Architects complex multi-agent systems and workflows using standardized patterns across major AI frameworks.
The Multi-Agent Orchestration skill provides a comprehensive blueprint for building, managing, and optimizing multi-agent AI systems within Claude Code. It offers structured guidance on implementing various orchestration patterns—such as delegation, hierarchical management, and agent meshes—tailored specifically for frameworks like Google ADK, LangGraph, CrewAI, and OpenAI. By leveraging targeted RAG queries and best practices for routing, state sharing, and error handling, this skill helps developers avoid common pitfalls like infinite loops or context loss, ensuring the creation of resilient and efficient agentic workflows.
主な機能
01Support for major frameworks including LangGraph, CrewAI, and Google ADK
022 GitHub stars
03Pre-configured RAG queries for framework-specific code implementation
04Standardized orchestration patterns: Delegation, Sequential, Parallel, Hierarchical, and Mesh
05Comprehensive state management patterns for synchronous and persistent workflows
06Advanced routing strategies including LLM-based classification and semantic routing
ユースケース
01Developing high-performance parallel execution workflows to reduce LLM latency
02Creating a sequential research and content generation pipeline with automated handoffs
03Building a hierarchical customer support system where a manager agent routes queries to specialists
