sumik

このコマンドは、前回のgitタグから現在までの変更履歴を分析し、Keep a Changelog形式でCHANGELOG.mdエントリーを自動生成します。

会話履歴とステージ済み変更からConventional Commits準拠のコミットメッセージを生成・表示

docs/e2e-testing/RUNBOOK.md を読み込んでその中身を実行する

Webアプリケーションのコードを分析し、ユーザーストーリードキュメント（`docs/user-story.md`）とE2Eテストケースドキュメント（`docs/e2e.md`）を作成または更新します。

このコマンドは、CHANGELOG.mdから該当バージョンのエントリーを抽出し、annotated tagを自動作成します。

このコマンドは、現在のブランチの変更点とコミットメッセージを分析し、GitHubプルリクエスト用の文章を自動生成します。

Run react-doctor to diagnose React code for security, performance, correctness, and architecture issues. Outputs a 0-100 health score with actionable diagnostics.

CLAUDE.mdを再読み込みしてcompaction後のコンテキストを復元

現在のプロジェクトのserenaデータを最新化

Token-efficient Serena MCP command for structured app development and problem-solving

GitHub風差分ビューアでdiffを表示

Token-efficient app development agent using /serena command for structured problem-solving. Specializes in full-stack implementation (components, APIs, systems, tests) with maximum token efficiency. Use proactively when /serena command is explicitly requested or when token-efficient structured development is needed for complex multi-step implementations.

AWS cloud specialized Tachikoma execution agent. Handles Lambda, API Gateway, DynamoDB, CDK, EKS, S3, Bedrock, and all AWS services including security (IAM, KMS, GuardDuty), cost optimization, and SRE operations. Use proactively when working with AWS services, CDK infrastructure, serverless applications, or AWS-related code. Detects: cdk.json, samconfig.toml, serverless.yml, @aws-sdk in package.json, or boto3 in Python deps.

Google Cloud specialized Tachikoma execution agent. Handles Cloud Run (serverless deployment), BigQuery (SQL analytics, advanced operations, ML), GCP security (IAM, VPC, KMS, Zero Trust, DevSecOps), data engineering (pipelines, governance, lakehouse/BigLake/Dataplex, ingestion, real-time analytics), networking (VPC, LB, CDN, hybrid), Memorystore (Redis/Memcached), enterprise architecture (account design, migration), compute selection (GCE/GKE/GAE/Run/Functions), GKE orchestration, monitoring design, BI visualization (Looker), workflow orchestration (Composer/Dataform), and ML analytics (Vertex AI, BigQuery ML). Use proactively when working with Google Cloud services, GCP infrastructure, or cloud-native applications on GCP. Detects: cloudbuild.yaml, .gcloudignore, @google-cloud packages, Looker, Dataplex.

Infrastructure/DevOps specialized Tachikoma execution agent. Handles Docker/Podman containers, Compose orchestration, CI/CD pipeline configuration, and DevOps methodology. Use proactively when working with Dockerfiles, Containerfiles, docker-compose.yml, CI/CD configs, or implementing DevOps practices. Detects: Dockerfile, Containerfile, docker-compose.*, or podman-related files.

Terraform IaC specialized Tachikoma execution agent. Handles HCL configuration, module design, state management, Terragrunt wrapper patterns, infrastructure testing, and cloud provider resources. Use proactively when working with .tf files, terragrunt.hcl, or infrastructure as code definitions. Detects: .tf files or terragrunt.hcl.

AI/ML development specialized Tachikoma execution agent. Handles Vercel AI SDK integration, LangChain.js, RAG system building, MCP server/client development, LLMOps operations, and AI-assisted development patterns. Use proactively when building AI-powered web features, RAG pipelines, MCP integrations, or LLM application deployment. Detects: ai/@vercel/ai/@langchain in package.json, google-adk in Python deps, or @modelcontextprotocol/sdk.

Database specialized Tachikoma execution agent. Handles relational database design, normalization, SQL optimization, schema migrations, database testing, and database internals understanding. Use proactively when designing database schemas, writing complex SQL, optimizing queries, planning migrations, writing database tests, or troubleshooting database performance. Detects: .sql files, schema.prisma, or DB-related packages.

Documentation and technical writing specialized Tachikoma execution agent. Handles technical documentation (7Cs principle), README creation, LaTeX academic reports, Zenn tech articles, and AI-assisted copywriting. Use proactively when creating documentation, writing READMEs, writing tech blog posts, preparing academic reports, or crafting marketing copy. Does NOT write application code.

HTML slide creation specialized Tachikoma execution agent. Handles HTML slide deck creation using 3-layer separation model (Engine/Theme/Content), source material conversion to slides, slide theme customization, and new slide project initialization. Use proactively when creating presentation slides, converting markdown/notes to HTML slide decks, or initializing slide starter projects. Detects: decks/ directory with index.html, engine/slide.css, slide.js files.

Training program design and presentation improvement specialized agent. Handles training needs analysis (KSA, ADDIE, participant analysis), curriculum design (90/20/8 rule, EAT framework), behavior change design (CREATE Action Funnel, cheat/habit/conscious strategies), presentation content improvement (storytelling, 3-act structure, AIDMA, PUNCH principle, delivery techniques), and effective prose writing (7Cs, AI smell detection). Self-improving: updates own definition based on user feedback. Use proactively when designing training programs, reviewing presentation materials, improving workshop content, or creating training curricula.

Design system construction and governance specialized Tachikoma execution agent. Handles design system architecture, component pattern libraries, Figma variable/token management, design system governance, and organizational adoption strategy. Use proactively when building or evolving design systems, establishing component governance, defining design tokens at architecture level, or planning design system rollout across teams. Detects: design system architecture tasks, design-tokens directories, or component library governance needs.

Figma-to-code implementation specialized Tachikoma execution agent. Handles Figma MCP integration (all 13 tools), Code Connect mappings, design token synchronization, visual validation, and Tailwind CSS styling methodology. Use proactively when converting Figma designs to code, syncing design tokens, managing Code Connect mappings, or implementing pixel-perfect UI from Figma mockups. Detects: Figma URLs in user prompts, .figma/ directory, or design-system-rules files.

Frontend component implementation specialized Tachikoma execution agent. Handles component implementation with shadcn/ui, Storybook story creation and interaction testing, and data visualization (charts/dashboards). Use proactively when creating UI components with shadcn/ui, writing Storybook stories, building interactive interfaces, or creating data charts/dashboards. For design principles, Figma integration, design systems, or Tailwind CSS architecture, use タチコマ（デザイン） instead. Detects: components.json, .stories.tsx/.ts files.

UX strategy, visual design, and creative specialized Tachikoma execution agent. Handles UI/UX philosophy (Fluid Interfaces, motion theory, constraint design), design thinking process (empathize, define, ideate, prototype, test), graphic design fundamentals (form, color, typography, layout), AI experience design (mental models, maturity frameworks), and AI-assisted creative generation. Use proactively when conducting UX research, making visual design decisions, running design thinking workshops, designing AI user experiences, or creating design creatives. Does NOT handle Figma-to-code conversion or design system code implementation.

Full-stack JavaScript specialized Tachikoma execution agent. Handles NestJS/Express backend development, RESTful API design, structured logging, and full-stack integration. Use proactively when building backend APIs with NestJS or Express, implementing API endpoints, designing request/response patterns, or configuring application logging. Detects: package.json with express, nestjs, fastify, koa, or hapi dependency.

Next.js/React specialized Tachikoma execution agent. Handles Next.js 16 App Router, Server Components, React 19 features, Turbopack, Cache Components, and next-devtools MCP integration. Use proactively when implementing Next.js pages, components, API routes, middleware, or React features in Next.js projects. Detects: package.json with 'next' dependency.

Bash shell scripting specialized Tachikoma execution agent. Handles shell script automation, I/O pipelines, process control, system administration, and script testing/debugging. Use proactively when writing or maintaining shell scripts, automating system tasks, or building CLI tools. Detects: .sh files.

Go specialized Tachikoma execution agent. Handles Go development including clean code practices, GoF/concurrency/DDD design patterns, and Go internals (type system, memory, reflection). Use proactively when working on Go projects or writing Go code. Detects: go.mod.

Python specialized Tachikoma execution agent. Handles modern Python development with uv/ruff/mypy tooling, FastAPI, Google ADK agent building, Pythonic patterns, DDD Tactical Patterns (Entity/Value Object/Aggregate Root), Event-Driven Architecture (Domain Events, CQRS, Message Bus), Unit of Work, and Architectural Testing. Use proactively when working on Python projects, building FastAPI services, creating Google ADK AI agents, or implementing domain models with Clean Architecture. Detects: pyproject.toml or requirements.txt.

TypeScript specialized Tachikoma execution agent. Handles advanced type system patterns, generics, conditional types, GoF design patterns in TypeScript, and type-safe architecture. Use proactively when deep TypeScript expertise is needed: complex type definitions, type refactoring, generic utility creation, or migrating JavaScript to TypeScript. Detects: tsconfig.json.

Code review specialized Tachikoma (READ-ONLY). Reviews code for bugs, logic errors, security vulnerabilities, code quality issues, and adherence to project conventions, using confidence-based filtering to report only high-priority issues that truly matter. Use proactively after code implementation for code quality audits, PR review assistance, or identifying technical debt. Does NOT modify code - produces review reports and recommendations only.

E2E testing and browser automation specialized Tachikoma execution agent. Handles Playwright test design, browser automation via agent-browser CLI, visual testing, accessibility testing, and CI/CD integration. Use proactively when writing E2E tests with Playwright, automating browser interactions, or setting up browser-based test infrastructure. Detects: playwright.config.* files.

Observability specialized Tachikoma execution agent. Handles monitoring system design, OpenTelemetry instrumentation (traces, metrics, logs), structured logging architecture, SLO/SLI design, and alerting strategies. Use proactively when implementing observability, distributed tracing, metrics collection, log pipelines, or monitoring dashboards. Detects: @opentelemetry/* packages or prometheus.yml.

Security review specialized Tachikoma (READ-ONLY). Reviews code for OWASP Top 10 vulnerabilities, serverless security threats, IAM patterns, dynamic authorization (ABAC/ReBAC/Cedar), Keycloak IAM, and AI development security. Use proactively after code implementation for security audits, penetration test planning, or access control design. Does NOT modify code - produces security reports and recommendations only.

Unit/Integration testing specialized Tachikoma execution agent. Handles TDD methodology, test design across all languages (Vitest/Jest, Go testing, pytest, etc.), React Testing Library, mock strategies, coverage optimization, and test refactoring. Use proactively when writing unit tests, integration tests, improving test coverage, or setting up test infrastructure. Detects: files containing 'test' or 'spec' in name (*test*, *spec*, *_test.go, test_*.py), vitest.config.*, jest.config.*, or pytest.ini.

Architecture design specialized Tachikoma (READ-ONLY). Analyzes systems using DDD strategic/tactical patterns, microservices architecture (CQRS, Saga, Event Sourcing), trade-off analysis methodology, data architecture patterns, and Clean Architecture (dependency rules, concentric layer model, component principles). Use proactively for architecture reviews, system design, domain boundary analysis, modernization planning, or multi-tenant SaaS design. Does NOT write implementation code - produces design documents and recommendations only.

Product management specialized Tachikoma agent (READ-ONLY). Handles PRD creation, roadmap planning, prioritization (RICE/ICE/MoSCoW), user story writing, A/B test design, growth metrics analysis (AARRR pirate metrics), AI product maturity assessment, PM-UX collaboration design, technical trade-off analysis for product decisions, and behavior change design (CREATE Action Funnel). Use proactively when creating PRDs, planning roadmaps, prioritizing features, designing experiments, analyzing product metrics, evaluating AI product readiness, or making product-technical trade-off decisions. Does NOT write implementation code - produces product documents and recommendations only.

General-purpose Tachikoma execution agent for tasks not covered by specialized Tachikomas. Handles development, testing, documentation, and other technical tasks. Uses /serena for efficient development. When a task matches a specialized domain (Next.js, Python, AWS, etc.), prefer the domain-specific Tachikoma instead. PARALLEL EXECUTION: Claude Code can launch multiple Tachikoma instances for independent tasks.

Behavioral design methodology for products that change user behavior using CREATE Action Funnel (Cue, Reaction, Evaluation, Ability, Timing) and three strategies (cheat, habit, conscious action). Use when designing product features that aim to change user habits, increase engagement, or guide users toward beneficial actions. For visual UI/UX design principles, use designing-ux instead. For training program design and facilitation methodology, use designing-training instead. For human-centered design thinking process (user research, ideation, prototyping), use practicing-design-thinking instead.

Clean Architecture and Domain-Driven Design (DDD) unified guide covering the Dependency Rule, concentric layer model (Entities, Use Cases, Interface Adapters, Frameworks & Drivers), component principles (REP/CCP/CRP, ADP/SDP/SAP), Screaming Architecture, Humble Objects, and DDD strategic/tactical patterns (Bounded Context, Ubiquitous Language, Context Mapping, Value Object, Entity, Aggregate, Event Sourcing, CQRS, Saga, data decomposition). Use when designing application layer structure, enforcing dependency rules, separating business logic from frameworks/UI/databases, modeling complex domains with DDD patterns, or mapping Bounded Contexts to Clean Architecture layers. For code-level clean practices, use writing-clean-code. For distributed/infrastructure patterns, use architecting-infrastructure. For evolutionary architecture and microservices intro, see ARCHITECTURE-EVOLUTION.md. For project foundations and team practices, use practicing-software-engineering.

Semantic Versioning 2.0.0仕様に基づくバージョン判断ガイド。MAJOR/MINOR/PATCH判定、プレリリース・ビルドメタデータ、範囲指定を提供。 REQUIRED for all version-related decisions. Use when determining version bumps, creating releases, or managing dependencies.

Data architecture patterns covering read-side optimization (replicas, materialized views, CQRS, CDC, event sourcing), domain-based decomposition, polyglot persistence, and caching strategies (cache-aside, read-through, write-through, write-around). Use when designing data flow architecture, choosing read scalability strategies, or implementing caching for enterprise systems. For microservices patterns (Saga, granularity), use architecting-infrastructure instead. For DDD domain modeling, use applying-domain-driven-design instead. For database engine internals, use developing-databases instead. For relational DB schema design, use developing-databases instead. For GCP-specific data services (BigQuery, Dataflow, Dataproc), lakehouse (BigLake, Dataplex), and BigQuery advanced operations (editions, HA/DR), use developing-google-cloud instead.

Vendor-neutral infrastructure and system architecture patterns. Infrastructure: 127 patterns for availability (HA, DR, backup), security, performance/scalability, operations, network/storage configuration, and cloud. Modernization: socio-technical strategy, domain redesign, trade-off analysis (code/API/system/meta levels). Microservices: CQRS, Event Sourcing, Saga, distributed transactions, service granularity, data ownership, resilience, messaging. Use when designing infrastructure, evaluating non-functional requirements, modernizing legacy systems, making architectural trade-offs, or designing microservices. For DevOps (CI/CD, IaC), use practicing-devops. For DDD domain modeling, use applying-domain-driven-design. For Clean Architecture, use applying-clean-architecture. For observability patterns, use implementing-observability instead. For GCP enterprise architecture, use developing-google-cloud instead. For GPU/CUDA performance tuning and AI workload optimization, use designing-genai-patterns instead.

Claude Code Plugin開発ガイド（Agent定義・Skill定義・コマンド定義の作成・最適化）。 Use when creating or modifying agents, skills, or commands for Claude Code plugins. フロントマター仕様、description設計、Progressive Disclosure、ツール制限、テスト手法を含む。

Browser Agent CLIによるブラウザ操作自動化（セマンティックロケーター、状態永続化、ネットワーク傍受）。 Use when automating browser interactions via agent-browser CLI (NOT for E2E testing). E2Eテストは testing-e2e-with-playwright スキルを参照。

AIエージェント構築ガイド（LangChain/LangGraph・Google ADK・リアルタイムマルチモーダル）。 Use when building AI agents with LangChain, LangGraph, Google ADK, or Gemini Live API. フレームワーク選択、ツール定義、マルチエージェント、A2A、リアルタイム音声/動画を含む。

Design system construction, governance, Figma implementation, and practical launch/adoption methodology covering system foundations, pattern taxonomy (CEV naming), organizational adoption (consensus building, proposal templates), pattern library operations, UI pattern catalog (20+ patterns), Figma variable/token architecture, incremental launch, buy-in strategies, content creation, CI/CD automation, update notification framework (6 steps), operational ownership patterns, case studies, and self-assessment. Use when architecting, evaluating, or restructuring a design system, planning organizational adoption, building a pattern library, constructing a design system within Figma, starting a DS from scratch, or creating design system content. NOT for component styling (applying-design-guidelines), design-to-code (implementing-design), frontend codegen (designing-frontend), or Tailwind config (styling-with-tailwind).

Multi-tenant SaaS architecture covering deployment models, tenant isolation, and data partitioning. Use when designing SaaS platforms or evaluating silo vs pool strategies. Covers identity, onboarding, tiering, and operations. For Next.js single-tenant SaaS (auth, payments), use building-nextjs-saas instead.

Claude Code Agent定義（.md）をCodex subagent定義（agents/*.toml）に変換するガイド。 フィールドマッピング・developer_instructions変換・skills.config展開・検証手順・よくある失敗パターンを網羅。 Use when converting Claude Code agents to Codex format, migrating agent definitions, or setting up Codex subagents. Triggers: "agentをcodexに変換", "codex agent変換", "agent migration"。

コンテンツ変換ガイド（画像ベースEPUB→テキストOCR変換・LM Studio英日翻訳）。 Use when converting image-based EPUBs to text or translating content with LM Studio. pandoc、recognize-image.py OCR、ローカルLLM翻訳ワークフローを含む。

コンテンツ制作統合スキル。AIコピーライティング（15テクニック＋心理的トリガーでマーケティングコピー・ブログタイトル・広告見出し・SNS投稿を生成）とAIデザインクリエイティブ（バナー・SNS・ポスター等の広告ビジュアル制作プロンプト設計）を統合。 Use when creating marketing copy, blog titles, ad headlines, social media posts, or generating visual creatives (banners, social images, posters) with AI tools. For eliminating AI smell from text use writing-effective-prose. For UI/UX design principles use designing-ux. For presentations use creating-slides. For LaTeX → writing-latex.

ダイアグラム作成ガイド（Mermaid構文・draw.io MCP・24種類のダイアグラム対応）。 Use when creating diagrams, charts, or visual documentation with Mermaid or draw.io. C4モデル、ER図、シーケンス図、フローチャート、ガントチャート等を含む。

Creates Anki flashcards from EPUB/PDF files via Anki MCP Server, covering MCP setup, deck/note-type management, and bulk import with HTML formatting. Use when converting textbooks, question banks, or study materials into spaced repetition flashcards, or when managing Anki cards via MCP tools. Covers full workflow: MCP setup → file conversion → content analysis → batch card creation. For MCP server/client development, use developing-mcp instead.

HTMLスライド作成スキル。slides repoの3層分離モデル（Engine/Theme/Content）に基づく16:9 HTMLプレゼンデッキの作成・テーマカスタマイズ・ソース素材変換をガイド。認知科学・ロジック構築（接続関係診断・基本ロジック図）・ストーリーテリング・聴衆分析・スライドデザイン・ビジュアルデザイン実践・提案書構成術・デリバリー（実践Tips集含む）・準備プロセスの9つのリファレンスで品質を担保。 Use when creating HTML slide decks, converting source materials to slides, customizing slide themes, or initializing new slide projects. Delegates actual slide creation to tachikoma-slide agent. For presentation quality improvement (storytelling, delivery, engagement) use designing-training. For text polishing use writing-effective-prose. For marketing copy use creating-content. For proposal structure (ロジック3つの力, MECE, メッセージマップ) see QUALITY-PROPOSAL-CONSTRUCTION.md.

Comprehensive data visualization principles covering chart type selection, color scales, design best practices, and data storytelling. Use when creating charts, dashboards, or any visual data representation. For UI/UX design principles, use designing-ux instead. For technical diagrams (flowcharts, architecture), use mermaid-diagrams instead.

Creates distinctive, production-grade frontend code with shadcn/ui integration, object-oriented UI (OOUI) methodology, and micro-frontend architecture patterns. Use when implementing web components, pages, or applications requiring creative, polished UI code; when designing UI structure using object-oriented approach (object extraction, view/navigation patterns, layout pattern selection); or when architecting micro-frontend systems (team splitting, Module Federation, BFF patterns, migration strategy, Conway's Law application). For Storybook story creation and component testing, use developing-storybook instead. For theoretical UI/UX design principles, use designing-ux instead. For Tailwind CSS methodology, component design patterns, and design system construction, use styling-with-tailwind instead. For design system methodology (pattern language, organizational strategy, UI pattern catalog, anti-patterns), use building-design-systems instead.

32 production-grade GenAI design patterns covering content control (Logits Masking, Grammar, Style Transfer), RAG architecture, model capabilities (Chain of Thought, Adapter Tuning), reliability (LLM-as-Judge, Reflection, Prompt Optimization), agentic systems (Tool Calling, Multiagent Collaboration), deployment optimization (SLM, Prompt Caching, Inference Optimization), safety guardrails (Self-Check, Guardrails), RAG implementation (11 source types, chunking, vector stores), and LLMOps/AgentOps (maturity L0–2, Tool/Agent Registry, Memory Governance, LLM metrics, LLMSecOps). Also covers AI performance (GPU/CUDA, LLM inference) via PERF- references. Use when designing GenAI applications, choosing RAG strategies, implementing agent architectures, optimizing LLM reliability, or operating AI systems in production. For web AI (Vercel AI SDK, LangChain.js), use integrating-ai-web-apps. For GCP-specific ML (BigQuery ML, Vertex AI), use developing-google-cloud.

Training program design and facilitation methodology covering needs analysis (participant analysis, stakeholder mapping), KSA framework, instructional design theory (ADDIE, TOTE, Gagné, mastery learning), objective setting, activity catalog, training operations (instructor, space, logistics), brain science principles (7±2 rule, long-term memory, habit formation), learner-centered facilitation (90/20/8 rule, EAT), online/hybrid delivery, skill map creation, and training material development. Use when designing training programs, facilitating workshops, creating training materials, building skill maps, or managing training operations. For behavior design→applying-behavior-design. For presentation content→creating-presentations. For design thinking process methodology (empathize/define/ideate/prototype/test)→practicing-design-thinking.

UXデザイン総合ガイド（認知心理学・UIガイドライン・グラフィック基礎・デザイン思考・AIエクスペリエンス設計・Webビジュアルデザイン実践・UXメソッド実践）。 Use when designing user experiences, UI, conducting design thinking workshops, designing AI interfaces, creating web page layouts and visual designs, defining web design concepts using functionality/emotionality frameworks, or implementing UX methods such as usability evaluations, prototyping, user research, journey mapping, and organizational UX adoption. デザイン思考（d.school 5ステップ）、AIインターフェース設計（メンタルモデル、Copilotパターン）、Webレイアウト・配色・UIコンポーネント・モーションデザイン、Webデザイン機能性7軸・情緒性6軸フレームワーク、デザインコンセプト立案プロセス、レイアウト実践パターン（即応用可能な判断パターン集）、イメージワードシステム（デザイン方向性の言語化・共有手法）、クリエイティブプロセスパターン（プロの制作プロセス・発想法・試作サイクル）を含む。実践メソッド（PRACTICE-*）としてユーザビリティ評価・プロトタイピング・構造化シナリオ法・ユーザーリサーチ・カスタマージャーニーマップ・組織導入戦略を含む。

AWS development guide covering serverless (Lambda, API Gateway, DynamoDB), CDK, EKS, ECS/Fargate, SRE, FinOps, security (IAM, VPC, KMS, GuardDuty), GenAI (Bedrock, RAG), databases (Aurora, ElastiCache), data engineering (Glue, Athena, Redshift), 57 Cloud Design Patterns, VPC architecture, enterprise architecture (multi-account, Landing Zone, Cognito), cloud migration (7R, DMS), and HA/fault tolerance. Use when working with AWS services, CDK infrastructure, or serverless applications. For GCP, use developing-google-cloud. For Terraform IaC, use developing-terraform. MUST load when working with AWS services, detected by aws-cdk or @aws-sdk in package.json, cdk.json, samconfig.toml, serverless.yml, template.yaml (SAM), or eksctl configs. For GCP→developing-google-cloud, Terraform→developing-terraform, Docker→managing-containers, monitoring→implementing-observability, RAG→designing-genai-patterns, LLMOps→practicing-llmops, serverless security→securing-serverless, logging→implementing-observability.

Comprehensive Bash shell scripting and automation guide covering fundamentals, control flow, I/O pipelines, process control, system administration automation, testing/debugging, security, design patterns, and penetration testing automation (Nmap, web scanning, exploitation, Wi-Fi assessment, reporting). MUST load when .sh files are detected or shell scripts are being written. For Docker-specific patterns, use managing-containers instead. For broader DevOps methodology, use practicing-devops instead.

Comprehensive database development guide covering relational design (ER modeling, normalization, PostgreSQL implementation), SQL antipatterns (25 patterns with detection signals and solutions), database internals (storage engines B-tree/LSM, distributed systems, consensus algorithms), and PostgreSQL operations (server config, extensions, query tuning, MVCC/VACUUM, backup/PITR, replication/HA, monitoring). Use when designing database schemas, writing SQL queries, reviewing data models, choosing storage engines, tuning PostgreSQL performance, managing backup/replication, or debugging distributed system behavior. Replaces: designing-relational-databases, avoiding-sql-antipatterns, understanding-database-internals. For data architecture patterns (CQRS, event sourcing), use architecting-data instead. For GCP managed databases (Cloud SQL, Spanner, Firestore, Bigtable) and BigQuery analytics, use developing-google-cloud instead.

Full-stack JavaScript development covering backend (NestJS/Express), frontend (React), deployment (CI/CD), and quality; also covers JavaScript language fundamentals (types, closures, prototypes, async/await, modules, and metaprogramming). MUST load when package.json contains Express, NestJS, or similar backend frameworks. Covers API design, state management, caching, and testing. Also use when working with JavaScript language-level concerns — type coercion, scoping, this binding, Promise patterns, Proxy/Reflect, or iterator protocols. For application logging design and structured logging patterns, use implementing-observability. For TypeScript type system and advanced type patterns, use mastering-typescript.

Comprehensive Go development guide covering clean code practices, design patterns (GoF/concurrency/DDD), and internals (type system/memory/reflection). MUST load when go.mod is detected or Go code is being written. Covers naming, error handling, concurrency, testing, project structure, function design, data structures, refactoring strategies, GoF patterns, architectural patterns, type system internals, and performance optimization. For application logging design and structured logging patterns, use implementing-observability.

Google Cloud guide: Cloud Run, GCP security (IAM/VPC/KMS/Zero Trust), data engineering (BigQuery/Dataflow/pipelines/governance/lakehouse), networking (VPC/LB/CDN), Memorystore, enterprise architecture, compute selection (GCE/GKE/GAE/Run/Functions), GKE, GAE, monitoring (SLO/SLI), BigQuery analytics (SQL/window functions/GA4), BigQuery ops (editions/HA-DR/slots), workflow (Composer/Dataform/Data Fusion), BI (Looker/Looker Studio/BI Engine), ingestion (DTS/Datastream CDC/GA4), real-time (Pub/Sub/Dataflow streaming), ML (BigQuery ML/Vertex AI/GIS). MUST load when google-cloud packages, cloudbuild.yaml, BigQuery/Dataflow/Memorystore/Looker/Dataplex detected. For Docker→managing-containers. For monitoring→implementing-observability. For OWASP→securing-code. For CQRS→architecting-data. For AWS→developing-aws. For GenAI→designing-genai-patterns.

Comprehensive MCP (Model Context Protocol) development guide covering architecture (Host/Client/Server roles, Tools/Resources/Prompts, Control Segregation), server and client implementation with TypeScript SDK, protocol specification (JSON-RPC 2.0, stdio/Streamable HTTP), and security threats (Tool Poisoning, Shadowing, Rug Pull, Prompt Injection). MUST load when building MCP servers or clients. For Claude Code plugin MCP configuration, use plugin-dev:mcp-integration instead. For MCP integration with Vercel AI SDK in web apps, see integrating-ai-web-apps. For consuming MCP tools from LangGraph agents, see building-langchain-agents.

Next.js 16.x development guide covering App Router, Server Components, Turbopack, React Compiler, proxy.ts, Cache Components, opt-in caching APIs (updateTag/revalidateTag/refresh), strict TypeScript, Tailwind CSS 4.x, Prisma 7.x, Zod 4.x, Vitest 4.x, Playwright, Docker. MUST load when package.json contains 'next' or next.config.* is detected. For React-specific concerns (internals, performance rules, animation, RTL testing), use developing-react. SaaS構築パターン（認証・決済・AI統合）を含む。 For AI integration with Vercel AI SDK and LangChain.js, use integrating-ai-web-apps. For application logging design and structured logging patterns, use implementing-observability. For Tailwind CSS methodology, component design patterns, and customization, use styling-with-tailwind.

Modern Python development guide covering project setup, tooling, Pythonic best practices, OOP design principles, GoF design patterns, SE process, and type safety. MUST load when pyproject.toml or requirements.txt is detected. Covers Python 3.13 + uv + ruff + mypy, FastAPI/FastMCP, pytest, Docker, Effective Python idioms, software design patterns (encapsulation, LSP, Design by Contract), SE practices (SDLC, Scrum/Kanban, CI/CD, deployment), type safety rules (Any prohibition, TypedDict, Protocol, dataclass), Architecture Patterns (Repository, Unit of Work, Aggregates, Domain Events, CQRS, Event-Driven Architecture), and DDD Tactical Patterns (Value Object, Entity, Aggregate Root). For language-agnostic clean code principles, use writing-clean-code. For application logging design, use implementing-observability. For Clean Architecture in Python, see references/CA-PYTHON.md. For general Clean Architecture principles, use applying-clean-architecture.

React 19.x development guide covering internals (rendering, reconciliation, Fiber), design patterns (Container/Presenter, HOC, Render Props, Compound Components), state management (useState to Zustand/TanStack Query, nuqs/Jotai), error handling (ErrorBoundary, React 19 error APIs), accessibility (ARIA, focus management, keyboard navigation, WCAG 2.1 AA), performance optimization (react-doctor, React Compiler, memoization), UI animation patterns (CSS transitions, easing), React Testing Library (RTL queries, interactions, TDD patterns), and Storybook (CSF3, interaction testing, a11y). Use when package.json contains 'react' (without 'next'), or when working on React-specific concerns in any framework. Also covers Storybook story creation, play function interaction tests, and axe-core a11y testing. For Next.js-specific features, use developing-nextjs instead. For E2E testing, use testing-e2e-with-playwright. For general testing methodology, use testing-code.

Terraform IaC development with HCL syntax, module design, state management, and Terragrunt wrapper patterns. MUST load when .tf files or terragrunt.hcl are detected, or go.mod contains 'terraform'. Covers AWS/GCP infrastructure patterns and mise task automation. For Docker infrastructure, use managing-containers instead. For broader DevOps methodology and IaC tool comparison, use practicing-devops instead. For AWS CDK (TypeScript/Python IaC), use developing-aws instead.

Web API development guide covering design (endpoints, responses, HTTP spec, versioning, security), spec-first methodology (spec writing, E2E test framework architecture, technical debt repayment), and testing strategy (design testing, contract testing, automation, performance, security testing). Use when designing REST APIs, writing API specs, adopting spec-first development, or building API test strategies. For framework-specific implementation (Express, NestJS), use respective framework skill. For microservices patterns and trade-off analysis of API usability vs maintenance, use architecting-infrastructure instead.

AI-assisted development methodology covering prompt engineering, context engineering, code generation workflows, quality assurance, collaborative debugging, agent collaboration patterns, LLM dialogue design patterns (8 conversational patterns, Grammar of Intention), team LLM adoption workshops, and role-specific augmentation (Dev/PO/Coach/Manager). MUST load when working with AI coding tools or discussing AI-assisted development practices. Use when crafting prompts for code generation, managing AI context, reviewing AI-generated code, coordinating multi-agent workflows, or designing structured LLM dialogues. For AI copywriting, use crafting-ai-content. For testing-specific AI techniques, use testing-code. For AI dev security (trust frameworks, governance, AI-SPM), use securing-ai-development. For PM-specific Claude Code workflows, use using-claude-code-as-pm. For LLM production operations, use practicing-llmops. For AI-enhanced workflows and team AI adoption, use practicing-software-engineering instead.

LLM evaluation and red-teaming toolkit using promptfoo. Covers promptfooconfig.yaml configuration, 40+ assertion types (deterministic, model-graded, RAG), provider setup (OpenAI, Anthropic, Google, Ollama, HTTP, custom JS/Python), red teaming (134+ plugins, jailbreak strategies, compliance frameworks), CLI commands, caching, and CI/CD integration. Use when writing promptfooconfig.yaml, designing LLM test suites, running adversarial red team evaluations, or integrating LLM quality gates in CI/CD. Detects: promptfooconfig.yaml or promptfoo in package.json. For general LLMOps operations, use practicing-llmops. For general test methodology (TDD/AAA), use testing-code.

Operates as Tachikoma Agent (Implementation Worker) performing actual code implementation. Use when receiving task assignments from Claude Code for implementation work, test creation, or documentation. Adapts to frontend, backend, or testing roles based on task assignment.

Comprehensive design-to-code implementation skill covering general design principles, Figma MCP integration, and Figma UI design workflows. General: design system integration, visual parity validation, responsive implementation, accessibility. Figma MCP (13 tools): basic/advanced workflows (Figma Make, Code Connect, Design System Rules, design token sync), visual validation. Figma UI design: wireframe→prototype→detailed design→handoff workflow, 8pt grid, component-driven design, UIStack (5 states), style naming conventions, engineer collaboration. Use when implementing designs from any source (screenshots, mockups, Figma URLs, specs) or when designing mobile UI in Figma and preparing design handoff. Requires Figma MCP server for Figma-specific code generation workflows. For design system architecture and organizational adoption strategy, use building-design-systems instead.

Dynamic authorization design covering ABAC, ReBAC, PBAC models and Cedar policy language. Use when designing access control systems, choosing authorization models, or writing Cedar policies. Distinct from securing-code (code-level) by focusing on authorization model selection and policy-based access control.

Unified observability guide covering monitoring system design (anti-patterns, 6-layer strategy, alerting, SLOs, telemetry pipelines, observability maturity model), OpenTelemetry implementation (tracing/metrics/logs API/SDK, Collector deployment, semantic conventions), logging design (structured logging, collection pipelines, analysis, anomaly detection, security compliance), and observability engineering practices (core analysis loop, debugging, ODD, build-vs-buy ROI, CI/CD observability, high-cardinality data). Use when designing monitoring strategy, implementing OpenTelemetry instrumentation, building log pipelines, setting up alerts, establishing SLO-based reliability, or adopting observability practices. Replaces: designing-monitoring, implementing-opentelemetry, implementing-logging. For LLM monitoring, use practicing-llmops instead. For DevOps (CI/CD, IaC), use practicing-devops instead. For GPU/CUDA profiling (Nsight Systems/Compute, PyTorch Profiler), use designing-genai-patterns instead.

Comprehensive guide to integrating generative AI into web applications using Vercel AI SDK, LangChain.js, and MCP with React/Next.js. Use when building AI-powered web features (streaming chat, RAG, tool calling, structured data generation) in JavaScript/TypeScript projects. For RAG system internals, use designing-genai-patterns. For MCP protocol development, use developing-mcp. For Next.js framework specifics, use developing-nextjs. For Python LangChain/LangGraph agent development, use building-langchain-agents. For framework-agnostic GenAI design patterns (32 patterns covering content control through safety guardrails), use designing-genai-patterns.

Manages CLAUDE.md creation, improvement, and maintenance following 8 core principles (size limits, progressive disclosure, essential elements, living documentation). Use when starting new projects, reviewing existing CLAUDE.md, or when Claude repeatedly makes the same mistakes. For skill file authoring, use authoring-skills instead.

Comprehensive Keycloak identity and access management guide covering OIDC/SAML protocols, SSO, Realms, Clients, Users, Roles, authentication flows, MFA, authorization policies, JWT token management, application integration (Java/Node.js/Spring Boot), Docker/Kubernetes deployment, clustering, and SPI extensions. MUST load when keycloak.json, keycloak.conf, or Keycloak-related dependencies are detected. For general security practices, use securing-code instead. For Cedar/ABAC/ReBAC authorization models, use implementing-dynamic-authorization instead.

Comprehensive TypeScript guide covering type system, advanced patterns, design patterns (GoF Creational/Structural/Behavioral), functional/reactive programming, framework integration, and 83 best practice decision criteria. MUST load when working in TypeScript projects detected by tsconfig.json. Covers Generics, Conditional Types, type inference, type design, SOLID/DDD in TypeScript, anti-patterns, and practical migration tips. Includes type safety rules (any prohibition, type guards, strict mode) formerly in enforcing-type-safety.

Codex CLI統合スキル（基本操作・プランレビュー・Agentオーケストレーション）。 Use when using Codex for code consultation, plan review, or orchestrating Codex agents in Wave parallel execution. Triggers: "codex", "codexと相談", "codexでレビュー", "codexでオーケストレーション"。Wave計画に基づき自然言語でagentを同時起動し、config.tomlのmax_threadsで並列度を宣言的に制御する。 For agent conversion to Codex format, use converting-agents-to-codex instead.

Agent Teamオーケストレーション（チーム編成・タチコマ並列起動・進捗管理）。 Use when multiple files need parallel changes with independent concerns (2+ files, frontend+backend, UI+API+test). Triggers: 複数ファイル並列変更、マルチ関心事開発、フロントエンド+バックエンド同時変更。

DevOps実践ガイド（進化3ステージ・IaCツール選定・オーケストレーション・CI/CD・コンテナ管理）。 Use when practicing DevOps, selecting IaC tools, designing CI/CD pipelines, or managing Docker/Podman containers. Docker/Podman、マルチステージビルド、Compose、セキュリティ強化を含む。 For continuous delivery philosophy, CD maturity assessment, and deployment pipeline design principles, see CONTINUOUS-DELIVERY.md subfile. For project foundations, team organization, and developer habits, use practicing-software-engineering instead. For monitoring and observability design, use implementing-observability instead. For GCP CI/CD and GKE container orchestration, use developing-google-cloud instead. For GPU-optimized Docker/Kubernetes environment tuning (NVIDIA Container Toolkit, MPS, MIG), use designing-genai-patterns instead.

プロダクトマネジメント総合ガイド（基礎・AIプロダクト・A/Bテスト・成長戦略・GPM・PLG・カスタマーサクセス・Claude Code PM活用）。 Use when practicing product management, managing AI products, designing experiments, growth product management, PLG strategy, customer success, or using Claude Code as PM tool. AARRR海賊指標、AIプロダクトライフサイクル、MLプロダクト化、AI統合戦略、オンライン実験設計、PM調査パターン、GPM実践、リテンション戦略、拡張収益を含む。 For engineer-focused user story writing techniques (templates, common mistakes, splitting), use writing-user-stories instead.

Comprehensive SW engineering practices covering project foundations (fast feedback, small steps, DORA metrics), team organization (Team Topologies, 4 types), pair programming (4 patterns), developer habits (GREAT framework), IC effectiveness mindset (outcomes vs outputs, strategic prioritization), career-stage skills (junior to staff, IC vs management), cross-functional influence (PM/design, authority-free leadership), 20 anti-patterns (15 individual + 5 team-level), sustainable performance (burnout, remote work), and AI-enhanced workflows (daily AI, 90-day rollout). Use when starting projects, organizing teams, for IC effectiveness, career growth, or team AI culture. For TDD/BDD/ATDD, use testing-code instead. For SOLID/refactoring, use writing-clean-code instead. For CI/CD, use practicing-devops instead. For DDD/Clean Architecture, use applying-clean-architecture instead. For ODD/debugging, use implementing-observability instead. For prompt engineering techniques, use developing-with-ai instead.

Researches existing libraries before implementation to prevent reinventing the wheel. REQUIRED before writing any new functionality. Use when evaluating npm packages, pip packages, Go modules, or any third-party libraries to find and evaluate existing solutions.

コードレビューガイドライン（PR構成・効果的コメント・トーン原則・CodeRabbit統合）。 Use when reviewing code, creating PRs, or using CodeRabbit for automated review. 3原則（客観性・具体性・明確なアウトカム）、自動修正ループを含む。

Web検索統合スキル。Exa MCP（第一優先）による高度検索（7カテゴリ: Company, Code, People, Financial Report, Research Paper, Personal Site, Tweet/X）と、gemini CLI（フォールバック）による複雑なクエリ検索を統合。 Use when performing web searches, researching technologies, companies, people, academic papers, or social media. Load Exa MCP first; fall back to gemini CLI when Exa is unavailable. MUST load as first-choice search tool. context7 for library docs, deepwiki for GitHub wiki, WebSearch built-in as last resort.

Organizational security strategy for AI-powered software development covering trust frameworks, adaptive guardrails, AI-BOM, AI-SPM, governance models, and cross-functional ownership. Use when establishing security controls for AI coding assistants, agentic systems, or AI-accelerated SDLC workflows. For code-level security (OWASP, CodeGuard), use securing-code instead. For AI development methodology (prompts, context engineering), use developing-with-ai instead. For LLM-specific security (prompt injection, LLMSecOps), use practicing-llmops.

REQUIRED after all code implementations. Automatically load when implementation is complete to run CodeGuard security check. Covers input validation, secrets management, OWASP top 10 countermeasures, authentication/authorization patterns, web penetration testing (reconnaissance, attack techniques, bypass methods), and serverless security (IAM/storage/functions across AWS/GCP/Azure, 17 threat categories, supply chain attacks). Use when implementing any code, handling external input, or developing serverless applications (Lambda, Cloud Run, Azure Functions). For dynamic authorization model design (ABAC/ReBAC/Cedar), use implementing-dynamic-authorization instead. For organizational AI development security strategy (trust frameworks, governance, AI-BOM), use securing-ai-development instead. For security logging patterns and compliance logging, use implementing-observability.

Comprehensive algorithm and data structure reference for competitive programming covering sorting, searching, trees, graphs, dynamic programming, computational geometry, and number theory with complexity analysis and language-agnostic implementations. Use when solving algorithmic problems, optimizing code for time/space complexity, or implementing classic data structures (stack, queue, heap, BST, union-find). For database-specific data structures (B-tree, LSM), use developing-databases instead.

Tailwind CSS styling methodology covering utility-first philosophy, v4 CSS-first config, utility/modifier reference, component design patterns, customization (plugins, presets, JS API), migration strategies, and design system construction with design tokens. Use when styling web applications with Tailwind CSS or making CSS architecture decisions. For frontend code generation with shadcn/ui and Storybook, use designing-frontend instead. For general UI/UX design principles, use designing-ux instead. For design system strategy beyond CSS (pattern language, organizational adoption, UI pattern catalog, measurement), use building-design-systems instead.

REQUIRED for all feature implementations. Automatically load when writing or reviewing tests. Enforces TDD approach with AAA pattern, actual/expected variables, and 100% coverage goal for business logic. Covers Vitest, Jest, Playwright, and AI-augmented testing strategies. Includes Khorikov's Four Pillars (regression protection, refactoring resistance, fast feedback, maintainability), three testing styles (output/state/communication-based), code classification, and test anti-patterns. For RTL patterns, use developing-react. For Web API testing, use developing-web-apis. For A/B testing, use conducting-ab-tests. For TDD mindset/ATDD/BDD, see TESTING-PRACTICES.md. For project foundations and team organization, use practicing-software-engineering.

Playwright E2Eテストの設計・実装・運用ガイド。 Use when writing, reviewing, or maintaining E2E tests with @playwright/test. playwright.config.* 検出時に自動ロード。testing-code（テスト全般）とは異なりPlaywright E2E固有のパターンに特化。

Next.js development integration tools via next-devtools MCP. MUST load when working on Next.js projects detected by package.json. Provides diagnostics, version upgrades, Cache Components optimization, and automatic error fixes. Primary tool for Next.js tasks.

Token-efficient structured development via /serena command. Use when implementing components, APIs, system designs, tests, bug fixes, or optimizations with structured problem-solving approach. Available for Claude Code本体 and all Agents including Tachikoma.

REQUIRED for all code implementations. Automatically load when writing or reviewing any code. Covers SOLID principles, MAPPER principles with 25 code smell categories, practical refactoring, software design laws, Kent Beck's 4 rules of simple design, Uncle Bob's 66 code smell heuristics, formatting principles, and boundary management patterns. Language-agnostic clean code guide. For language-specific practices, use developing-go, mastering-typescript, or developing-python. For architecture-level clean design, use applying-clean-architecture. For legacy code refactoring, see CHANGEABLE-CODE.md subfile. For refactoring mindset (spikes, feature toggles, structural discovery), see REFACTORING-MINDSET.md reference.

Conventional Commits 1.0.0仕様に基づくコミットメッセージフォーマットガイド。type/scope/BREAKING CHANGEの判定とSemVerとの対応を提供。 REQUIRED for all commit messages and version descriptions. Use when writing git commit messages.

Unified writing craft guide: prose fundamentals, AI smell detection (always active), technical docs (7Cs), engineering design docs (requirements spec, functional design, operation design), academic writing (Harvard, dissertation), university report/thesis writing, tech blog writing, Japanese prose craft, clarity/explanation techniques, web/digital writing, FAQ writing, revision, README creation, and Zenn technical article publishing. REQUIRED for all text output — AI smell check is always active regardless of document type. Use when writing, reviewing, proofreading, or creating any document (technical docs, articles, engineering design documents (requirements specs, system design docs), reports, slides, emails, academic papers, dissertations, university reports, graduation theses, business documents, web content, FAQ/help docs, design specs, README.md files, Zenn tech articles). For LaTeX → writing-latex. For presentations → creating-presentations.

LaTeX document creation for Japanese academic reports. MUST load when .tex files are detected. Covers upLaTeX + dvipdfmx setup, minted code highlighting with Japanese font support, equations, figures, tables, and cover pages.

Guides effective user story creation for software projects covering story templates (As a.../I want.../So that...), common mistakes, technical requirements to story conversion, acceptance criteria writing, and story splitting techniques. Use when writing user stories, converting technical requirements to stories, or improving backlog quality. For product management practices (PRD, roadmap, prioritization), use practicing-product-management instead.

1 hook across 1 event