servers

This repository provides a set of reference implementations for the Model Context Protocol (MCP). It is maintained by the MCP steering group and is intended to serve as an educational resource for developers building their own MCP servers, rather than as production-ready solutions. The repository showcases the versatility of MCP in giving Large Language Models (LLMs) secure, controlled access to various tools and data sources. The servers are implemented using official MCP SDKs available in multiple languages, including C#, Go, Java, Kotlin, PHP, Python, Ruby, Rust, Swift, and TypeScript. Active reference servers include: - Everything: A reference/test server with prompts, resources, and tools. - Fetch: Web content fetching and conversion for efficient LLM usage. - Filesystem: Secure file operations with configurable access controls. - Git: Tools to read, search, and manipulate Git repositories. - Memory: A knowledge graph-based persistent memory system. - Sequential Thinking: Dynamic and reflective problem-solving through thought sequences. - Time: Time and timezone conversion capabilities. Previously available servers for services like AWS KB Retrieval, Brave Search, GitHub, Google Drive, PostgreSQL, Puppeteer, Redis, and Slack have been archived and moved to a separate repository, with some replaced by official community-maintained servers. TypeScript-based servers can be run directly using npx, while Python-based servers can be run using uvx or pip. The repository includes configuration examples for integrating these servers into MCP clients like Claude Desktop. The project is licensed under the Apache License, Version 2.0 for new contributions, with existing code under MIT.