed_visual_state_change"
                }
              ]
            }
          }
        ],
        "scene_type": "normal",
        "dependent_scene_index": 7,
        "dependency_reason": "exact_background rule linked to S7_Shot3 to maintain spatial continuity of the mountain forest environment.",
        "_shot_index": 13,
        "visible_entities": [
          "L12"
        ],
        "render_prompt_card": {
          "schema_version": 1,
          "shot_key": {
            "scene_index": 7,
            "shot_index": 13
          },
          "render_strategy": {
            "mode": "direct",
            "primary_subject": "",
            "framing_scale": "wide",
            "moment_lock": "single still moment from the selected shot only",
            "camera_direction": "At eye height near the group's position, hold a static view across the forest with no people entering the frame. The distant signal occupies a compact central area beyond the intervening trees, captured at the exact instant its fully illuminated pattern forms the same three-branched tree emblem.",
            "lighting_mood": "The dark night forest remains cool and muted while the briefly illuminated signal provides the shot's precise point of contrast.",
            "perception_mode": "direct",
            "spatial_consistency": {
              "camera_frame_rule": {
                "applies_when": "camera_direction expresses a vertical camera height or viewpoint position and the scene description includes frame-edge surfaces",
                "consistency_check_summary": "elements described inside the frame must be at vertical/depth positions consistent with the camera's height — low cameras place feet/lower-body in the lower frame, overhead cameras place head/shoulders in upper frame, hip-height cameras place torso in frame with head/feet implied beyond edges",
                "core_principles": [
                  "low cameras at ground/quay/floor level place feet and lower legs in the lower frame; chest-height frame-edge surfaces behind the subject are geometrically forbidden",
                  "overhead high angles from above place the top of head and shoulders in the upper-center frame; floor surfaces visible below feet at the lower frame edge are forbidden",
                  "hip-height side cameras place torso and arms in frame; head and feet are implied beyond frame edges, not both fully visible together",
                  "full body shots require either eye-level or high-angle framing where head and feet both fit; low cameras + full silhouette in one frame is forbidden"
                ],
                "forbidden_combinations": [
                  "low camera at ground/quay/floor level + frame-edge surface described at chest/waist height behind the subject",
                  "overhead high angle + feet visible at the lower frame edge with floor visible below them",
                  "hip-height side camera + full silhouette including head and feet visible in same frame",
                  "low angle looking up + ceiling and floor both fully visible in same frame"
                ],
                "recommended_phrasings": [
                  "low observer at ground level, shoes and lower legs occupying the lower frame, upper body extending toward the upper edge",
                  "overhead high angle from above, the top of head and shoulders fill the upper-center frame, the floor visible around them",
                  "at hip height beside the subject, torso and arms in frame, head and lower legs implied beyond frame edges",
                  "low angle looking up at the figure, ceiling lamps in the upper frame, the wall behind extending downward"
                ],
                "self_check_steps": [
                  "identify the camera position phrasing (height + angle)",
                  "list elements described inside the frame (foreground prop, body parts, background surfaces)",
                  "verify each element's vertical/depth placement is geometrically consistent with that camera position",
                  "if mismatch found, adjust frame-edge description to match camera position"
                ],
                "rationale_summary": "T2I 모델은 카메라 위치와 frame 안 요소 vertical 좌표 모순 시 절충 합성 — 인물이 가구에 박히거나 신체 일부가 잘리거나 prop 이 떠 있는 부자연스러운 결과"
              },
              "fg_bg_shared_anchor_rule": {
                "not_applicable": true,
                "not_applicable_reason": "이 샷의 visible_entities 에 인물이 0명이라 두 인물 사이의 공유 anchor 규칙은 성립하지 않는다",
                "applies_when": "two characters share an interaction (dialogue / handing object / facing / co-watching) AND are described as foreground/background separated AND a shared surface or spatial anchor exists between them (bench / table / door / room)"
              },
              "primary_framing_rule": {
                "applies_when": "framing_scale ∈ {close, medium, wide, insert} from `shot_staging.framing_scale` enum (render_strategy.framing_scale, production SOT). LLM consumes the enum only; keyword classification is forbidden.",
                "close_framing_rules": {
                  "not_applicable": true,
                  "not_applicable_reason": "framing_scale=wide — close 규칙은 이 샷에 해당하지 않는다"
                },
                "wide_medium_rules": {
                  "primary_and_secondary_both_visible": "primary subject + secondary entities all renderable",
                  "body_part_close_up_forbidden": "body-part close-up of any character is forbidden in wide/medium framing — close-up belongs in a separate shot",
                  "fg_bg_separation_requires_shared_anchor": "if two characters are rendered fg/bg separated, also apply fg_bg_shared_anchor_rule",
                  "subject_reference_policy_cross_ref": "see id_policy.subject_reference_policy (subject_reference_policy SOT — policy enum 에 따라 ID/outlook 사용)"
                },
                "self_check_steps": [
                  "read framing_scale enum from render_strategy.framing_scale (do NOT infer from camera_direction)",
                  "if close: verify only primary subject is fully visible; other entities are partial / soft bg / absent",
                  "verify same character's full body + body-part close-up are not combined",
                  "if close: verify two faces are not both rendered sharp"
                ],
                "rationale_summary": "한 shot 안 framing scale 가 일관되어야 모델이 단일 카메라 셔터로 해석. close/wide/medium 혼합은 multi-camera 합성으로 인식되어 인물 분열 / 부위 분리 / 시점 혼합 합성"
              },
              "rationale_summary": "Rule F/G/J 의 spatial consistency 룰 통합 — camera 위치 vs frame edge 모순, fg/bg 분리 시 공유 anchor 누락, framing scale 혼합. 3 sub-rule 모두 single t2i_prompt = single camera shutter 원칙의 spatial 측면."
            },
            "frame_spatial_contract": {
              "reason": "required_background_position",
              "constraints": [
                {
                  "constraint_id": "fsc_001",
                  "target_kind": "background",
                  "target_id": "",
                  "label": "three-branched signal light",
                  "screen_zone": "middle_center",
                  "depth_plane": "background",
                  "gesture_action": "none",
                  "gesture_target_label": ""
                }
              ]
            },
            "constraints": [
              "one prompt captures one still moment",
              "do not combine full-body framing and body-part close-up in one frame",
              "camera position and frame-edge elements must be geometrically consistent — see render_strategy.spatial_consistency.camera_frame_rule",
              "framing_scale binds entity rendering — close = primary subject only fully visible, wide/medium = no body-part close-up — see render_strategy.spatial_consistency.primary_framing_rule"
            ]
          },
          "id_policy": {
            "allowed_base_entity_ids": [
              "L12"
            ],
            "allowed_outlook_pairs": [],
            "must_use_composite_character_ids": true,
            "common_noun_required_when": [
              "a reproduced face inside any declared reproduction surface",
              "unregistered extra"
            ],
            "id_use_required_when": [
              "a physical photo/poster/document/picture-frame/map/key prop appearing in the scene — use its P## ID and preserve canonical content from entity_canon"
            ],
            "demographic_fallback_required": true,
            "face_identifiability_rule": {
              "use_entity_id_when": [
                "face identifiable: front, profile, three-quarter, or eyes closed",
                "OTS framing with any visible face"
              ],
              "common_noun_required_when": [
                "back to camera with no visible face",
                "silhouette or blurred outline only",
                "OTS with only back of head/shoulder, face entirely hidden"
              ],
              "id_use_summary": "C##/C##O## allowed only when face is identifiable AND framing is not dominated by an isolated body part",
              "rationale_summary": "C## 사용 시 얼굴 ref 가 자동 결합 — 얼굴이 안 보이는 인물에 C## 부착하면 ref 와 본문 묘사가 충돌"
            },
            "reproduction_surface_rule": {
              "applies": true,
              "id_use": "forbidden — generic descriptor with demographic only",
              "rationale_summary": "C##O## 은 얼굴 reference 이미지를 원본 해상도로 inject 하므로 사진/화면/반사 표면 안 얼굴이 표면 밖 실물 크기로 합성됨. 이 rule 은 표면 안 얼굴 (face) 에만 적용 — 물리적 사진/액자/문서 prop 자체는 P## ID 보존 (Patch A)."
            },
            "demographic_descriptor_policy": {
              "required_on_first_appearance": true,
              "applies_to_id_forms": [
                "C##",
                "C##O##"
              ],
              "token_count_range": [
                1,
                2
              ],
              "components": {
                "ethnicity": [
                  "Asian",
                  "East Asian",
                  "South Asian",
                  "Southeast Asian",
                  "Black",
                  "Middle Eastern",
                  "Hispanic",
                  "Latina",
                  "Latino",
                  "Caucasian"
                ],
                "gender": [
                  "man",
                  "woman",
                  "figure"
                ],
                "age_band": [
                  "in his/her 20s",
                  "in his/her 30s",
                  "in his/her 40s",
                  "young",
                  "middle-aged",
                  "elderly"
                ],
                "role_hint_from_outfit": [
                  "in worker uniform",
                  "in fisher workwear",
                  "in detective coat",
                  "in business suit"
                ]
              },
              "format_template_a": "C##O## in <옷 1-3 단어>, <demographic descriptor>",
              "format_template_b": "C##O##, <demographic descriptor>, <자세 표현>",
              "source_priority": [
                "visual_world_rules.region",
                "visual_world_rules.era",
                "shot context outfit hint"
              ],
              "scenario_dependency_ban": "no work-specific proper nouns (character names, place names) — common nouns only",
              "regional_consistency": "한 작품 내 region 표기는 일관 — 선택된 세분 ethnicity (East Asian / Southeast Asian / South Asian / Caucasian / Black / Hispanic / Latina|o / Middle Eastern) 를 작품 단위 고정"
            },
            "subject_reference_policy": [],
            "constraints": [
              "ID format for visible subjects (C## bare, C##O## composite IDs, or common-noun descriptor) is determined per-subject by the id_policy per-subject policy array — see constraints[2] below",
              "do not use C##O## for a reproduced face inside any element flagged by id_policy.reproduction_surface_rule.applies (declared reproduction surface) — use a generic descriptor with demographic instead. NOTE: the physical photo/poster/document/picture-frame/map/key prop itself MUST use its P## ID — the generic-descriptor rule applies only to a face reproduction, not to the prop entity.",
              "per-subject identity reference policy is provided in id_policy.subject_reference_policy[] — apply each subject's policy when emitting references: (a) id_and_outlook_required ⇒ subject must use C##O## composite ID; (b) base_id_required ⇒ subject must use bare C## only (no outlook); (c) generic_descriptor_allowed ⇒ subject must NOT use C## or C##O##, render with a common/demographic descriptor only. omitted subjects default to id_and_outlook_required",
              "when an ID is required for a subject (policy = id_and_outlook_required or base_id_required) and a C##O## (or C##) is introduced for the first time in this t2i_prompt, append a 1-2 token demographic descriptor (ethnicity + age band, or ethnicity + role hint) consistent with visual_world_rules.region — never use scenario-specific proper nouns. does not apply to generic_descriptor_allowed subjects (no ID emitted)",
              "when an ID is required (policy = id_and_outlook_required or base_id_required), use the format `C##O## in <옷 1-3 단어>, <demographic descriptor>` or `C##O##, <demographic descriptor>, <자세 표현>` — main characters with face refs may include this descriptor too (image inject takes precedence). does not apply to generic_descriptor_allowed subjects — those use a common/demographic descriptor with no C##/C##O##"
            ]
          },
          "background_binding": {
            "mode": "background_ref_attached",
            "bg_id": "L12B01",
            "reference_usage": "exact_background",
            "owned_objects": [
              "pursuit light",
              "rock outcrop",
              "rockfall",
              "signal lamp",
              "transmitter",
              "tunnel entrance"
            ],
            "camera_reference": null,
            "close_framing_skips_background_ref": false,
            "constraints": [
              "owned_objects lists all environment objects already drawn in the background reference — do not create new ones",
              "if focusing on an owned object, declare its reference-kind via the reference_phrase_kinds sidecar (producer contract)"
            ]
          },
          "continuity_elements_used": {
            "fixed_elements": [
              {
                "element_id": "ambient_night_lighting",
                "element_type": "environment_state",
                "character_name": "",
                "description": "The overall environment is enveloped in deep night darkness with a cool, muted lighting tone. This low-key illumination provides stark but restrained contrast across the scene, deliberately keeping the background shrouded in shadow while emphasizing specific points of interest or isolated light sources.",
                "element_scope": "full",
                "applies_to_shots": [
                  3,
                  10,
                  13
                ],
                "source_facts": [
                  "Night darkness and the concentrated warning-red lights create a cool, muted long view",
                  "Night-appropriate low-key illumination renders the engraved relief with cool, muted contrast",
                  "The dark night forest remains cool and muted"
                ],
                "visual_inferences": [
                  "The absence of a strong moon or general ambient light unifies the vast outdoor landscapes and the extreme macro close-ups under the same heavy, oppressive shadow."
                ],
                "creative_decisions": [
                  "Applying a consistent cool color palette and restrained lighting model across both wide shots and macro shots to maintain a tense, unified nocturnal atmosphere."
                ],
                "confidence": "high"
              },
              {
                "element_id": "three_branched_tree_emblem",
                "element_type": "environment_state",
                "character_name": "",
                "description": "A distinct symbol depicting a three-branched tree serves as a recurring visual mark. The precise structural design and geometric proportions of this emblem remain absolutely identical, whether it appears as an engraved physical relief on a tiny metal surface or as a brightly illuminated signal pattern in the distance.",
                "element_scope": "close",
                "applies_to_shots": [
                  10,
                  13
                ],
                "source_facts": [
                  "The three-branched tree emblem occupies the central portion of the image",
                  "its fully illuminated pattern forms the same three-branched tree emblem",
                  "송신기 표면에 Wyoming 공동체의 세 갈래 나무 문장이 새겨져 있다",
                  "같은 문장의 신호등을 한 번 켰다가 끈다"
                ],
                "visual_inferences": [
                  "The emblem is composed of sharp, well-defined lines that allow its shape to be instantly identifiable and perfectly matched regardless of the medium (metal engraving vs. light source)."
                ],
                "creative_decisions": [
                  "Ensuring the exact silhouette of the three branches perfectly matches between the macro physical prop and the distant light signal to visually confirm the narrative link of an internal traitor."
                ],
                "confidence": "high"
              }
            ],
            "previous_shot_refs": [
              {
                "scene_index": 7,
                "shot_index": 3,
                "ref_usage": "",
                "kind": "location"
              }
            ],
            "forward_zoom_targets": [],
            "cross_shot_id_substitution_rule": {
              "applies_when": "fixed_elements[i].character_name maps to a C## allowed in id_policy.allowed_base_entity_ids OR a C##O## in id_policy.allowed_outlook_pairs",
              "substitution": "replace the common-noun person reference inside fixed_elements[i].description (e.g. 'An Asian man' / 'a woman' / 'a figure') with the matched C## or C##O##",
              "double_description_forbidden": "do not describe the same character with both a C##/C##O## and a separate common-noun person reference in the same t2i_prompt — produces phantom extra figures",
              "preserve_pose_unchanged": "fixed_elements[i].description 의 자세/위치/상태 keyword 는 그대로 유지 — 인물 참조 토큰만 교체",
              "no_repeat_after_pose": "이미 t2i_prompt 안에서 같은 인물의 자세를 묘사했다면, 같은 fixed_element description 을 별도 문장으로 반복하지 마라",
              "rationale_summary": "fixed_element description 이 보통명사로 작성되었고 같은 인물의 C## 묘사가 t2i_prompt 에 따로 있으면 모델이 두 인물로 인식해 인원 수가 부풀려진다"
            },
            "ref_usage_constraints": {
              "zoom_in_detail": {
                "scope_summary": "same moment, same space, same primary subject; camera moved/zoomed onto a sub-region only",
                "forbidden_additions": [
                  "new persons not in the prior shot",
                  "new props not in the prior shot",
                  "new background elements not in the prior shot",
                  "new poses, gestures, or expressions not in the prior shot",
                  "different moments of action"
                ],
                "required_phrasings": [
                  "describe only the focused sub-region in concrete terms",
                  "explicitly state the same environment (lighting, background) is preserved",
                  "use continuity tokens such as 'the same figure', 'the same room', 'the same surface'"
                ],
                "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy applies per-subject; policy enum (id_and_outlook_required / base_id_required / generic_descriptor_allowed) determines C##/C##O## use",
                "rationale_summary": "zoom_in_detail = 앞 샷의 특정 영역을 확대한 '같은 사진'. 새 정보 추가 금지 — 현재 프레임이 보여주는 부분만 정확히 묘사"
              },
              "exact_background": {
                "scope_summary": "same room, possibly different moment or angle; background carries through, persons/actions are described fresh for this shot",
                "permitted_additions": [
                  "new persons, poses, expressions, actions for this shot's moment",
                  "different camera angle within the same room"
                ],
                "background_consistency_rule": "background elements established in the prior shot must remain consistent (no contradictory walls, doors, furniture)",
                "ignore_keep_handled_elsewhere": "ignore/keep directives are auto-applied via reference image processing — do not echo them inside t2i_prompt",
                "rationale_summary": "같은 공간, 새 순간 — 인물·동작은 이 샷의 instantaneous moment 로 묘사하되 배경은 prior shot 과 일관"
              },
              "atmosphere_reference": {
                "scope_summary": "different room/floor/angle entirely — only mood/tone is shared with the prior shot",
                "forbidden_imports": [
                  "furniture layout from the prior shot",
                  "wall positions from the prior shot",
                  "specific architectural features from the prior shot"
                ],
                "required_handling": "describe the new space's background directly in this shot; carry through only color tone or lighting mood (e.g. 'cool gray ambient' / 'dim warm pool')",
                "rationale_summary": "atmosphere_reference = 분위기 reference 만, 레이아웃 reference 아님. 가구·벽 배치를 끌어오면 모델이 두 공간을 한 공간으로 합성"
              }
            },
            "view_consistency": {
              "framing_scope_options": [
                "third_person",
                "close_up"
              ],
              "single_camera_rule": "one t2i_prompt = one camera position. third-person full body and close-up of the same character must not be combined in a single prompt.",
              "third_person_handling": "describe persons with full or partial body in frame; single-camera consistency forbids combining a partial close-up and the same character's full body in one prompt",
              "close_up_handling": "an isolated framed area of a single subject must not coexist with that subject's full-body description in the same prompt",
              "focus_on_is_focus_area_not_view_switch": "any framing-intent wording designates the area shown by the single camera position; it does not switch the view",
              "subject_reference_policy_cross_ref": "id_policy.subject_reference_policy — subjects with policy = 'base_id_required' or 'generic_descriptor_allowed' must not use outlook reference (C##O## form); additionally, subjects with policy = 'generic_descriptor_allowed' must not use bare C## either — render with a common/demographic descriptor only",
              "prop_orientation_rule": "props the character holds must face one direction only — toward camera OR toward character, not both. forbidden contradiction example: 'photo angled partly toward the camera' + 'eyes fixed on the photograph' (양방향 모순)",
              "mixing_forbidden": true,
              "rationale_summary": "single t2i_prompt = single camera shutter at 1/1000s. 시점 혼합은 모델이 두 카메라 합성으로 해석해 인물이 두 번 등장 또는 부위 분리"
            },
            "constraints": [
              "fixed_elements are continuity inputs derived from prior shots — treat them as binding context, not as source facts to be invented",
              "when fixed_elements[i].character_name maps to a C## in id_policy.allowed_base_entity_ids or a C##O## in id_policy.allowed_outlook_pairs, replace the common-noun person reference inside fixed_elements[i].description with that C##/C##O## and do not also write a separate common-noun reference for the same character in the same t2i_prompt",
              "previous_shot_refs with ref_usage='zoom_in_detail' must preserve the same moment, the same space, and the same lighting as the prior shot — no new persons, props, background elements, poses, or moments",
              "previous_shot_refs with ref_usage='exact_background' may freely describe this shot's persons, poses, and moment, but must keep background elements consistent with the prior shot's established room",
              "previous_shot_refs with ref_usage='atmosphere_reference' must describe a new space directly — do not import furniture layout or wall positions from the prior shot, only mood/tone",
              "one t2i_prompt = one camera position — third-person full body and close-up of the same character must not be combined in a single shot. any framing-intent wording designates the area shown by the single camera position, not a view switch (id_policy.subject_reference_policy applies per-subject)",
              "props held by characters face one direction only — toward camera OR toward character, never both"
            ]
          },
          "asset_requirements": {
            "required_refs": [
              {
                "kind": "background",
                "id": "L12B01",
                "policy": "required"
              }
            ],
            "forbidden_refs": [],
            "readiness_policy": "block_if_missing",
            "constraints": [
              "do not imply or describe a reference image that is not listed in required_refs (no phantom references)",
              "if a required character_outlook ref is missing, scene image generation must block before the image stage"
            ]
          },
          "render_contracts": [],
          "_card_metadata": {
            "lift_status": {
              "rule_a_lifted": true,
              "rule_c_lifted": true,
              "rule_e_lifted": false,
              "rule_h_lifted": true,
              "id_policy_composite_lifted": true,
              "id_policy_close_framing_face_lifted": true,
              "id_policy_reproduction_surface_lifted": true,
              "continuity_cross_shot_id_substitution_lifted": true,
              "continuity_ref_usage_constraints_lifted": true,
              "continuity_view_consistency_lifted": true,
              "continuity_constraints_extended": true,
              "spatial_camera_frame_lifted": true,
              "spatial_fg_bg_shared_anchor_lifted": true,
              "spatial_primary_framing_lifted": true,
              "spatial_constraints_extended": true
            },
            "rule_source": {
              "camera_rule": "A",
              "owned_rule": "C",
              "close_skip_rule": "E",
              "demographic_rule": "H",
              "composite_id_rule": "C## 사용 규칙 — 참조 이미지 연동 (v17:32-64)",
              "close_framing_face_rule": "극단 클로즈업 표현 (v17:66-85)",
              "reproduction_surface_rule": "사진·포스터·화면·거울 속 인물 규칙 (v17:133-146)",
              "cross_shot_id_substitution_rule": "v20/scene_detail/system.md#continuity-section",
              "ref_usage_rule": "v20/scene_detail/system.md#continuity-section",
              "view_consistency_rule": "v20/scene_detail/system.md#continuity-section",
              "camera_frame_rule": "v20/scene_detail/system.md#spatial-consistency",
              "fg_bg_shared_anchor_rule": "v20/scene_detail/system.md#spatial-consistency",
              "primary_framing_rule": "v20/scene_detail/system.md#spatial-consistency"
            }
          }
        },
        "render_prompt_card_hash": "9f3b453756eaffbc"
      }
    ],
    "consumed_bg_catalog_hash": "ef9206e193be155b",
    "consumed_shot_binding_hash": "82ab9c737571f47f"
  },
  "step_id": "scene_detail",
  "run_id": "140ccd91-2512-46ee-8649-d489f1360462",
  "resolved_model": "gemini-pro",
  "updated_at": "2026-09-05T08:28:08.534312+00:00",
  "project_config_snapshot": {}
}