LLM & Agents

Implementar serviços de lógica de negócio em services/. Use quando criar service, lógica de negócio, orquestração de repositórios, ou integração com LLM.

Deploys ProAgentic DfX (H2 Tank Designer) using the full CI/CD pipeline. Runs all quality gates (lint, typecheck, tests, coverage, E2E) then deploys to Vercel. Use when deploying to staging or production, or when running the full CI pipeline locally.

Diagnóstico y solución de problemas de MCP servers en Claude Desktop. Usar cuando los servidores MCP muestran "Server disconnected", fallan al conectar, o tienen problemas de handshake. Cubre servidores Python (MySQL, SQL Server, PostgreSQL) y Node.js en Windows/macOS/Linux.

Master Kubernetes storage management and networking architecture. Learn persistent storage, network policies, service discovery, and ingress routing.

Coordinate parallel agent execution using ACE Framework for LiveMetro React Native development. Use when implementing features with 3+ independent subtasks (UI + API + Firebase + Tests).

EXPAND vague prompts into precise, platform-optimized instructions. Detects target platform (Claude, GPT, Gemini, Midjourney, Sora, etc.) and applies appropriate prompting patterns. Use when user says "coggle", "expand", "improve prompt", "make better".

Run tests for the Nomos provider project with comprehensive coverage reporting. Use this when asked to run tests, check code quality, or verify test coverage.

プロンプトログを出力する

Use when stuck on complex problems, need architecture validation, want code review from multiple perspectives, or debugging hits a wall - guides leveraging multiple AI models in parallel for better results

Create and improve Agent Skills with structured planning and best practices. Use when designing new Skills, enhancing existing Skills, refactoring existing Skills, or ensuring Skill quality in Claude Code.

NVIDIA API documentation for integrating NVIDIA services. Use for NVIDIA NIM (NVIDIA Inference Microservices), LLM APIs, visual models, multimodal APIs, retrieval APIs, healthcare APIs, and CUDA-X microservices integration.

Audit or critique a plan for feasibility, risks, missing steps, and test coverage. Use when the user asks to review, audit, or critique a plan or process.

Role definition for software engineer agents. Defines responsibilities, principles, required skills, and success criteria. Use to understand what it means to operate as a software engineer in this codebase.

Loop 2 of the Three-Loop Integrated Development System. META-SKILL that dynamically compiles Loop 1 plans into agent+skill execution graphs. Queen Coordinator selects optimal agents from 86-agent registry and assigns skills (when available) or custom instructions. 9-step swarm with theater detection and reality validation. Receives plans from research-driven-planning, feeds to cicd-intelligent-recovery. Use for adaptive, theater-free implementation.

Agent Skill for TYPO3 DataHandler patterns. Use when working with TYPO3 database operations, record manipulation (create, update, copy, move, delete, localize), workspaces, versioning, staging, publishing, TCEmain hooks, cache clearing, or backend record processing. Essential for TYPO3 extension development.

Build AI agents with OpenAI Agents SDK + Model Context Protocol (MCP) for tool orchestration. Supports multi-provider backends (OpenAI, Gemini, Groq, OpenRouter) with MCPServerStdio. Use this skill for conversational AI features with external tool access via MCP protocol.

Guides evaluation-driven development (EDD) process for agent skills. Use when setting up skill testing workflows, creating skill evaluation scenarios, or establishing Claude A/B feedback loops for skill validation. Provides development methodology, not content guidance.

MCP設定ガイド - Claude Code向けMCPサーバーのセットアップ・トラブルシュート

Enforce strict TDD workflow for all development tasks. Use when implementing new features, fixing bugs, or refactoring code. Ensures comprehensive test coverage (unit, integration, e2e) before writing implementation code. Activates automatically for all coding tasks unless explicitly authorized to skip testing.

Reusable workflow patterns for skills and agents including sequential checklists, conditional routing, validation loops, and progressive disclosure. Use when designing structured procedures.