Comprehensive feature development with exploration, architecture, TDD, and quality gates. Use when user mentions 機能開発, 新機能, 機能追加, feature development. Do NOT use for implementation without planning (use /code), bug fixes (use /fix), or planning only without implementation (use /think).
From shipnpx claudepluginhub thkt/dotclaude --plugin toolkitThis skill is limited to using the following tools:
Provides UI/UX resources: 50+ styles, color palettes, font pairings, guidelines, charts for web/mobile across React, Next.js, Vue, Svelte, Tailwind, React Native, Flutter. Aids planning, building, reviewing interfaces.
Fetches up-to-date documentation from Context7 for libraries and frameworks like React, Next.js, Prisma. Use for setup questions, API references, and code examples.
Calculates TAM/SAM/SOM using top-down, bottom-up, and value theory methodologies for market sizing, revenue estimation, and startup validation.
Chain /think → /code → /audit for end-to-end feature development.
| Plugin | Required for | Install |
|---|---|---|
| agent-browser | Phase 4.5 only | claude plugin add agent-browser |
Phase 4.5 is skipped gracefully if agent-browser is not installed.
$1 (optional)Detect project type → present relevant options:
| Pattern | Detection | Options |
|---|---|---|
| Claude Code config | ~/.claude/ or .claude/ with skills/hooks/ | Add skill, Add hook, Add agent |
| React/Next.js | package.json has react, next | Add component, Add page, Add API route |
| API server | Express, Fastify, Hono, or src/routes/ | Add endpoint, Add middleware, Add service |
| CLI tool | bin in package.json or src/cli/ | Add command, Add option, Add subcommand |
| Library | main/exports in package.json | Add function, Add class, Add type |
| Fallback | No match | New feature, Extension, Refactoring |
[@../skills/lib/sow-resolution.md]
| Phase | Name | Action | User Checkpoint |
|---|---|---|---|
| 1 | Discovery | Context scan → PRE_TASK_CHECK | [?] or [→] resolution |
| 2 | Design | Skill: /think | Design approval |
| 3 | Implementation | Skill: /code | — |
| 4 | Quality | /audit → /fix loop (max 3) | Remaining issues only |
| 4.5 | Visual Verification | Browser check (UI tasks only) | Visual approval |
| 5 | Summary | AC coverage + scope report | Completion |
$1 empty → AskUserQuestion with context-aware options/code (skip Phases 2-5)Execute Skill("think", $1).
Output: .claude/workspace/planning/YYYY-MM-DD-[feature]/sow.md + spec.md
Execute Skill("code", $1).
/code auto-discovers SOW from Phase 2 output.
| Step | Action | Exit |
|---|---|---|
| 1 | Skill: /audit (changed files from git) | 0 critical/high → Step 4 |
| 2 | Skill: /fix for each critical/high | → Step 3 |
| 3 | Increment iteration (max 3) → Step 1 | Max reached → Step 5 |
| 4 | Skill: /polish → verify tests pass | Tests fail → fix (max 2). Still failing → Step 5 |
| 5 | Present remaining issues (if any) | User decides |
Changed files: git diff main...HEAD --name-only (or base branch).
| # | Check | How | On fail |
|---|---|---|---|
| 1 | UI files in changed files | Match .tsx, .jsx, .css, .scss, .html | skip (silent) |
| 2 | agent-browser installed | which agent-browser | skip (silent) |
| 3 | Dev server detected in package.json | Match dev, start:dev, start scripts | skip (silent) |
From package.json scripts:
| Priority | Script name pattern | Default URL |
|---|---|---|
| 1 | dev, start:dev | http://localhost:5173 |
| 2 | start | http://localhost:3000 |
| 3 | storybook, storybook:dev | http://localhost:6006 |
Extract port from script value if specified (--port, -p, PORT=).
agent-browser --headed open {url} → navigate to page matching SOW scope
(infer route from changed file paths or AC descriptions)agent-browser screenshot → capture current stateType: e2e scenarios:
Agent(subagent_type: "e2e-test-generator",
prompt: "spec_path: <path>\ndev_server_url: <url>",
run_in_background: true)
agent-browser closePresent summary. AC coverage is already verified by /code Quality Gates.
Detect resume point from existing artifacts:
| Artifact | Resume |
|---|---|
| No SOW | Phase 1 |
SOW draft | Phase 2 |
SOW in-progress + no implementation | Phase 3 |
| Implementation done + quality not completed | Phase 4 |
| Quality passed + UI files changed | Phase 4.5 |
| Quality passed + no UI files | Phase 5 |
Implementation evidence: git diff main...HEAD --name-only shows files matching
SOW scope.
| Error | Action |
|---|---|
| /think cancelled or fails | Save context, exit |
| /code fails | Present error, ask user |
| Quality loop exhausted (3 rounds) | Present remaining, user decides |
| agent-browser not installed | Skip Phase 4.5, continue to Phase 5 |
| Dev server not running | Skip Phase 4.5, continue to Phase 5 |
| /code Quality Gates shows unmet ACs | Offer to re-enter Phase 3 or 4 |
| Check | Required |
|---|---|
| PRE_TASK_CHECK passed? | Yes |
| SOW + Spec generated? | Yes |
| All tests pass? | Yes |
| /code AC coverage? | Yes |