メニュー
dure decomposition — decompose the active fixed spec into milestone/epic/issue per-item files (source of truth), review them, and regenerate the index. Every issue gets a non-empty, testable acceptance criterion.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
| name | plan |
| description | dure decomposition — decompose the active fixed spec into milestone/epic/issue per-item files (source of truth), review them, and regenerate the index. Every issue gets a non-empty, testable acceptance criterion. |
| argument-hint | [spec-slug] |
| allowed-tools | Bash Read Grep Glob Write Task |
| disable-model-invocation | true |
Target spec: $ARGUMENTS (if empty, use the slug in .dure/active)
Goal: per spec §5, decompose a FIXED spec into milestone ⊃ epic ⊃ issue.
Per-item files are the source of truth (.dure/roadmap/{milestones,epics,issues}/<id>.md);
index.md is generated. The interview runs in the user's language; artifacts are English (ADR-0007).
${CLAUDE_PLUGIN_ROOT}/scripts/dure-bootstrap.sh (idempotent).$ARGUMENTS > .dure/active > ask the user (in their language)..dure/specs/<slug>.md.Produce per-item files with YAML front matter + body:
id, slug, type: milestone, title, status: todo, github: null, epics: [<ids>].+ milestone: <id>, issues: [<ids>].+ milestone, epic, acceptance: [<testable criteria>] (Gherkin Given/When/Then; INVEST).Stable IDs (I1.3.1/I1.3.3): ids are derived deterministically from titles via
${CLAUDE_PLUGIN_ROOT}/scripts/dure-slug.sh "<title>" (ASCII kebab; hash fallback for non-ASCII),
disambiguated with a numeric suffix on collision. Re-running /dure:plan on the same spec MUST
reuse existing ids/slugs (read existing files first; do not renumber or overwrite unchanged content).
Every issue MUST have a non-empty, testable acceptance (spec AC3).
Task the redteam-critic subagent over the draft decomposition to flag missing,
duplicate, or oversized issues and untestable acceptance criteria. Revise until clean.
${CLAUDE_PLUGIN_ROOT}/scripts/dure-index.py → regenerate roadmap/index.md from per-item files.${CLAUDE_PLUGIN_ROOT}/scripts/dure-doctor.py → confirm integrity (status pass; warnings ok).Report the milestone/epic/issue tree and point the user to /dure:sync (push to GitHub) and
/dure:status (progress).
Implementation status (E1.3): the index generator (
dure-index.py) and integrity check (dure-doctor.py) are deterministic and tested. The decomposition itself (steps 1–2) is driven by these instructions; stable-id reuse on re-decomposition is enforced by reading existing files first.
dure direction — gather survey/audit/status evidence, run a deep interview that converges a project direction, write a validated .dure/directions/<slug>.md proposal, and hand the chosen direction to /dure:plan. Report-only at the validator boundary; never writes the roadmap.
dure survey — run the deterministic state survey and surface forward-looking planning signals (closable milestones, empty epics) as evidence for /dure:interview and /dure:plan. Report-only; never creates issues.
dure audit — run the deterministic repo scanner and surface findings (debt / missing-test / progress-consistency signals) as evidence for /dure:interview and /dure:plan. Report-only; never creates issues.
dure deep interview — interview a vague requirement until understanding converges quantitatively, then crystallize it into .dure/specs/<slug>.md. Four traits — evidence-grounded, red-team, auto-research, quantitative convergence gate.
dure sync — idempotently mirror the local roadmap to GitHub Issues/Milestones. plan/apply separation; structure pushed, status pulled, conflicts reported. gh CLI first.
dure progress report — aggregate per-item roadmap status into per-milestone completion, blockers, and local↔GitHub conflicts.