pgermishuys/opencode-weave/sdd-orchestration
Claude Code and agent skill by pgermishuys/opencode-weave
- community
Checking that the SKILL.md file is still in the repository.
Spec-Driven Development: Orchestration protocol — working-spec.json lifecycle, phase management, initialization, and delegation guidance for SDD workflows
Install
npx vskill@latest install pgermishuys/opencode-weave/sdd-orchestrationSKILL.md
<SDDOrchestration>
Spec-Driven Development Orchestration
When the user wants to build a feature using Spec-Driven Development (SDD), you manage the full lifecycle: initialize → constitute → specify → clarify → plan → analyze → implement → review. You handle the interactive phases (constitution, specification, clarification) yourself. You delegate autonomous phases to specialist agents.
Loading SDD Skills
Before each phase, load the relevant skill using the skill tool:
| Phase | Skill to load | Why | |
|---|---|---|---|
| Constitution | sdd-constitution | Constitution template, versioning rules, quality rules | |
| Specification | sdd-specification | Spec format (FR-001, SC-001), validation checklist | |
| Clarification | sdd-clarify | Ambiguity taxonomy, prioritization, questioning protocol |
When delegating, tell the delegate which skill to load:
- Pattern → tell it to load
sdd-planning(SDD plan format, task breakdown, Weave plan bridge) - Thread → tell it to load
sdd-analysis(6-pass consistency analysis methodology)
Shared State: .specify/working-spec.json
This file tracks SDD progress across sessions. Read it at the start of any SDD-related conversation.
Status values (ordered phases): initialized → constituting → specifying → clarifying → planning → analyzing → implementing → reviewing → complete
Update after each phase transition: set status, append to phase_history, update updated_at, populate paths as artifacts are created.
Starting a New Feature
When the user says "I want to build X" (or similar, with or without mentioning SDD):
- Generate a feature slug (lowercase, hyphens — e.g. "user-authentication")
- Confirm slug and goal with the user
- Create directories:
.specify/memory/(if needed),.specify/features/{slug}/,.specify/features/{slug}/checklists/ - Write
.specify/working-spec.jsonwith status"initialized", all paths populated - If
.specify/memory/constitution.mdexists: read it, summarize, ask if updates needed - If not: load the
sdd-constitutionskill and proceed to constitution drafting
Phase Guidance
Constitution (interactive — you do this): Load sdd-constitution skill. Update status → "constituting". Ask about core principles (3-5) and governance. Offer sensible defaults. Write .specify/memory/constitution.md using the format from the skill.
Specification (autonomous — you do this): Load sdd-specification skill. Update status → "specifying". Read constitution, write spec at .specify/features/{slug}/spec.md using the format from the skill. Create requirements checklist at checklists/requirements.md.
Clarification (interactive — you do this): Load sdd-clarify skill. Update status → "clarifying". Scan spec for ambiguities using the methodology from the skill. Ask up to 5 questions, one at a time, multiple-choice with recommended option. Update spec after each answer.
Planning (delegate to Pattern): Tell Pattern to load the sdd-planning skill. Delegate with: feature goal, slug, spec path, constitution path, feature directory. Pattern creates plan.md, tasks.md, and .weave/plans/{slug}.md. Update status → "planning".
Analysis (delegate to Thread): Tell Thread to load the sdd-analysis skill. Delegate with all artifact paths. Thread writes analysis.md. Update status → "analyzing".
Review (delegate to Weft/Warp): Give Weft artifact paths, ask for APPROVE/REJECT. For security review, use Warp. Update status → "reviewing".
Implementation (hand off to Tapestry): Tell the user to run /start-work. The Weave plan is at .weave/plans/{slug}.md. Update status → "implementing".
Soft Sequencing
Before each SDD action, check working-spec.json status. If the user skips steps, note it conversationally and suggest the recommended next step — but always proceed if they insist.
Example: "I notice we haven't clarified the spec yet. The clarification step often catches ambiguities that save rework later. Want me to run through it, or should I plan from the current spec?"
</SDDOrchestration>
Security
- Security scanPassed, 93 of 100Pattern checks on the skill's files.
- Model reviewNot runA model review is scheduled for the next rescan.
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingexamples/config/github-speckit/skills/sdd-orchestration/SKILL.md
Popularity
- Repository stars
- 142
- Repository forks
- 9
- 7-day trend score
- 25
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/pgermishuys/opencode-weave/sdd-orchestration)