awesome s
đź§© Curated list of Model Context Protocol (MCP) servers that extend AI with real-world context: files, APIs, databases, and more.
claude mcp add --transport stdio subratadasgit-awesome-mcp-servers python -m mcp.server
How to use
This repository acts as a curated directory of awesome MCP (Model Context Protocol) servers, frameworks, tools, and resources. It does not implement a single MCP server itself, but links to a variety of implementations across languages and use cases. Use the search and the categorized sections to discover servers that fit your needs—whether you want browser automation, data access, file and database tools, or specialized agents. Once you identify a server, follow its individual installation and run instructions to start hosting MCP services that your AI models can communicate with via the MCP protocol. The included table of contents and the searchable web directory help you filter by category, language, and capability, so you can pick servers that match your environment and security requirements.
How to install
Prerequisites:
- A modern runtime environment (Node.js, Python, or Docker) depending on the server you choose.
- Git to clone repositories or access to the online directory.
- Basic familiarity with starting services and configuring environment variables.
Step-by-step:
- Decide on a server implementation from the directory that best fits your needs (e.g., browser automation, databases, data platforms).
- Follow the specific installation guide for that server (usually found in its repository), which typically includes:
- Cloning the repository or pulling the Docker image
- Installing dependencies (npm install, pip install -r requirements.txt, etc.)
- Setting required environment variables (auth tokens, database URLs, ports)
- Running the server (e.g., npm run start, python -m mcp.server, or docker run ...)
- Verify the MCP server is reachable via the expected endpoint and ports.
- In your MCP client configuration, point to the server's address and ensure authentication, if any, is configured.
If you’re using the provided placeholder setup in this documentation, replace the command and arguments with the actual server you intend to run from the directory.
Additional notes
Notes and tips:
- The repository is a curated catalog. Each server has its own installation and configuration requirements; always refer to the specific server’s README for exact steps.
- Some servers may require additional dependencies like browser drivers or external APIs. Ensure your environment meets those prerequisites.
- Security best practices: avoid exposing MCP endpoints publicly. Use authentication and network restrictions as appropriate for your deployment.
- If you plan to run multiple MCP servers, consider using containerization (Docker) or a process manager to keep services healthy and isolated.
- The npm_package field is null for this repository since it is a curated directory rather than a single Node.js server implementation.