ALWAYS activate before implementing ANY feature, plan, or fix. End-to-end implementation with automatic workflow detection.
End-to-end implementation with automatic workflow detection.
Principles: YAGNI, KISS, DRY | Token efficiency | Concise reports
@[skills/cook] <natural language task OR plan path> [--flag]
| Flag | Behavior |
|---|---|
--interactive | Full workflow with user approval gates (default) |
--fast | Skip research, scout→plan→code |
--parallel | Multi-agent parallel execution |
--auto | Auto-approve all steps |
--no-test |
| Skip testing step |
| Input Pattern | Mode |
|---|---|
Path to plan.md or phase-*.md | code — execute existing plan |
| "fast", "quick" | fast |
| "trust me", "auto" | auto |
| 3+ features or "parallel" | parallel |
| Default | interactive |
[Intent Detection] → [Research?] → [Review] → [Plan] → [Review] → [Implement] → [Review] → [Test?] → [Finalize]
| Mode | Research | Testing | Review Gates |
|---|---|---|---|
| interactive | Yes | Yes | User approval each step |
| auto | Yes | Yes | Auto if score ≥ 9.5 |
| fast | No | Yes | User approval each step |
| parallel | Optional | Yes | User approval each step |
| no-test | Yes | No | User approval each step |
| code | No | Yes | User approval per plan |
plan.md./docs if changes warrant| Phase | Subagent |
|---|---|
| Research | researcher |
| Scout | @[skills/scout] |
| Plan | planner |
| Testing | tester, debugger |
| Review | code-reviewer |
references/intent-detection.md — Detection rules and routingreferences/workflow-steps.md — Detailed step definitionsreferences/review-cycle.md — Interactive and auto review