Start the multi-AI pipeline with a given request. Guides through plan -> review -> implement -> review workflow.
Sign in to like and favorite skills
You are starting the multi-AI pipeline. Follow this process exactly.
First, read the standards that guide all reviews:
skill/multi-ai/reference/standards.mdRemove old
.task/rm -rf .task mkdir -p .task
Write the user's request to
.task/user-request.txtWrite
.task/plan.json{ "id": "plan-YYYYMMDD-HHMMSS", "title": "Short descriptive title", "description": "What the user wants to achieve", "requirements": ["req1", "req2"], "created_at": "ISO8601", "created_by": "claude" }
Research the codebase and create
.task/plan-refined.json{ "id": "plan-001", "title": "Feature title", "description": "What the user wants", "requirements": ["req1", "req2"], "technical_approach": "Detailed how-to", "files_to_modify": ["path/to/file.ts"], "files_to_create": ["path/to/new.ts"], "dependencies": [], "estimated_complexity": "low|medium|high", "potential_challenges": ["Challenge and mitigation"], "refined_by": "claude", "refined_at": "ISO8601" }
Run reviews in sequence. Fix issues after each before continuing:
Invoke /review-sonnet
.task/review-sonnet.jsonneeds_changes.task/plan-refined.jsonInvoke /review-codex
.task/review-codex.jsonneeds_changesapprovedInvoke /implement-sonnet
This skill will:
.task/plan-refined.json.task/impl-result.jsonRun reviews in sequence. Fix issues after each before continuing:
Invoke /review-sonnet
.task/review-sonnet.jsonneeds_changesInvoke /review-codex
.task/review-codex.jsonneeds_changesapprovedWrite
.task/state.json{ "state": "complete", "plan_id": "plan-001", "completed_at": "ISO8601" }
Report success to the user with:
| File | Purpose |
|---|---|
| Original user request |
| Initial plan |
| Refined plan with technical details |
| Implementation result |
| Sonnet review output |
| Codex review output |
| Pipeline state |
| Path | Purpose |
|---|---|
| Review criteria and coding standards |
| JSON schemas for structured output |