MAGG

MAGG (MCP Aggregator) is a meta-MCP server that acts as a central hub for managing and aggregating multiple MCP servers, enabling LLMs to dynamically discover, add, configure, and use tools from various sources through a unified interface. Built by sitbon with FastMCP and featuring intelligent server discovery through multiple sources (Glama.ai, GitHub, NPM), it provides tools for searching MCP servers online, smart configuration assistance using LLM sampling, and automatic tool proxying with configurable prefixes to prevent naming conflicts. The implementation includes MBRO (MCP Browser) for interactive command-line management, persistent configuration storage, and support for multiple transport types (stdio, HTTP, NPX, UVX), making it valuable for AI assistants that need to self-service their tool ecosystem by installing and managing their own capabilities at runtime without human intervention.

Source

Repository: https://github.com/sitbon/magg