メニュー
OpenSpec lifecycle orchestration for /workflow chains. TRIGGER when: coordinating the analyst-architect-developer-reviewer chain or managing spec status transitions. SKIP: writing an individual spec artifact (use spec-research, spec-design, or spec-review).
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 | spec-workflow |
| description | OpenSpec lifecycle orchestration for /workflow chains. TRIGGER when: coordinating the analyst-architect-developer-reviewer chain or managing spec status transitions. SKIP: writing an individual spec artifact (use spec-research, spec-design, or spec-review). |
Spec-driven development protocol for coordinating multi-agent workflows.
CRITICAL: The conversation_id MUST always be a UUID (format: xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx). NEVER use descriptive names, slugs, or human-readable strings. If no UUID is available, generate one with uuidgen or python3 -c "import uuid; print(uuid.uuid4())". Self-check: does the conversation_id match the pattern [0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}? If not, STOP and use a real UUID.
.scaffolding/conversations/{conversation_id}/specs/
proposal.md -- WHY + WHAT (analyst)
design.md -- WHAT + HOW (architect)
tasks.md -- checklist (architect)
Tracked in .scaffolding/conversations/{conversation_id}/context.json:
| Status | Meaning | Transition |
|---|---|---|
exploring | Research phase | -> drafting when proposal done |
drafting | Design + tasks | -> implementing when tasks done |
implementing | Code changes | -> reviewing when tasks complete |
reviewing | Verification | -> complete when verified |
complete | Done | -> archived on request |
archived | Historical | Terminal state |
| Phase | Agent | Reads | Writes | Gate |
|---|---|---|---|---|
| 1. Analyze | analyst | requirements | proposal.md | proposal written |
| 1b. Research (conditional) | researcher | requirements | ResearchPack | score >= 80 |
| 2. Design | architect | proposal.md | design.md, tasks.md | score >= 85 |
| 3. Apply | developer | tasks.md, design.md | source code | validation passes |
| 4. Verify | reviewer | design.md, tasks.md | review report | no criticals |
Analyst flags need for research in proposal.md. Researcher is invoked ONLY when the task involves:
Analyst writes proposal directly (no researcher) for:
| Task Type | Use Specs? | Reason |
|---|---|---|
| New feature | Yes | Needs design, multi-file |
| Architecture change | Yes | Needs decisions documented |
| Multi-agent workflow | Yes | Needs coordination |
| Bug fix (simple) | No | Direct to developer |
| Typo / config tweak | No | Direct to developer |
| Single-file refactor | No | Direct to developer |
uuidgen. MUST be a UUID, never a descriptive name.mkdir -p .scaffolding/conversations/{conversation_id}/specs/.scaffolding/conversations/{conversation_id}/specs/{
"conversation_id": "abc-123",
"status": "implementing",
"created_at": "2026-01-15T10:00:00Z",
"description": "Feature description",
"agents_involved": ["architect", "developer"]
}
Custom schema: .scaffolding/openspec/schemas/scaffolding-workflow/schema.yaml
Config: .scaffolding/openspec/config.yaml
Templates: .scaffolding/openspec/schemas/scaffolding-workflow/templates/
SendMessage recipient validation and worktreePath safety (CWE-59). TRIGGER when: validating a SendMessage `to:` recipient against the agent whitelist, or a worktreePath from another agent before acting on it. SKIP: routing topology (agent Comms Protocol); git ops (use git-operations); worktree recovery (use worktree-management).
Python backend patterns: layered architecture, async I/O, dependency injection, repository/service separation. TRIGGER when: creating routes, models, schemas, or services in a Python backend. SKIP: REST contract design (use api-design); schema/index tuning (use database-optimization). (Examples: FastAPI + SQLAlchemy + Pydantic.)
Knowledge distillation methodology: candidate extraction, confidence scoring, tier routing, conversation-scoped mode. TRIGGER when: running /learn or /distill, deciding if an insight is memory-worthy, or scoring a knowledge candidate. SKIP: ad-hoc memory reads/writes (use agent-memory); vector storage (use semantic-memory-store).
Build minimal, reversible implementation plans. TRIGGER when: turning research into a plan, or sequencing work before writing code. SKIP: gathering documentation (use research-methodology); writing OpenSpec design.md/tasks.md (use spec-design).
Score ResearchPacks (>=80) and Implementation Plans (>=85) against quality gates. TRIGGER when: gating research or a plan before implementation. SKIP: writing or running tests (use testing-strategy); security threat review (use security-review-checklists).
Optimize Claude Code context-window usage for accuracy and cost. TRIGGER when: hitting context limits, structuring prompts for an agent, or trimming what gets injected into a task. SKIP: persisting knowledge across sessions (use agent-memory); semantic recall tuning (use semantic-memory-mcp).