{"id":"9dfc9864-5cf5-4aed-bc8b-a2596244eadf","shortId":"wn6XU3","kind":"skill","title":"feature-development","tagline":"Run a full, policy-compliant feature development flow with structured review phases","description":"# feature-development skill\n\n## ⛔ CRITICAL: PLANNING ONLY - NO CODE\n\n**`/cook` produces ONLY an artifact file. Do NOT implement any code.**\n\n- ❌ NO adding/modifying source files\n- ❌ NO creating classes, functions, or components\n- ❌ NO running implementation tests\n- ❌ NO writing actual code beyond file path references\n- ❌ NO \"Implementation\" step in todo list\n- ✅ ONLY produce the artifact file (`cook/*.cook.md`)\n\nAfter cooking is complete, user will **separately** request implementation.\n\n**Correct todo list for /cook:**\n1. Create artifact file (FIRST!)\n2. Product review - scope & acceptance criteria\n3. UX review - interaction design\n4. Architecture review - technical approach\n5. Security review - risk assessment\n6. QA review - test plan\n7. Finalize artifact\n\n**WRONG - do NOT include:**\n- ❌ Implementation\n- ❌ Write code\n- ❌ Any coding step\n\n---\n\n## ⚡ IMMEDIATE FIRST ACTION - EXECUTE NOW\n\n**Your VERY FIRST tool call MUST be Write to create the artifact file.**\n\nExecute this Write call IMMEDIATELY:\n\n```\nWrite(\n  file_path=\"cook/<feature-slug>.<YYYY-MM-DD>.cook.md\",\n  content=\"<full skeleton below>\"\n)\n```\n\n**BLOCKED ACTIONS until artifact exists:**\n- ❌ Read (any file)\n- ❌ Glob/Grep (any search)\n- ❌ Task/Explore (any agent)\n- ❌ Bash (any command)\n\n**ONLY ALLOWED first action:** Write artifact file.\n\nA PreToolUse hook will BLOCK other tools until artifact exists.\n\n```markdown\n# Cooking Result\n\n## Dish\n<1-2 sentence description of what we're building>\n\n## Status\nraw\n\n## Cooking Mode\nwell-done\n\n## Current Phase\nStep 0.0 - Artifact Created\n\n## Ownership\n- Decision Owner: _TBD_\n- Reviewers: _TBD_\n- Approved by: _TBD_\n\n---\n\n# Phase 0 - Project Policy & Context\n\n## Sources Scanned\n| File | Status | Key Rules |\n|------|--------|-----------|\n| CLAUDE.md | _Pending_ | |\n| README.md | _Pending_ | |\n| .claude/agents/*.md | _Pending_ | |\n\n## Hard Rules (must not be violated)\n_Pending..._\n\n## Preferred Patterns\n_Pending..._\n\n## Detected Conflicts\n_Pending..._\n\n## Policy Alignment Risk\n_Pending..._\n\n---\n\n# Step 1 - Read the Order\n\n## Feature Summary\n_Pending..._\n\n## Affected Modules/Components\n| Module | Impact | Risk Level |\n|--------|--------|------------|\n| | | |\n\n## Dependencies\n_Pending..._\n\n## Microwave Blocker Check\n_Pending..._\n\n---\n\n# Step 2 - Ingredient Approval (Product Review)\n\n## Product Decision\n_Pending: Approved / Rejected / Deferred_\n\n## Scope\n\n### In Scope\n- _Pending..._\n\n### Out of Scope\n- _Pending..._\n\n### Non-goals\n- _Pending..._\n\n## User Value\n_Pending..._\n\n## Assumptions\n- _Pending..._\n\n---\n\n# Step 3 - Presentation Planning (UX Review)\n\n## UX Decision\n_Pending: Required / Not Required_\n\n## User Flow\n_Pending..._\n\n## UI Components Affected\n| Component | Change Type | Notes |\n|-----------|-------------|-------|\n| | | |\n\n## Accessibility Considerations\n_Pending..._\n\n---\n\n# Step 4 - Implementation Plan\n\n## Architecture Decision\n\n### Selected Approach\n_Pending..._\n\n### Alternatives Considered\n| Option | Pros | Cons | Decision |\n|--------|------|------|----------|\n| Option A | | | Rejected: _reason_ |\n| Option B | | | **Selected**: _reason_ |\n\n### Trade-offs\n- Sacrificing: _what we give up_\n- Gaining: _what we get_\n\n## Patch Plan\n\n### Files to Modify\n| File | Change | Risk |\n|------|--------|------|\n| | | |\n\n### Commit Sequence\n1. _commit message_\n2. _commit message_\n\n### High-risk Areas\n- _area needing extra attention_\n\n---\n\n# Step 5 - QA Review\n\n## Test Plan\n\n### Test Cases\n| # | Scenario | Given | When | Then |\n|---|----------|-------|------|------|\n| 1 | Happy path | | | |\n| 2 | Edge case | | | |\n| 3 | Error case | | | |\n\n### Edge Cases\n- _edge case 1_\n- _edge case 2_\n\n### Acceptance Criteria\n- [ ] Given _context_, when _action_, then _result_\n- [ ] Given _context_, when _action_, then _result_\n\n### Regression Checks\n- _existing feature to verify_\n\n---\n\n# Step 6 - Security Review\n\n## Security Status\n- Reviewed: _yes/no_\n- Risk level: _low/medium/high_\n\n## Security Checklist\n| Check | Status | Notes |\n|-------|--------|-------|\n| Input validation | _Pending_ | |\n| Auth/authz | _Pending_ | |\n| Data exposure | _Pending_ | |\n| Injection vectors | _Pending_ | |\n\n## Issues Found\n_Pending..._\n\n---\n\n# Step 7 - Documentation\n\n## Documentation Updates\n| File | Change Needed |\n|------|---------------|\n| | |\n\n## New Documentation Needed\n_Pending..._\n\n---\n\n# Risk Management\n\n## Pre-mortem (3 scenarios required)\n| # | What Could Go Wrong | Likelihood | Impact | Mitigation |\n|---|---------------------|------------|--------|------------|\n| 1 | | | | |\n| 2 | | | | |\n| 3 | | | | |\n\n## Rollback Plan\n1. _step 1_\n2. _step 2_\n\n## Blast Radius\n- Affected users/modules: _list_\n- Feature flag: _yes/no (name)_\n- Rollout strategy: _immediate/gradual/canary_\n\n---\n\n# Decision Log\n\n| Date | Phase | Decision | Rationale |\n|------|-------|----------|-----------|\n| <today> | Step 0.0 | Artifact created | Starting cook flow |\n```\n\n**DO NOT:**\n- Read CLAUDE.md first\n- Explore codebase first\n- Run any searches first\n- Use Task/Explore agents first\n\n**FIRST action = Create artifact file. No exceptions.**\n\n---\n\n## Purpose\nCook features through a structured, multi-phase development flow.\n\nThe goal is not speed, but correctness, safety, and product discipline.\nEvery dish must be properly prepared before serving.\n\n---\n\n## Inputs\n\n- **feature_description** (string, required)\n  Plain-language description of the feature or change (the order).\n\n- **instruction_file** (string, optional)\n  Related specification or requirements file (the recipe).\n\n- **mode** (enum: well-done | microwave, default: well-done)\n  Determines cooking thoroughness and review phases.\n\n- **dry-run** (boolean, default: false)\n  Preview mode - shows what would happen without executing.\n\n- **validate** (string, optional)\n  Path to existing artifact to validate without re-cooking.\n  Example: `/cook --validate cook/feature.cook.md`\n\n- **no-validate** (boolean, default: false)\n  Skip auto-validation after artifact generation.\n\n- **interactive** (boolean, default: false)\n  Launch interactive menu for artifact management.\n  Example: `/cook --interactive`\n\n---\n\n## Interactive Mode\n\nUse `--interactive` to launch the artifact management menu:\n\n```\n/cook --interactive\n```\n\n### What Interactive Mode Does\n\n1. **Scans for artifacts** in `cook/*.cook.md`\n2. **Presents a picker** with available artifacts\n3. **Shows action menu**:\n   - Validate artifact\n   - Compare artifacts (diff)\n   - View status summary\n4. **Executes selected action**\n\n### Interactive Flow\n\n```\n/cook --interactive\n   |\n   v\n┌─────────────────────────────────────┐\n│  Select artifact:                   │\n│  > dry-run-validation.2026-01-10    │\n│    user-auth.2026-01-09             │\n│    payment-flow.2026-01-08          │\n└─────────────────────────────────────┘\n   |\n   v\n┌─────────────────────────────────────┐\n│  Select action:                     │\n│  > Validate                         │\n│    Compare with another artifact    │\n│    View status summary              │\n└─────────────────────────────────────┘\n   |\n   v\n[Executes selected action]\n```\n\n### Actions Available\n\n| Action | Description | Command Equivalent |\n|--------|-------------|-------------------|\n| Validate | Run validation checks | `cook-validate <file>` |\n| Compare | Diff two artifacts | `cook-diff <a> <b>` |\n| Status | Show artifact summary | Quick view of status, mode, owner |\n\n---\n\n## Dry-Run Mode\n\nUse `--dry-run` to preview the cooking process without producing artifacts.\n\n```\n/cook <feature> --dry-run\n```\n\n### What dry-run does\n\n1. **Checks prerequisites**\n   - Is CLAUDE.md present?\n   - Are project-specific chefs configured?\n   - Which system chefs will be used as fallback?\n\n2. **Shows cooking plan**\n   - Lists all phases that would execute\n   - Shows which chefs will be consulted\n   - Identifies microwave blockers (if applicable)\n\n3. **Validates inputs**\n   - Parses feature description\n   - Checks for instruction file (if specified)\n   - Identifies potential issues early\n\n### Dry-run output\n\n```markdown\n# Dry-Run: /cook preview\n\n## Feature\n<parsed feature description>\n\n## Mode\nwell-done | microwave\n\n## Prerequisites Check\n- CLAUDE.md: found | NOT FOUND (will use defaults)\n- Project chefs: <list> | none (will use system chefs)\n- System chefs available: <list>\n\n## Cooking Plan\n1. Phase 0 - Project Policy & Context\n   - Chef: <project or system>\n2. Step 1 - Read the Order\n3. Step 2 - Ingredient Approval (well-done only)\n   - Chef: product_chef\n4. Step 3 - Presentation Planning (if UI changes)\n   - Chef: ux_chef\n5. Step 4 - Cooking\n   - Chef: engineer_chef, architect_chef\n6. Step 5 - Taste Testing\n   - Chef: qa_chef\n7. Step 6 - Safety Inspection\n   - Chef: security_chef\n8. Step 7 - Recipe Notes (if needed)\n   - Chef: docs_chef\n\n## Microwave Blockers (if --microwave)\n- <blocker topics detected> | none\n\n## Potential Issues\n- <early warnings> | none detected\n\n## Ready to Cook\nyes | no (reason: <why>)\n```\n\n### When to use dry-run\n\n- First time using `/cook` on a project\n- Verifying chef configuration\n- Checking if microwave mode is allowed\n- Understanding what phases will run\n\n---\n\n## Cooking Modes\n\n### well-done (default)\nFull governance cooking. No shortcuts, no raw ingredients.\n\nCooking phases:\n- Product scope check (ingredient approval)\n- UX/Design review (presentation planning)\n- Implementation (cooking)\n- QA review (taste testing)\n- Security review (safety inspection)\n- Documentation (recipe notes)\n\nBlocking allowed: YES\n\n---\n\n### microwave\nSpeed-optimized cooking for low-risk changes.\n\nCooking phases:\n- Implementation (quick heat)\n- QA review (light taste test)\n- Security review (only if API/auth touched)\n\nBlocking allowed: YES (security only)\n\nRules:\n- No scope expansion (no adding ingredients)\n- No architecture changes (no changing the recipe)\n- Should be followed by well-done cooking for verification\n\n---\n\n## Cooking Statuses\n\nEvery feature progresses through these stages:\n\n| Status | Meaning |\n|--------|---------|\n| `raw` | Feature requested, not yet evaluated |\n| `cooking` | /cook in progress, review phases running |\n| `blocked` | Specific blocker identified (requires owner + next step) |\n| `needs-more-cooking` | Rejected, incomplete, or killed (+ reason field) |\n| `well-done` | Approved and ready to implement |\n| `ready-for-merge` | Post QA/Security, ready for merge |\n| `plated` | Shipped to production |\n\n**Note:** `killed` is NOT a separate status. Use `needs-more-cooking` with `reason: killed - <why>`\n\n---\n\n## Microwave Blockers\n\nMicrowave mode is **BLOCKED** for these topics. Use `--well-done` instead:\n\n- **auth / permissions / crypto / network security** - any authentication, authorization, encryption, or security-related changes\n- **schema / migrations / storage** - database schema changes, migrations, storage layer modifications\n- **public API contracts** - any changes to public-facing API signatures or behavior\n- **UI flow changes** - even small changes to user flows or navigation\n- **payments / purchase / paywall** - anything touching billing, payments, or monetization\n\nIf microwave mode is requested for a blocked topic, automatically escalate to well-done.\n\n---\n\n## Definition of Done\n\n### Well-Done Mode Requirements\nReference: `~/.claude/templates/well-done-checklist.md`\n\n**MUST include:**\n- Scope definition (in/out)\n- Risks + mitigations (min 3)\n- Test plan (min 3 test cases)\n- Security checklist (all items addressed)\n- Rollout/rollback plan\n- Pre-mortem (3 failure scenarios)\n- Trade-offs documented\n- Ownership assigned\n\n### Microwave Mode Requirements\nReference: `~/.claude/templates/microwave-checklist.md`\n\n**MUST include:**\n- Problem statement + reproduction steps\n- Minimal fix plan\n- Tests (1-2)\n- \"Why safe\" (1 sentence)\n- Pre-mortem (1 failure scenario)\n\n---\n\n## Stop Rules (Kill Switch)\n\nIn Step 2 (Ingredient Approval), automatically set status to `needs-more-cooking` with `reason: killed` if:\n\n1. **No measurable effect** - feature has no clear, testable outcome\n2. **Risk > value** - implementation risk outweighs user benefit\n3. **No owner** (well-done mode) - no one assigned as Decision Owner\n4. **No testable AC** - acceptance criteria cannot be verified\n\nWhen killed, document the specific reason and stop processing.\n\n---\n\n## MANDATORY FIRST STEP: Create Artifact File\n\n**BEFORE ANY OTHER ACTION**, you MUST create the artifact file with skeleton structure.\n\n### Why Artifact-First?\n\n1. **Progress visibility** - User can see cooking progress in real-time\n2. **Interrupt safety** - Partial results are preserved if execution stops\n3. **Phase enforcement** - Each phase MUST write to artifact before proceeding\n4. **State machine** - Artifact tracks which phases are complete\n\n### Step 0.0 - Create Artifact Skeleton (REQUIRED)\n\n**DO THIS IMMEDIATELY UPON /cook INVOCATION:**\n\n1. Generate artifact filename: `cook/<slug>.<YYYY-MM-DD>.cook.md`\n   - `<slug>` = kebab-case of feature description (max 40 chars)\n   - `<YYYY-MM-DD>` = today's date\n\n2. Create file with this skeleton:\n\n```markdown\n# Cooking Result\n\n## Dish\n<feature description>\n\n## Status\nraw\n\n## Cooking Mode\n<well-done | microwave>\n\n## Current Phase\nPhase 0 - Starting...\n\n---\n\n## Phase 0 - Project Policy & Context\n_Pending..._\n\n## Step 1 - Read the Order\n_Pending..._\n\n## Step 2 - Ingredient Approval\n_Pending..._\n\n## Step 3 - Presentation Planning\n_Pending..._\n\n## Step 4 - Cooking\n_Pending..._\n\n## Step 5 - Taste Testing (QA)\n_Pending..._\n\n## Step 6 - Safety Inspection (Security)\n_Pending..._\n\n## Step 7 - Recipe Notes\n_Pending..._\n\n---\n\n## Decision Log\n| Date | Phase | Decision | Rationale |\n|------|-------|----------|-----------|\n```\n\n3. **CONFIRM** artifact file exists before proceeding\n\n**CRITICAL:** Do NOT proceed to Phase 0 until artifact file is created and confirmed.\n\n---\n\n## Cooking Steps\n\n### Phase 0 - Project Policy & Context (REQUIRED)\n\nThis phase runs BEFORE scope, UX, or implementation planning. No code, no design, no solutions are allowed in this phase. Project rules override user intent.\n\n#### Step 0.1 - Discover Project Context Files\n\nSearch for and read the following files (do not fail if missing):\n\n**Priority order:**\n1. `CLAUDE.md` - project rules and constraints\n2. `POLICY.md` - explicit policies\n3. `ENGINEERING.md` - engineering standards\n4. `README.md` - project overview\n5. `docs/**/*.md` - architecture, ADRs, decisions\n6. `.claude/agents/*.md` - project-specific chefs\n\n**Chef Resolution Order:**\n1. Project-specific chefs in `<project>/.claude/agents/`\n2. System-wide chefs in `~/.claude/agents/`\n\n**System-Wide Chefs Available:**\n- `engineer.md` - Head chef (implementation)\n- `product.md` - Menu curator (scope decisions)\n- `designer.md` - Presentation specialist (UX/flow)\n- `security.md` - Health inspector (security audit)\n- `qa.md` - Taste tester (quality assurance)\n- `architect.md` - Kitchen designer (architecture)\n- `docs.md` - Recipe writer (documentation)\n\n#### Step 0.2 - Extract and Normalize Rules\n\nFrom discovered files, extract and classify rules into:\n\n**A) Hard rules (MUST / MUST NOT)**\n- Non-negotiable constraints\n- Security, architecture, legal, platform limitations\n- Explicit \"do not\" statements\n\n**B) Preferred patterns**\n- Recommended libraries, architectures, conventions\n- Style or process preferences\n- Defaults the project expects\n\n**C) Explicit non-goals / forbidden approaches**\n- Deprecated patterns\n- Known bad ideas\n- Things intentionally avoided\n\n**D) Implicit assumptions (derived)**\n- Assumptions inferred due to missing or unclear documentation\n- MUST be clearly marked as assumptions\n\n#### Step 0.3 - Detect Conflicts\n\nIf user request conflicts with extracted rules:\n- Do NOT resolve it yet\n- Do NOT propose alternatives\n- Record the conflict clearly\n\n#### Step 0.4 - Risk Classification\n\nBased on documentation completeness, classify alignment risk:\n- **LOW** - clear policies found\n- **MEDIUM** - partial policies\n- **HIGH** - no meaningful policies found\n\n#### Step 0.5 - Output Format (MANDATORY)\n\nProduce this section in the cook artifact:\n\n```markdown\n## Phase 0 - Project Policy & Context\n\n### Sources scanned\n- <file or \"not found\">\n\n### Hard rules (must not be violated)\n- <rule>\n\n### Preferred patterns\n- <pattern>\n\n### Explicit non-goals / forbidden approaches\n- <non-goal>\n\n### Assumptions due to missing documentation\n- <assumption>\n\n### Detected conflicts with request\n- None OR <conflict description>\n\n### Policy alignment risk\n- LOW | MEDIUM | HIGH\n```\n\n#### Step 0.6 - Blocking Rule\n\nIf ANY of these conditions are true:\n- A hard rule directly blocks the requested feature\n- Alignment risk is HIGH\n\nThen:\n1. Set status: `needs-more-cooking`\n2. Document the blocking reason\n3. STOP - do not proceed to Step 1\n\nThe issue must be acknowledged before continuing.\n\nThis Phase 0 output informs ALL subsequent cooking steps.\n\n#### GATE: Write Phase 0 to Artifact\n\n**STOP. Before proceeding to Step 1, you MUST:**\n\n1. Update artifact file - replace `_Pending..._` under `## Phase 0` with actual output\n2. Update `## Current Phase` to `Phase 0 - Complete`\n3. Update `## Status` to `cooking`\n4. Add entry to Decision Log\n\n**DO NOT proceed until artifact is updated.**\n\n---\n\n### Step 1 - Read the Order\n- Restate feature in concrete terms (what dish are we making?)\n- Identify affected modules/components (which stations are involved?)\n- Identify risks and dependencies (allergens, timing)\n- Note any project-specific constraints from Step 0\n- **Check Microwave Blockers**: If microwave mode requested but touches blocked topics -> escalate to well-done\n\nStatus: `raw` -> `cooking`\n\n#### GATE: Write Step 1 to Artifact\n\n**STOP. Before proceeding to Step 2, you MUST:**\n\n1. Update artifact file - replace `_Pending..._` under `## Step 1` with actual output\n2. Update `## Current Phase` to `Step 1 - Complete`\n3. Add entry to Decision Log\n\n**DO NOT proceed until artifact is updated.**\n\n---\n\n### Step 2 - Ingredient Approval (well-done only)\n- Is this in scope for the project? (Is it on our menu?)\n- Does it add user value? (Will customers order it?)\n- Decision: Approve / Reject / Defer\n\nIf rejected -> Status: `needs-more-cooking`. STOP.\n\n#### GATE: Write Step 2 to Artifact\n\n**STOP. Before proceeding to Step 3, you MUST:**\n\n1. Update artifact file - replace `_Pending..._` under `## Step 2` with actual output (or `_Skipped (microwave mode)_`)\n2. Update `## Current Phase` to `Step 2 - Complete`\n3. Add Product Decision to artifact header section\n4. Add entry to Decision Log\n\n**DO NOT proceed until artifact is updated.**\n\n---\n\n### Step 3 - Presentation Planning (conditional)\nTriggered if:\n- New UI components (new plating style)\n- Changed user flow (changed service sequence)\n- Risk of user confusion (unfamiliar dish)\n\nOutput:\n- Flow description\n- UX considerations\n\n#### GATE: Write Step 3 to Artifact\n\n**STOP. Before proceeding to Step 4, you MUST:**\n\n1. Update artifact file - replace `_Pending..._` under `## Step 3` with actual output (or `_Skipped (no UI changes)_`)\n2. Update `## Current Phase` to `Step 3 - Complete`\n3. Add entry to Decision Log\n\n**DO NOT proceed until artifact is updated.**\n\n---\n\n### Step 4 - Implementation Plan (NO CODE YET)\n- Implement only approved scope (follow the recipe)\n- Respect existing patterns and architecture (house style)\n- Follow project tech stack and conventions (use the right tools)\n- Reference all changed files (ingredient list)\n\n#### GATE: Write Step 4 to Artifact\n\n**STOP. Before proceeding to Step 5, you MUST:**\n\n1. Update artifact file - replace `_Pending..._` under `## Step 4` with Patch Plan\n2. Update `## Current Phase` to `Step 4 - Complete`\n3. Add Trade-offs section to artifact\n4. Add entry to Decision Log\n\n**DO NOT proceed until artifact is updated.**\n\n---\n\n### Step 5 - Taste Testing (QA)\n- Validate acceptance criteria (does it taste right?)\n- Identify edge cases (unusual orders)\n- Flag potential regressions (did we break another dish?)\n- Verify tests exist or are added (documented tasting notes)\n\nBlockers must be resolved before proceeding.\nIf blocked -> Status: `needs-more-cooking`\n\n#### GATE: Write Step 5 to Artifact\n\n**STOP. Before proceeding to Step 6, you MUST:**\n\n1. Update artifact file - replace `_Pending..._` under `## Step 5` with QA Status\n2. Update `## Current Phase` to `Step 5 - Complete`\n3. Add Pre-mortem section to artifact (min 3 scenarios for well-done, 1 for microwave)\n4. Add entry to Decision Log\n\n**DO NOT proceed until artifact is updated.**\n\n---\n\n### Step 6 - Safety Inspection (Security)\n- Input validation (ingredient safety)\n- Authentication/authorization (who can order this?)\n- Data exposure risks (customer privacy)\n- Injection vulnerabilities (contamination)\n\nSecurity blockers override everything.\nIf blocked -> Status: `needs-more-cooking`\n\n#### GATE: Write Step 6 to Artifact\n\n**STOP. Before proceeding to Step 7, you MUST:**\n\n1. Update artifact file - replace `_Pending..._` under `## Step 6` with Security Status\n2. Update `## Current Phase` to `Step 6 - Complete`\n3. Add Blast Radius & Rollout section to artifact\n4. Add entry to Decision Log\n\n**DO NOT proceed until artifact is updated.**\n\n---\n\n### Step 7 - Recipe Notes (conditional)\nTriggered if:\n- Assumptions were made\n- Behavior changed\n- New concepts introduced\n\n#### GATE: Write Step 7 to Artifact (FINAL)\n\n**STOP. Before marking cooking complete, you MUST:**\n\n1. Update artifact file - replace `_Pending..._` under `## Step 7` with docs updates (or `_Skipped (no docs needed)_`)\n2. Update `## Current Phase` to `Cooking Complete`\n3. Update `## Status` to `well-done` or `ready-for-merge`\n4. Add Ownership section to artifact header\n5. Add Assumptions & Notes section\n6. Add Next Actions section\n7. Add final entry to Decision Log\n\n**Artifact is now complete. Proceed to validation.**\n\n---\n\n## Phase Rollback Rules\n\nCooking is NOT strictly linear. These rollback rules apply:\n\n```\n+--------------------------------------------------------------+\n|                        ROLLBACK FLOW                          |\n+--------------------------------------------------------------+\n|                                                               |\n|  Step 2 (Scope) <-------------------------------------+      |\n|       |                                               |      |\n|       v                                               |      |\n|  Step 3 (UX) <------------------------------------+   |      |\n|       |                                           |   |      |\n|       v                                           |   |      |\n|  Step 4 (Implementation) <--------------------+   |   |      |\n|       |                                       |   |   |      |\n|       v                                       |   |   |      |\n|  Step 5 (QA) --------- blocker ------------->|   |   |      |\n|       |                                           |   |      |\n|       v                                           |   |      |\n|  Step 6 (Security) -- blocker ------------------>|   |      |\n|       |                     |                         |      |\n|       |                     +-- scope change -------->+      |\n|       v                                                      |\n|  Step 7 (Docs)                                               |\n|                                                              |\n+--------------------------------------------------------------+\n```\n\n### Rollback Triggers:\n\n| From | To | Trigger |\n|------|----|---------|\n| Step 5 (QA) | Step 4 (Implementation) | Test failure, missing edge case |\n| Step 5 (QA) | Step 2 (Scope) | Scope creep discovered |\n| Step 6 (Security) | Step 4 (Implementation) | Security vulnerability found |\n| Step 6 (Security) | Step 2 (Scope) | Fundamental design flaw |\n| Step 3 (UX) | Step 2 (Scope) | UX requirements change scope |\n\nWhen rollback occurs:\n1. Document the reason in Decision Log\n2. Update status to `blocked` with owner and next step\n3. Re-execute from the rollback target step\n\n---\n\n## Output Format (MANDATORY)\n\nTwo formats exist based on cooking mode. Use the appropriate one.\n\n### WELL-DONE MODE OUTPUT\n\n```markdown\n# Cooking Result\n\n## Dish\n<short description>\n\n## Status\nraw | cooking | blocked | needs-more-cooking | well-done | ready-for-merge | plated\n(if killed: needs-more-cooking + reason: killed - <why>)\n\n## Cooking Mode\nwell-done\n\n## Ownership (REQUIRED)\n- Decision Owner: <name/role>\n- Reviewers: <list or \"auto\">\n- Approved by: <name> on <date>\n\n## Product Decision\nApproved / Rejected / Deferred\n- Reason: <why>\n\n## Pre-mortem (REQUIRED - 3 scenarios)\n1. <scenario> -> mitigation: <action>\n2. <scenario> -> mitigation: <action>\n3. <scenario> -> mitigation: <action>\n\n## Trade-offs\n- Sacrificing: <perf/UX/maintainability/time>\n- Reason: <why>\n- Rejected alternatives:\n  - <alternative 1> - rejected because <reason>\n  - <alternative 2> - rejected because <reason>\n\n## Patch Plan\n- Files to modify:\n  1. <file> - <what changes>\n  2. <file> - <what changes>\n- Commit sequence:\n  1. <commit message>\n  2. <commit message>\n- High-risk areas: <list>\n- Tests to run: <list>\n\n## QA Status\n- Tests: <coverage>\n- Edge cases considered: <list>\n- Regressions checked: <list>\n\n## Security Status\n- Reviewed: yes/no\n- Issues found: <list or \"none\">\n- Risk level: low/medium/high\n\n## Blast Radius & Rollout\n- Affected users/modules: <list>\n- Feature flag: yes/no (name: <flag_name>)\n- Rollout strategy: immediate/gradual/canary\n- Rollback steps:\n  1. <step>\n  2. <step>\n\n## Assumptions & Notes\n<list>\n\n## Next Actions\n<list>\n\n## Decision Log\n| Date | Decision | Rationale |\n|------|----------|-----------|\n```\n\n### MICROWAVE MODE OUTPUT\n\n```markdown\n# Cooking Result\n\n## Dish\n<short description>\n\n## Status\nraw | cooking | blocked | needs-more-cooking | well-done | ready-for-merge | plated\n\n## Cooking Mode\nmicrowave\n\n## Problem Statement\n<what's broken + how to reproduce>\n\n## Fix Plan\n<minimal fix description>\n\n## Why Safe\n<1 sentence explaining why this is low risk>\n\n## Pre-mortem (REQUIRED - 1 scenario)\n1. <scenario> -> mitigation: <action>\n\n## Tests\n- <test 1: verifies the fix>\n- <test 2: regression check> (optional)\n\n## Security Status (only if touches auth/API)\n- Reviewed: yes/no\n- Issues found: <list or \"none\">\n\n## Next Actions\n<list>\n```\n\n---\n\n## Constraints\n\n- This skill orchestrates reasoning and decision-making\n- Each cooking phase must complete before the next begins\n- **Step 0 (Mise en Place) is MANDATORY** - cannot cook without prep\n- Project-specific rules from CLAUDE.md override generic behavior\n- Chef resolution: project-specific chefs take precedence over system-wide chefs\n- Security and QA standards from project policies take precedence\n\n## Chef Assignments\n\n**For each cooking phase, use chefs in this priority:**\n\n1. **Project-specific chef** (if exists): `<project>/.claude/agents/<project>-<role>.md`\n2. **System-wide chef** (fallback): `~/.claude/agents/<role>.md`\n\n| Phase | Project Chef | System Chef |\n|-------|-------------|-------------|\n| Cooking | `*-engineer.md` | `engineer.md` |\n| Menu Approval | `*-product.md` | `product.md` |\n| Presentation | `*-designer.md` | `designer.md` |\n| Safety | `*-security.md` | `security.md` |\n| Tasting | `*-qa.md` | `qa.md` |\n| Kitchen Design | `*-architect.md` | `architect.md` |\n| Recipe Notes | `*-docs.md` | `docs.md` |\n\n## Why Project Configuration Is Recommended\n\nWhile system-wide chefs provide generic best practices, project-specific configuration enables:\n- Enforcing **your** kitchen's standards and techniques\n- Validating against your house recipes\n- Using project-specific safety requirements\n- Context-aware reviews tailored to your cuisine\n\n**Recommendation:** Set up `CLAUDE.md` and project-specific chefs for production cooking.\n\n---\n\n## Artifact Storage\n\nCooking results are stored as decision records:\n- Location: `cook/*.cook.md`\n- Contains: scope decisions, review outcomes, final status\n\n---\n\n## Auto-Validation (Final Step)\n\nAfter generating the artifact, **automatically run validation**:\n\n```bash\n./scripts/cook-validate cook/<artifact>.cook.md\n```\n\n### Validation Behavior\n\n1. **If VALID** - Report success and proceed\n2. **If INVALID (errors)** - Show issues and offer to fix them\n3. **If WARNINGS only** - Report warnings but consider artifact ready\n\n### Example Output\n\n```\n[cook-validate] Validating artifact...\n\nValidating: user-auth.2026-01-10.cook.md\nMode: well-done\n\n[PASS] Scope sections present\n[PASS] Pre-mortem (3 scenarios)\n[PASS] Test cases (5 defined)\n[PASS] Ownership assigned\n\nResult: VALID\n\n✓ Artifact ready for implementation\n```\n\n### Skip Validation\n\nUse `--no-validate` to skip auto-validation:\n\n```\n/cook Add feature X --no-validate\n```\n\n### Validation-Only Mode\n\nTo validate an existing artifact without re-cooking:\n\n```\n/cook --validate cook/existing-artifact.cook.md\n```","tags":["feature","development","cook","pjuniszewski","agent-skills","antigravity","claude-code","feature-development","skills"],"capabilities":["skill","source-pjuniszewski","skill-feature-development","topic-agent-skills","topic-antigravity","topic-claude-code","topic-feature-development","topic-skills"],"categories":["cook"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/PJuniszewski/cook/feature-development","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add PJuniszewski/cook","source_repo":"https://github.com/PJuniszewski/cook","install_from":"skills.sh"}},"qualityScore":"0.456","qualityRationale":"deterministic score 0.46 from registry signals: · indexed on github topic:agent-skills · 13 github stars · SKILL.md body (28,727 chars)","verified":false,"liveness":"unknown","lastLivenessCheck":null,"agentReviews":{"count":0,"score_avg":null,"cost_usd_avg":null,"success_rate":null,"latency_p50_ms":null,"narrative_summary":null,"summary_updated_at":null},"enrichmentModel":"deterministic:skill-github:v1","enrichmentVersion":1,"enrichedAt":"2026-04-22T13:03:33.344Z","embedding":null,"createdAt":"2026-04-19T00:41:18.003Z","updatedAt":"2026-04-22T13:03:33.344Z","lastSeenAt":"2026-04-22T13:03:33.344Z","tsv":"'-01':749,755,761 '-08':762 '-09':756 '-10':750 '-2':198,1359 '/.claude/agents':1721,1728,3149,3157 '/.claude/templates/microwave-checklist.md':1347 '/.claude/templates/well-done-checklist.md':1308 '/cook':26,85,662,689,701,739,824,898,1022,1153,1515,3358,3378 '/scripts/cook-validate':3277 '0':229,929,1556,1559,1620,1631,1907,1997,2007,2026,2036,2092,3090 '0.0':216,532,1506 '0.1':1662 '0.2':1766 '0.3':1847 '0.4':1871 '0.5':1894 '0.6':1945 '1':86,197,264,382,408,421,502,507,509,707,833,927,936,1358,1362,1367,1391,1463,1517,1565,1681,1715,1968,1987,2015,2018,2057,2115,2126,2134,2144,2214,2303,2392,2494,2529,2592,2662,2826,2925,2948,2952,2992,3041,3053,3055,3142,3282 '2':91,284,385,411,424,503,510,512,714,853,934,942,1376,1401,1475,1535,1571,1687,1722,1975,2030,2123,2138,2160,2203,2222,2230,2236,2320,2404,2506,2604,2679,2744,2790,2808,2817,2833,2927,2949,2953,2993,3151,3289 '2026':748,754,760 '3':97,313,414,492,504,721,874,940,954,1317,1321,1334,1409,1485,1576,1607,1691,1980,2038,2146,2211,2238,2260,2292,2311,2326,2328,2412,2514,2523,2612,2686,2748,2814,2843,2923,2929,3300,3331 '4':102,338,733,952,965,1422,1496,1581,1695,2043,2246,2300,2342,2381,2400,2410,2420,2532,2620,2698,2752,2779,2799 '40':1530 '5':107,397,963,974,1585,1699,2389,2434,2483,2502,2512,2705,2756,2776,2787,3336 '6':112,446,972,982,1591,1705,2491,2546,2581,2600,2610,2710,2761,2796,2805 '7':117,476,980,990,1597,2589,2634,2651,2670,2715,2768 '8':988 'ac':1425 'accept':95,425,1426,2439 'access':334 'acknowledg':1992 'action':132,160,179,430,436,555,723,736,765,777,778,780,1449,2713,2997,3070 'actual':53,2028,2136,2224,2313 'ad':1117,2463 'add':2044,2147,2181,2239,2247,2329,2413,2421,2515,2533,2613,2621,2699,2706,2711,2716,3359 'adding/modifying':38 'address':1328 'adr':1703 'affect':271,329,515,2072,2981 'agent':172,552 'align':260,1879,1939,1963 'allergen':2082 'allow':177,1034,1079,1108,1652 'altern':346,1865,2938 'anoth':769,2456 'anyth':1278 'api':1252,1260 'api/auth':1105 'appli':2740 'applic':873 'approach':106,344,1819,1926 'appropri':2864 'approv':225,286,292,944,1060,1180,1378,1573,2162,2189,2350,2910,2915,3168 'architect':970 'architect.md':1757,3182,3183 'architectur':103,341,1120,1702,1760,1790,1803,2359 'area':391,392,2957 'artifact':30,68,88,119,146,162,181,191,217,533,557,654,676,686,698,710,720,726,728,743,770,794,800,823,1444,1454,1461,1493,1499,1508,1519,1609,1622,1904,2009,2020,2053,2117,2128,2156,2205,2216,2243,2256,2294,2305,2338,2383,2394,2419,2430,2485,2496,2521,2542,2583,2594,2619,2630,2653,2664,2703,2722,3245,3272,3308,3316,3343,3373 'artifact-first':1460 'assess':111 'assign':1342,1418,3132,3340 'assumpt':310,1830,1832,1845,1927,2640,2707,2994 'assur':1756 'attent':395 'audit':1751 'auth':753,1227 'auth/api':3064 'auth/authz':464 'authent':1233 'authentication/authorization':2554 'author':1234 'auto':673,3265,3356 'auto-valid':672,3264,3355 'automat':1293,1379,3273 'avail':719,779,924,1733 'avoid':1827 'awar':3227 'b':357,1798 'bad':1823 'base':1874,2858 'bash':173,3276 'begin':3088 'behavior':1263,2643,3108,3281 'benefit':1408 'best':3200 'beyond':55 'bill':1280 'blast':513,2614,2978 'block':159,187,1078,1107,1159,1218,1291,1946,1959,1978,2102,2474,2572,2837,2878,3013 'blocker':280,871,999,1161,1214,2095,2467,2568,2758,2763 'boolean':637,668,679 'break':2455 'broken':3033 'build':205 'c':1813 'call':139,151 'cannot':1428,3096 'case':403,413,416,418,420,423,1323,1525,2447,2785,2965,3335 'chang':331,378,481,604,959,1090,1121,1123,1240,1246,1255,1266,1269,2272,2275,2319,2374,2644,2765,2821 'char':1531 'check':281,440,458,787,834,880,907,1029,1058,2093,2968 'checklist':457,1325 'chef':843,847,865,916,921,923,933,949,951,960,962,967,969,971,977,979,985,987,995,997,1027,1711,1712,1719,1726,1732,1736,3109,3114,3121,3131,3138,3146,3155,3161,3163,3197,3241 'class':43 'classif':1873 'classifi':1776,1878 'claude.md':239,541,837,908,1682,3105,3236 'claude/agents':243,1706 'clear':1398,1842,1869,1882 'code':25,36,54,126,128,1646,2346 'codebas':544 'command':175,782 'commit':380,383,386,2950 'compar':727,767,791 'complet':75,1504,1877,2037,2145,2237,2327,2411,2513,2611,2659,2685,2725,3084 'compliant':9 'compon':46,328,330,2268 'con':350 'concept':2646 'concret':2064 'condit':1952,2263,2637 'configur':844,1028,3190,3205 'confirm':1608,1627 'conflict':257,1849,1853,1868,1933 'confus':2281 'consid':347,2966,3307 'consider':335,2288 'constraint':1686,1788,2089,3071 'consult':868 'contain':3257 'contamin':2566 'content':158 'context':232,428,434,932,1562,1634,1665,1910,3226 'context-awar':3225 'continu':1994 'contract':1253 'convent':1804,2367 'cook':70,73,156,194,208,536,562,629,660,712,789,796,819,855,925,966,1009,1040,1048,1054,1066,1085,1091,1133,1136,1152,1170,1209,1386,1469,1521,1542,1547,1582,1628,1903,1974,2002,2042,2111,2198,2479,2577,2658,2684,2732,2860,2872,2877,2882,2896,2899,3007,3012,3017,3026,3081,3097,3135,3164,3244,3247,3255,3278,3313,3377 'cook-diff':795 'cook-valid':788,3312 'cook.md':71,157,713,1522,3256,3279 'cook/existing-artifact.cook.md':3380 'cook/feature.cook.md':664 'correct':81,578 'could':496 'creat':42,87,144,218,534,556,1443,1452,1507,1536,1625 'creep':2793 'criteria':96,426,1427,2440 'critic':21,1614 'crypto':1229 'cuisin':3232 'curat':1740 'current':213,1553,2032,2140,2232,2322,2406,2508,2606,2681 'custom':2185,2562 'd':1828 'data':466,2559 'databas':1244 'date':527,1534,1603,3000 'decis':220,290,319,342,351,525,529,1420,1601,1605,1704,1742,2047,2150,2188,2241,2250,2332,2424,2536,2624,2720,2831,2906,2914,2998,3001,3078,3252,3259 'decision-mak':3077 'default':624,638,669,680,914,1045,1809 'defer':294,2191,2917 'defin':3337 'definit':1299,1312 'depend':277,2081 'deprec':1820 'deriv':1831 'descript':200,593,599,781,879,1528,2286 'design':101,1648,1759,2811,3181 'designer.md':1743,3172,3173 'detect':256,1006,1848,1932 'determin':628 'develop':3,11,19,570 'diff':729,792,797 'direct':1958 'disciplin':582 'discov':1663,1772,2794 'dish':196,584,1544,2067,2283,2457,2874,3009 'doc':996,1700,2672,2677,2769 'docs.md':1761,3186,3187 'document':477,478,484,1075,1340,1433,1764,1839,1876,1931,1976,2464,2827 'done':212,622,627,904,947,1044,1132,1179,1225,1298,1301,1304,1414,1551,2108,2165,2528,2692,2868,2885,2903,3020,3322 'dri':635,745,809,814,826,830,891,896,1017 'dry-run':634,808,813,825,829,890,895,1016 'dry-run-valid':744 'due':1834,1928 'earli':889 'edg':412,417,419,422,2446,2784,2964 'effect':1394 'en':3092 'enabl':3206 'encrypt':1235 'enforc':1487,3207 'engin':968,1693 'engineer.md':1734,3165,3166 'engineering.md':1692 'entri':2045,2148,2248,2330,2422,2534,2622,2718 'enum':619 'equival':783 'error':415,3292 'escal':1294,2104 'evalu':1151 'even':1267 'everi':583,1138 'everyth':2570 'exampl':661,688,3310 'except':560 'execut':133,148,647,734,775,862,1483,2846 'exist':163,192,441,653,1611,2356,2460,2857,3148,3372 'expans':1115 'expect':1812 'explain':3043 'explicit':1689,1794,1814,1921 'explor':543 'exposur':467,2560 'extra':394 'extract':1767,1774,1855 'face':1259 'fail':1676 'failur':1335,1368,2782 'fallback':852,3156 'fals':639,670,681 'featur':2,10,18,268,442,518,563,592,602,878,900,1139,1147,1395,1527,1962,2062,2983,3360 'feature-develop':1,17 'field':1176 'file':31,40,56,69,89,147,154,166,182,235,374,377,480,558,608,615,883,1445,1455,1537,1610,1623,1666,1673,1773,2021,2129,2217,2306,2375,2395,2497,2595,2665,2945 'filenam':1520 'final':118,2654,2717,3262,3267 'first':90,131,137,178,542,545,549,553,554,1019,1441,1462 'fix':1355,3037,3298 'flag':519,2450,2984 'flaw':2812 'flow':12,325,537,571,738,759,1265,1272,2274,2285,2742 'follow':1128,1672,2352,2362 'forbidden':1818,1925 'format':1896,2853,2856 'found':473,909,911,1884,1892,2803,2974,3068 'full':6,1046 'function':44 'fundament':2810 'gain':368 'gate':2004,2112,2200,2289,2378,2480,2578,2648 'generat':677,1518,3270 'generic':3107,3199 'get':371 'give':366 'given':405,427,433 'glob/grep':167 'go':497 'goal':305,573,1817,1924 'govern':1047 'happen':645 'happi':409 'hard':246,1780,1913,1956 'head':1735 'header':2244,2704 'health':1748 'heat':1095 'high':389,1888,1943,1966,2955 'high-risk':388,2954 'hook':185 'hous':2360,3217 'idea':1824 'identifi':869,886,1162,2071,2078,2445 'immedi':130,152,1513 'immediate/gradual/canary':2989 'immediate/gradual/canary_':524 'impact':274,500 'implement':34,49,60,80,124,339,1065,1093,1184,1404,1643,1737,2343,2348,2753,2780,2800,3346 'implicit':1829 'in/out':1313 'includ':123,1310,1349 'incomplet':1172 'infer':1833 'inform':1999 'ingredi':285,943,1053,1059,1118,1377,1572,2161,2376,2552 'inject':469,2564 'input':461,591,876,2550 'inspect':984,1074,1593,2548 'inspector':1749 'instead':1226 'instruct':607,882 'intent':1660,1826 'interact':100,678,683,690,691,694,702,704,737,740 'interrupt':1476 'introduc':2647 'invalid':3291 'invoc':1516 'involv':2077 'issu':472,888,1004,1989,2973,3067,3294 'item':1327 'kebab':1524 'kebab-cas':1523 'key':237 'kill':1174,1199,1212,1372,1389,1432,2892,2898 'kitchen':1758,3180,3209 'known':1822 'languag':598 'launch':682,696 'layer':1249 'legal':1791 'level':276,454,2976 'librari':1802 'light':1098 'likelihood':499 'limit':1793 'linear':2736 'list':64,83,517,857,2377 'locat':3254 'log':526,1602,2048,2151,2251,2333,2425,2537,2625,2721,2832,2999 'low':1088,1881,1941,3047 'low-risk':1087 'low/medium/high':2977 'low/medium/high_':455 'machin':1498 'made':2642 'make':2070,3079 'manag':488,687,699 'mandatori':1440,1897,2854,3095 'mark':1843,2657 'markdown':193,894,1541,1905,2871,3006 'max':1529 'md':244,1701,1707,3150,3158 'mean':1145 'meaning':1890 'measur':1393 'medium':1885,1942 'menu':684,700,724,1739,2178,3167 'merg':1188,1193,2697,2889,3024 'messag':384,387 'microwav':279,623,870,905,998,1001,1031,1081,1213,1215,1285,1343,1552,2094,2097,2228,2531,3003,3028 'migrat':1242,1247 'min':1316,1320,2522 'minim':1354 'mise':3091 'miss':1678,1836,1930,2783 'mitig':501,1315,2926,2928,2930,3056 'mode':209,618,641,692,705,806,811,901,1032,1041,1216,1286,1305,1344,1415,1548,2098,2229,2861,2869,2900,3004,3027,3319,3368 'modif':1250 'modifi':376,2947 'modul':273 'modules/components':272,2073 'monet':1283 'mortem':491,1333,1366,2518,2921,3051,3330 'multi':568 'multi-phas':567 'must':140,248,585,1309,1348,1451,1490,1782,1783,1840,1915,1990,2017,2125,2213,2302,2391,2468,2493,2591,2661,3083 'name':521,2986 'name/role':2908 'navig':1274 'need':393,482,485,994,1168,1207,1384,1972,2196,2477,2575,2678,2880,2894,3015 'needs-more-cook':1167,1206,1383,1971,2195,2476,2574,2879,2893,3014 'negoti':1787 'network':1230 'new':483,2266,2269,2645 'next':1165,2712,2841,2996,3069,3087 'no-valid':665,3350,3362 'non':304,1786,1816,1923 'non-goal':303,1815,1922 'non-negoti':1785 'none':917,1002,1005,1936 'normal':1769 'note':333,460,992,1077,1198,1599,2084,2466,2636,2708,2995,3185 'occur':2825 'off':362,1339,2416,2933 'offer':3296 'one':1417,2865 'optim':1084 'option':348,352,356,610,650,3058 'orchestr':3074 'order':267,606,939,1568,1680,1714,2060,2186,2449,2557 'outcom':1400,3261 'output':893,1895,1998,2029,2137,2225,2284,2314,2852,2870,3005,3311 'outweigh':1406 'overrid':1658,2569,3106 'overview':1698 'owner':221,807,1164,1411,1421,2839,2907 'ownership':219,1341,2700,2904,3339 'pars':877 'partial':1478,1886 'pass':3323,3327,3333,3338 'patch':372,2402,2943 'path':57,155,410,651 'pattern':254,1800,1821,1920,2357 'payment':758,1275,1281 'payment-flow':757 'paywal':1277 'pend':240,242,245,252,255,258,262,270,278,282,291,298,302,306,309,311,320,326,336,345,463,465,468,471,474,486,1563,1569,1574,1579,1583,1589,1595,1600,2023,2131,2219,2308,2397,2499,2597,2667 'perf/ux/maintainability/time':2935 'permiss':1228 'phase':16,214,228,528,569,633,859,928,1037,1055,1092,1157,1486,1489,1502,1554,1555,1558,1604,1619,1630,1637,1655,1906,1996,2006,2025,2033,2035,2141,2233,2323,2407,2509,2607,2682,2729,3082,3136,3159 'picker':717 'place':3093 'plain':597 'plain-languag':596 'plan':22,116,315,340,373,401,506,856,926,956,1064,1319,1330,1356,1578,1644,2262,2344,2403,2944,3038 'plate':1194,2270,2890,3025 'platform':1792 'polici':8,231,259,931,1561,1633,1690,1883,1887,1891,1909,1938,3128 'policy-compli':7 'policy.md':1688 'post':1189 'potenti':887,1003,2451 'practic':3201 'pre':490,1332,1365,2517,2920,3050,3329 'pre-mortem':489,1331,1364,2516,2919,3049,3328 'preced':3116,3130 'prefer':253,1799,1808,1919 'prep':3099 'prepar':588 'prerequisit':835,906 'present':314,715,838,955,1063,1577,1744,2261,3171,3326 'preserv':1481 'pretoolus':184 'preview':640,817,899 'prioriti':1679,3141 'privaci':2563 'problem':1350,3029 'proceed':1495,1613,1617,1984,2012,2051,2120,2154,2208,2254,2297,2336,2386,2428,2472,2488,2540,2586,2628,2726,3288 'process':820,1439,1807 'produc':27,66,822,1898 'product':92,287,289,581,950,1056,1197,2240,2913,3243 'product.md':1738,3169,3170 'progress':1140,1155,1464,1470 'project':230,841,915,930,1025,1560,1632,1656,1664,1683,1697,1709,1717,1811,1908,2087,2173,2363,3101,3112,3127,3144,3160,3189,3203,3221,3239 'project-specif':840,1708,1716,2086,3100,3111,3143,3202,3220,3238 'proper':587 'propos':1864 'pros':349 'provid':3198 'public':1251,1258 'public-fac':1257 'purchas':1276 'purpos':561 'qa':113,398,978,1067,1096,1588,2437,2504,2757,2777,2788,2961,3124 'qa.md':1752,3178,3179 'qa/security':1190 'qualiti':1755 'quick':802,1094 'radius':514,2615,2979 'rational':530,1606,3002 'raw':207,1052,1146,1546,2110,2876,3011 're':204,659,2845,3376 're-cook':658,3375 're-execut':2844 'read':164,265,540,937,1566,1670,2058 'readi':1007,1182,1186,1191,2695,2887,3022,3309,3344 'readme.md':241,1696 'ready-for-merg':1185,2694,2886,3021 'real':1473 'real-tim':1472 'reason':355,359,1012,1175,1211,1388,1436,1979,2829,2897,2918,2936,3075 'recip':617,991,1076,1125,1598,1762,2354,2635,3184,3218 'recommend':1801,3192,3233 'record':1866,3253 'refer':58,1307,1346,2372 'regress':439,2452,2967 'reject':293,354,1171,2190,2193,2916,2937,2939,2941 'relat':611,1239 'replac':2022,2130,2218,2307,2396,2498,2596,2666 'report':3285,3304 'reproduc':3036 'reproduct':1352 'request':79,1148,1288,1852,1935,1961,2099 'requir':321,323,494,595,614,1163,1306,1345,1510,1635,2820,2905,2922,3052,3224 'resolut':1713,3110 'resolv':1859,2470 'respect':2355 'restat':2061 'result':195,432,438,1479,1543,2873,3008,3248,3341 'review':15,93,99,104,109,114,223,288,317,399,448,451,632,1062,1068,1072,1097,1102,1156,2909,2971,3065,3228,3260 'right':2370,2444 'risk':110,261,275,379,390,453,487,1089,1314,1402,1405,1872,1880,1940,1964,2079,2278,2561,2956,2975,3048 'rollback':505,2730,2738,2741,2770,2824,2849,2990 'rollout':522,2616,2980,2987 'rollout/rollback':1329 'rule':238,247,1112,1371,1657,1684,1770,1777,1781,1856,1914,1947,1957,2731,2739,3103 'run':4,48,546,636,746,785,810,815,827,831,892,897,1018,1039,1158,1638,2960,3274 'sacrif':363,2934 'safe':1361,3040 'safeti':579,983,1073,1477,1592,2547,2553,3174,3223 'scan':234,708,1912 'scenario':404,493,1336,1369,2524,2924,3054,3332 'schema':1241,1245 'scope':94,295,297,301,1057,1114,1311,1640,1741,2170,2351,2745,2764,2791,2792,2809,2818,2822,3258,3324 'search':169,548,1667 'section':1900,2245,2417,2519,2617,2701,2709,2714,3325 'secur':108,447,449,456,986,1071,1101,1110,1231,1238,1324,1594,1750,1789,2549,2567,2602,2762,2797,2801,2806,2969,3059,3122 'security-rel':1237 'security.md':1747,3175,3176 'see':1468 'select':343,358,735,742,764,776 'sentenc':199,1363,3042 'separ':78,1203 'sequenc':381,2277,2951 'serv':590 'servic':2276 'set':1380,1969,3234 'ship':1195 'shortcut':1050 'show':642,722,799,854,863,3293 'signatur':1261 'skeleton':1457,1509,1540 'skill':20,3073 'skill-feature-development' 'skip':671,2227,2316,2675,3347,3354 'small':1268 'solut':1650 'sourc':39,233,1911 'source-pjuniszewski' 'specialist':1745 'specif':612,842,1160,1435,1710,1718,2088,3102,3113,3145,3204,3222,3240 'specifi':885 'speed':576,1083 'speed-optim':1082 'stack':2365 'stage':1143 'standard':1694,3125,3211 'start':535,1557 'state':1497 'statement':1351,1797,3030 'station':2075 'status':206,236,450,459,731,772,798,805,1137,1144,1204,1381,1545,1970,2040,2109,2194,2475,2505,2573,2603,2688,2835,2875,2962,2970,3010,3060,3263 'step':61,129,215,263,283,312,337,396,445,475,508,511,531,935,941,953,964,973,981,989,1166,1353,1375,1442,1505,1564,1570,1575,1580,1584,1590,1596,1629,1661,1765,1846,1870,1893,1944,1986,2003,2014,2056,2091,2114,2122,2133,2143,2159,2202,2210,2221,2235,2259,2291,2299,2310,2325,2341,2380,2388,2399,2409,2433,2482,2490,2501,2511,2545,2580,2588,2599,2609,2633,2650,2669,2743,2747,2751,2755,2760,2767,2775,2778,2786,2789,2795,2798,2804,2807,2813,2816,2842,2851,2991,3089,3268 'stop':1370,1438,1484,1981,2010,2118,2199,2206,2295,2384,2486,2584,2655 'storag':1243,1248,3246 'store':3250 'strategi':523,2988 'strict':2735 'string':594,609,649 'structur':14,566,1458 'style':1805,2271,2361 'subsequ':2001 'success':3286 'summari':269,732,773,801 'switch':1373 'system':846,920,922,1724,1730,3119,3153,3162,3195 'system-wid':1723,1729,3118,3152,3194 'tailor':3229 'take':3115,3129 'target':2850 'task/explore':170,551 'tast':975,1069,1099,1586,1753,2435,2443,2465,3177 'tbd':222,224,227 'tech':2364 'technic':105 'techniqu':3213 'term':2065 'test':50,115,400,402,976,1070,1100,1318,1322,1357,1587,2436,2459,2781,2958,2963,3057,3334 'testabl':1399,1424 'tester':1754 'thing':1825 'thorough':630 'time':1020,1474,2083 'today':1532 'todo':63,82 'tool':138,189,2371 'topic':1221,1292,2103 'topic-agent-skills' 'topic-antigravity' 'topic-claude-code' 'topic-feature-development' 'topic-skills' 'touch':1106,1279,2101,3063 'track':1500 'trade':361,1338,2415,2932 'trade-off':360,1337,2414,2931 'trigger':2264,2638,2771,2774 'true':1954 'two':793,2855 'type':332 'ui':327,958,1264,2267,2318 'unclear':1838 'understand':1035 'unfamiliar':2282 'unusu':2448 'updat':479,2019,2031,2039,2055,2127,2139,2158,2215,2231,2258,2304,2321,2340,2393,2405,2432,2495,2507,2544,2593,2605,2632,2663,2673,2680,2687,2834 'upon':1514 'use':550,693,812,850,913,919,1015,1021,1205,1222,2368,2862,3137,3219,3349 'user':76,307,324,752,1271,1407,1466,1659,1851,2182,2273,2280 'user-auth':751 'user-auth.2026-01-10.cook.md':3318 'users/modules':516,2982 'ux':98,316,318,961,1641,2287,2749,2815,2819 'ux/design':1061 'ux/flow':1746 'v':741,763,774,2746,2750,2754,2759,2766 'valid':462,648,656,663,667,674,725,747,766,784,786,790,875,2438,2551,2728,3214,3266,3275,3280,3284,3314,3315,3317,3342,3348,3352,3357,3364,3366,3370,3379 'validation-on':3365 'valu':308,1403,2183 'vector':470 'verif':1135 'verifi':444,1026,1430,2458 'view':730,771,803 'violat':251,1918 'visibl':1465 'vulner':2565,2802 'warn':3302,3305 'well':211,621,626,903,946,1043,1131,1178,1224,1297,1303,1413,1550,2107,2164,2527,2691,2867,2884,2902,3019,3321 'well-don':210,620,625,902,945,1042,1130,1177,1223,1296,1302,1412,1549,2106,2163,2526,2690,2866,2883,2901,3018,3320 'wide':1725,1731,3120,3154,3196 'without':646,657,821,3098,3374 'would':644,861 'write':52,125,142,150,153,180,1491,2005,2113,2201,2290,2379,2481,2579,2649 'writer':1763 'wrong':120,498 'x':3361 'yes':1010,1080,1109 'yes/no':520,2972,2985,3066 'yes/no_':452 'yet':1150,1861,2347","prices":[{"id":"ca7ae6a0-6e84-45b0-9436-465cf21237ae","listingId":"9dfc9864-5cf5-4aed-bc8b-a2596244eadf","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"PJuniszewski","category":"cook","install_from":"skills.sh"},"createdAt":"2026-04-19T00:41:18.003Z"}],"sources":[{"listingId":"9dfc9864-5cf5-4aed-bc8b-a2596244eadf","source":"github","sourceId":"PJuniszewski/cook/feature-development","sourceUrl":"https://github.com/PJuniszewski/cook/tree/main/skills/feature-development","isPrimary":false,"firstSeenAt":"2026-04-19T00:41:18.003Z","lastSeenAt":"2026-04-22T13:03:33.344Z"}],"details":{"listingId":"9dfc9864-5cf5-4aed-bc8b-a2596244eadf","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"PJuniszewski","slug":"feature-development","github":{"repo":"PJuniszewski/cook","stars":13,"topics":["agent-skills","antigravity","claude-code","feature-development","skills"],"license":"mit","html_url":"https://github.com/PJuniszewski/cook","pushed_at":"2026-02-07T08:41:12Z","description":null,"skill_md_sha":"ed72dbd63edbf45c4f6c5e009da7c7b30ec2908a","skill_md_path":"skills/feature-development/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/PJuniszewski/cook/tree/main/skills/feature-development"},"layout":"multi","source":"github","category":"cook","frontmatter":{"name":"feature-development","description":"Run a full, policy-compliant feature development flow with structured review phases"},"skills_sh_url":"https://skills.sh/PJuniszewski/cook/feature-development"},"updatedAt":"2026-04-22T13:03:33.344Z"}}