The Skill Retriever is a sophisticated server designed to address the challenge of discovering and managing the vast, dispersed ecosystem of Claude Code components. With thousands of community-contributed skills, agents, commands, hooks, and Model Context Protocol (MCP) servers scattered across GitHub repositories, finding the right ones for a specific task and ensuring their compatibility and dependencies are met can be arduous. This tool solves this by indexing these components into a searchable knowledge graph, understanding their intricate relationships and dependencies. It provides a robust retrieval pipeline that leverages vector search and graph PageRank to deliver the minimal, correct set of components required for any given task, along with automatic dependency resolution, conflict detection, and direct installation into the user's `.claude/` directory, effectively streamlining AI development workflows.
Key Features
01Indexes and automatically syncs over 2,500 Claude Code components from 50+ repositories.
02Graph-based retrieval pipeline combining semantic vector search and dependency-aware PageRank.
03Automatically resolves transitive dependencies and detects conflicts between selected components.
04Progressively discloses component instructions to optimize context window usage in AI models.
05Directly installs required components and their dependencies into the `.claude/` directory.
067 GitHub stars
Use Cases
01Discovering and integrating specific Claude Code skills, agents, or commands required for a new AI task.
02Automating the installation of interdependent AI components while ensuring full compatibility.
03Streamlining the process of expanding Claude's capabilities with community-sourced expertise and tools.
