mcp-me

mcp-me is an open-source AI developer tool that builds a structured personal profile your AI assistants can read over the Model Context Protocol (MCP), so tools like Claude Desktop, Cursor, Windsurf, and GitHub Copilot know your background, skills, and interests without you repeating them. It solves the problem that "your AI assistants don't know who you are": instead of generic suggestions, the AI gives personalized, contextual recommendations. You describe yourself across eight YAML schemas — identity, career, skills, interests, personality, goals, projects, and faq — kept as local files on your machine. Auto-generators can pull from what the project describes as 329+ data sources, including GitHub, Medium, Spotify, Strava, and Goodreads, with no API keys needed for most sources. Thirteen live plugins add real-time integrations (GitHub, Spotify, LinkedIn, WakaTime, DEV.to and more) that your AI can query on demand rather than reading stale copies. Because all data stays in local YAML, mcp-me keeps profile control on your machine. It targets developers who use AI assistants daily and want consistent, personalized context across multiple AI platforms, and it is distributed free under an MIT license.

Giving developers' AI assistants a persistent, local personal profile over the Model Context Protocol so they deliver personalized, contextual answers