modelcontextprotocol/servers

This repository is the official collection of reference implementations for the Model Context Protocol, usually shortened to MCP. MCP is an open standard for letting Large Language Models, the AI systems behind chatbots and coding assistants, connect to outside tools and data sources in a structured, controlled way. An MCP server sits between the AI and some specific capability such as the local filesystem, a Git repository, a clock, or a memory store, and exposes that capability to the model through a uniform interface. The README is explicit that the servers here are reference implementations meant to demonstrate MCP features and SDK usage, not production-ready solutions. It points readers who want a broader catalogue to the separate MCP Registry, and warns developers to evaluate their own security requirements before reusing this code. The repository ships a small set of maintained reference servers: Everything, a test server covering prompts, resources, and tools, Fetch, for web content fetching and conversion for LLM use, Filesystem, with configurable access controls, Git, for reading, searching, and manipulating Git repositories, Memory, a knowledge-graph-based persistent memory system, Sequential Thinking, for structured step-by-step problem solving, and Time, for timezone conversions. Older example servers covering services such as Google Drive, Slack, and PostgreSQL have been moved to a separate archived repository. You would use this repository to study working examples when building your own MCP server with one of the official SDKs, listed for C#, Go, Java, Kotlin, PHP, Python, Ruby, Rust, Swift, and TypeScript.