# context-package — 단일 컴파일러(context_package.py --compile)가 발급한 스켈레톤.
# spawn 전: 아래 placeholder를 채우고 `python3 .claude/hooks/context_package.py hyeonworks/state/context-packages/hyeonworks-vnext-v1-direction/DES-VISUAL-20260718T114118Z.pkg.yaml` 가
# exit 0(통과)이어야 워커를 spawn한다. 스키마·규칙: org-os/06-agent-work/context-package-spec.yaml
# 아직 채워야 하는 필드(발급 시점):
#   - 필수 필드 비어있음(placeholder 미충전): objective
#   - 필수 필드 비어있음(placeholder 미충전): allowed-tools
#   - 필수 필드 비어있음(placeholder 미충전): task-boundaries
#   - 필수 필드 비어있음(placeholder 미충전): must-read
#   - 필수 필드 비어있음(placeholder 미충전): acceptance-tests
#   - 필수 필드 비어있음(placeholder 미충전): non-goals
#   - 필수 필드 비어있음(placeholder 미충전): evidence-plan
workflow-id: hyeonworks-vnext-v1-direction
task-id: direction-comparison-audit
mode: converge
tier: standard
model: sonnet
effort: medium
assigned-lens: null
target-role-agent: DES-VISUAL
workspace: hyeonworks
target-repo: hyeonworks
objective: >-
  세 개의 독립 coded direction과 동일 viewport full-size preview를 함께 읽고, 모든 쌍이
  layout/navigation/type/imagery/motion/primitives 중 최소 네 축에서 실제로 갈라지는지 감사한다.
  reference 중복·primitive collision·반응형 위장 여부를 확인해 pass/revise/re-diverge를 판정한다.
output-format: report.yaml (report-header BLUF로 시작) + 최종 메시지로 report-path + 1줄 bottom-line
  반환
allowed-tools: [Read, exec_command, view_image, apply_patch]
task-boundaries: >-
  이 run은 sibling isolation의 유일한 예외로 세 방향 원본·concept·full-size preview와 direction-set을
  모두 읽는다. 비교 결과는 지정된 comparative-divergence-audit report 한 파일에만 기록하며 방향 코드,
  concept, direction-set 또는 제품 결정을 수정하지 않는다.
non-goals:
  - 선호 방향 선택 또는 요소 혼합
  - coded slice 수정, 제품 전략 변경, prototype 구현
  - 스크린샷 존재만으로 시각 품질 pass 선언
  - 외부 네트워크·배포·사용자 효과 검증 주장
must-read:
  - hyeonworks/completion-records/hyeonworks-vnext-v1-direction/DES-DIRECTOR-20260718T111440Z-1.report.yaml
  - hyeonworks/completion-records/hyeonworks-vnext-v1-direction/DES-VISUAL-20260718T113705Z.report.yaml
  - hyeonworks/design-direction/hyeonworks-vnext-v1/directions/ledger-studio/concept.yaml
  - hyeonworks/design-direction/hyeonworks-vnext-v1/directions/ledger-studio/index.html
  - hyeonworks/design-direction/hyeonworks-vnext-v1/directions/signal-trace/concept.yaml
  - hyeonworks/design-direction/hyeonworks-vnext-v1/directions/signal-trace/index.html
  - hyeonworks/design-direction/hyeonworks-vnext-v1/directions/field-manual/concept.yaml
  - hyeonworks/design-direction/hyeonworks-vnext-v1/directions/field-manual/index.html
  - hyeonworks/design-direction/hyeonworks-vnext-v1/previews/ledger-desktop.png
  - hyeonworks/design-direction/hyeonworks-vnext-v1/previews/signal-desktop.png
  - hyeonworks/design-direction/hyeonworks-vnext-v1/previews/manual-desktop.png
  - hyeonworks/design-direction/hyeonworks-vnext-v1/previews/comparison.png
inherited-decisions: []
acceptance-tests:
  - criterion: "ledger↔signal, ledger↔manual, signal↔manual 세 쌍을 모두 비교하고 각 differing-axes가 4개 이상이다."
    verify: "python3 .claude/hooks/lint_design_direction.py hyeonworks/completion-records/hyeonworks-vnext-v1-direction/DES-VISUAL-20260718T113705Z-1.report.yaml comparative-divergence-audit"
  - criterion: "세 full-size preview ref/SHA를 실제 파일에 대조하고 primitive collision 또는 blocker를 숨기지 않는다."
    verify: "python3 .claude/hooks/lint_design_direction.py --divergence-bundle hyeonworks/completion-records/hyeonworks-vnext-v1-direction/DES-VISUAL-20260718T113705Z-1.report.yaml hyeonworks/completion-records/hyeonworks-vnext-v1-direction/DES-VISUAL-20260718T113705Z.report.yaml hyeonworks/completion-records/hyeonworks-vnext-v1-direction/DES-DIRECTOR-20260718T111440Z-1.report.yaml"
evidence-plan:
  - claim: "세 방향은 동일 viewport의 실제 브라우저 렌더에서 구조적으로 분리된다."
    grade: E4
    receipt: hyeonworks/design-direction/hyeonworks-vnext-v1/previews/comparison.png
  - claim: "audit가 charter와 direction-set의 exact SHA 및 세 pair를 완전하게 결속한다."
    grade: E5
    receipt: "lint_design_direction.py --divergence-bundle exit 0"
method-selection:
  method-id: compare-directions
expected-output:
  report-header:
    bottom-line: ''
    decision-needed:
      needed: false
      approver: null
    confidence:
      value: Med
      derived-from: evidence
    risks: []
    evidence: []
token-budget:
  max-input-tokens: 60000
  max-output-tokens: 8000