claude-code-skills

PR review severity を Haiku AI で再分類。regex 偽陽性を除外する片方向ゲート。/classify-review [PR番号] で実行。regex が CRITICAL/HIGH 検出した際の偽陽性チェック専用。

Codexにセカンドオピニオンを求める。AI同士の忖度なしガチレビュー。Use when user mentions 'Codex レビュー', 'セカンドオピニオン', 'Codex の意見', 'Codex でレビュー'. Do NOT load for: 'Codex に実装させて', 'Codex Worker', 'Codex に作らせて', '実装を依頼'.

Prevent GCP deployment failures by validating Docker builds, Cloud Run deploys, and GKE rollouts. Triggers on: docker build, docker push, gcloud run deploy, kubectl apply, Dockerfile/cloudbuild.yaml/nginx.conf edits. Catches arm64-amd64 platform mismatch, Mixed Content from http:// build args, VAD/WASM asset 404s, nginx MIME breakage. Use for all Apple Silicon to GCP deploy workflows. Do NOT use for local-only Docker builds, non-GCP deployments, or application logic changes.

Manage Context7 CLI skills: search, install, list, remove, info via ctx7. Use when the user wants to find, install, list, remove, or inspect Context7 skills. Triggers: 'search skills', 'install skill', 'list my skills', 'remove skill', 'ctx7', 'context7'. Do NOT use for general package management (npm, pip), non-Context7 CLI commands, or skill authoring.

Plan and execute multi-agent team compositions for parallel implementation tasks. Provides team patterns, wave-based execution, cross-review protocols, and TeamCreate vs parallel Agent decision matrix. Use when: planning team composition for complex multi-file features, choosing orchestration strategy for 3+ parallel tasks, setting up wave-based execution with dependencies, coordinating cross-review between agents. Do NOT use for: single-agent tasks, basic parallel execution of 2 independent tasks (rules handle that), Codex CLI delegation, or sequential single-file changes.

Investigate, fix, and regression-proof bugs across the codebase. Invoke with: /bugfix [bug description]. Performs root cause analysis with full codebase grep, applies fixes to ALL instances (not just the first match), adds regression tests, and generates a structured report. Uses web search for unknown error patterns and library-specific issues. Use when: user reports a bug, error log needs investigation, fixing a GitHub Issue labeled 'bug'. Do NOT use for: new feature implementation, refactoring without a bug, performance optimization, or UI styling changes.

Review code for quality, security, and best practices across TypeScript, JavaScript, Python, Go, Swift, and Kotlin. Analyzes PRs with automated scripts (pr_analyzer, code_quality_checker, review_report_generator), checks for OWASP vulnerabilities, validates naming/structure/error handling, and generates actionable review reports. Use when: reviewing a pull request, providing code feedback, checking code quality before merge, auditing security of changes. Do NOT use for: writing new code, fixing bugs (use /bugfix), generating tests, or architectural design (use /modern-architecture).

Design features using DDD + TDD + Clean Architecture + Event-Driven + API-First patterns for TypeScript, Python, and Go. Use when: designing features with domain complexity (Entities, Value Objects, Aggregates), structuring projects with layer separation, defining Bounded Contexts, applying Outside-In TDD, designing CQRS or Event Sourcing, creating API contracts before implementation, starting a new project architecture. Trigger phrases: DDD, TDD, clean architecture, hexagonal, ports and adapters, CQRS, event sourcing, API-first, domain model, bounded context. Do NOT use for: simple scripts, one-off utilities, UI-only components without domain logic, or CRUD-only endpoints with no business rules.

CI green + claude-review LGTM まで自動ループ。push → CI待ち → 全レビューソースチェック → 修正 → 再push を繰り返す

Interactive guide for creating, reviewing, and optimizing Claude skills. Walks users through use case definition, frontmatter generation, instruction writing, validation, and packaging. Use when user says "create a skill", "build a skill", "make a new skill", "design a skill", "optimize my skill", "review this skill", or "improve this skill". Do NOT use for general coding tasks, writing documentation, or non-skill file creation.

Design system toolkit: generate design tokens (colors, typography, spacing), document components, calculate responsive layouts, and produce developer handoff assets. Use when creating a design system, generating tokens from brand colors, ensuring visual consistency, or preparing dev handoff. Triggers: 'generate design tokens', 'create color palette', 'design system setup', 'typography scale'. Do NOT use for UX research, user testing, or backend architecture.

Opinionated UI constraints for Tailwind CSS, accessibility, animation, and layout. Use when building, reviewing, or fixing UI components. Triggers: 'review this component', 'check UI quality', 'build a form/modal/dialog', 'fix animation jank', 'accessible button'. Do NOT use for backend logic, API design, or database work.

UX research and design toolkit: data-driven persona generation, journey mapping, usability testing frameworks, and research synthesis. Use when creating personas, mapping user journeys, planning usability tests, or synthesizing research findings. Triggers: 'create persona', 'map user journey', 'plan usability test', 'synthesize research'. Do NOT use for UI implementation, visual design, or coding tasks.

Implement production-ready agents using Google Agent Development Kit (ADK). Use when creating ADK agents, building multi-agent systems, implementing workflow agents, reviewing ADK code, or deploying to Vertex AI / Cloud Run. Triggers: 'ADK agent', 'Google ADK', 'multi-agent system', 'Gemini agent', 'agent workflow', 'adk deploy'. Do NOT use for non-ADK agent frameworks (LangChain, CrewAI, AutoGen) or general Gemini API calls without ADK.

Explore ideas and design solutions collaboratively before implementation. Use BEFORE building features, components, or making architectural changes. Triggers: 'brainstorm', 'design this feature', 'let's think through', 'explore approaches', 'before we build', 'what should we consider'. Do NOT use for implementation, code review, or bug fixes.

Generates user-friendly changelogs from git commit history. Use when the user says: 'create changelog', 'generate release notes', 'write changelog', 'what changed since last release', 'summarize commits', 'prepare release notes', 'app store update description', 'product update summary'. Scans git history, categorizes changes, and converts technical commits into customer-readable language. Do NOT use for git operations themselves, code review, or generating internal technical documentation.

Analyze Claude Code chat history to identify coding patterns, skill gaps, and growth areas. Generates a personalized report with curated HackerNews learning resources and sends it to Slack DMs. Triggers: 'analyze my growth', 'review my coding patterns', 'developer growth report', 'what should I learn', 'analyze my recent work'. Do NOT use for code review, project analysis, or team performance evaluation.

Organizes files and folders by analyzing content, detecting duplicates, and restructuring directories. Use when the user says: 'organize my files', 'clean up downloads', 'find duplicates', 'restructure folder', 'organize photos', 'desktop cleanup', 'sort files', 'archive old projects', 'folder structure'. Handles Downloads cleanup, duplicate detection, project reorganization, and photo sorting. Do NOT use for code refactoring (use senior-fullstack), git operations, or codebase restructuring within a software project.

Analyze git diff and generate descriptive conventional commit messages. Use when staging changes, writing commit messages, or reviewing staged changes before committing. Triggers: 'commit this', 'write commit message', 'what should the commit say', 'review staged changes', 'help me commit'. Do NOT use for git branching strategy, merge conflict resolution, or PR creation.

Google Workspace CLI operations via gws (Drive/Sheets/Docs/Slides/Calendar/Gmail). Use when a user asks to search, list, read, export, upload, or update files on Google Drive; read/write Google Sheets or Docs; manage Calendar events; send/search Gmail; or troubleshoot gws auth/access errors. Triggers: 'Google Drive', 'Sheets', 'Docs', 'Slides', 'gws', 'spreadsheet', 'ドライブ', 'スプレッドシート'. Do NOT use for: MCP-native Gmail/Calendar operations (prefer mcp__claude_ai_Gmail/Google_Calendar when available), local file operations, or non-Google cloud storage.

Run a security review against code changes or files. Covers OWASP Top 10: secrets management, input validation, SQL injection, XSS, CSRF, auth/authz, rate limiting, and sensitive data exposure. Use when adding auth, handling user input, creating API endpoints, working with secrets, implementing payments, or integrating third-party APIs. Triggers: 'security review', 'check for vulnerabilities', 'audit this endpoint', 'is this safe'. Do NOT use for general code review, performance optimization, or UI/UX feedback.

Design scalable, maintainable system architectures with technology stack decisions, dependency analysis, and architecture diagrams. Use for system design, architecture decisions, trade-off evaluation, integration patterns, and technical decision frameworks. Triggers: 'design the system', 'architecture for', 'how should we structure', 'tech stack decision', 'dependency analysis', 'draw architecture diagram', 'evaluate trade-offs'. Do NOT use for implementation details, frontend component creation, or code-level refactoring.

Guides backend system development with Node.js, Express, Go, Python, PostgreSQL, GraphQL, and REST APIs. Use when the user says: 'design an API', 'optimize database queries', 'implement authentication', 'set up backend', 'create API endpoint', 'database migration', 'fix N+1 query', 'add rate limiting', 'backend code review'. Covers API scaffolding, database optimization, security implementation, and performance tuning. Do NOT use for frontend/UI work (use senior-fullstack), test writing (use tdd-workflow), or changelog generation.

Build high-performance web applications with React, Next.js, TypeScript, and Tailwind CSS. Use for frontend feature development, component creation, performance optimization, bundle analysis, UI/UX implementation, state management, and frontend code review. Triggers: 'create component', 'optimize performance', 'reduce bundle size', 'fix UI', 'frontend review', 'React pattern', 'Next.js optimization'. Do NOT use for backend-only tasks, database design, or infrastructure/DevOps work.

Guides fullstack web application development with React, Next.js, Node.js, GraphQL, and PostgreSQL. Use when the user says: 'scaffold a new project', 'set up fullstack app', 'analyze code quality', 'review architecture', 'create project structure', 'implement design pattern', 'fullstack development help'. Covers project scaffolding, code quality analysis, architecture patterns, and tech stack guidance. Do NOT use for backend-only tasks (use senior-backend), test writing (use tdd-workflow), or file organization (use file-organizer).

Diagnose and fix Supabase + Next.js App Router bugs on Vercel serverless. Use when: "Vercel 404/500", "Supabase RLS error", "pre-commit not found", "serverless state lost", "hydration error #418", "API works locally but not on Vercel", "env vars not working on Vercel", "NEXT_PUBLIC_ not available at runtime". Do NOT use for: general Next.js UI bugs, Supabase schema design, or non-Vercel deployments.

Enforces test-driven development with 80%+ coverage across unit, integration, and E2E tests. Use when the user says: 'write tests first', 'TDD', 'add test coverage', 'test this feature', 'write unit tests', 'create integration test', 'add E2E test', 'fix with TDD', 'red green refactor', 'increase coverage'. Guides the RED-GREEN-REFACTOR cycle for new features, bug fixes, and refactoring. Do NOT use for running existing tests without changes, CI/CD pipeline setup, or code review without test focus.

Verify that tests actually detect the bugs they claim to detect. Checks if assertions would fail when the declared bug exists -- catches 'tests that pass even when broken' (PR #321 incident). Invoke with /test-falsify or auto-trigger after E2E/integration test creation. Use when: reviewing test quality, validating regression tests, reporting test results in PRs, or when a test seems too permissive. Do NOT use for: unit test generation, test coverage analysis, performance benchmarking, or linting.