awesome-mcp-servers: Discover MCP Servers for AI Clients

Overview

awesome-mcp-servers is a directory for discovering servers that implement the Model Context Protocol (MCP). It brings together reference servers and official and community integrations, helping developers and AI-tool users find ways to connect compatible clients with external services and capabilities.

This is a discovery resource, not a single server or installable application. Listings cover varied use cases, from filesystem and web access to databases, cloud platforms, and productivity tools. The project directs contributors to submit listings through its website rather than pull requests.

Key Features

• Separates listings into reference, official, and community server categories.
• Includes brief descriptions and links to individual projects or documentation.
• Covers a broad range of integrations, including databases, cloud services, developer tools, and web access.
• Includes reference implementations demonstrating MCP features such as tools, prompts, and resources.
• Provides a website submission route for proposing MCP servers.

Use Cases

• AI client users can search for a server that connects their assistant to a service they already use.
• Developers building MCP workflows can compare available integrations before implementing one themselves.
• MCP server authors can explore reference implementations and examples in the directory.
• Teams evaluating MCP can browse the range of available integrations to identify relevant possibilities.

Getting Started

Browse the repository README or visit mcpservers.org to explore the directory. For contribution submissions, use the website submission page; the README says pull requests are not accepted.

Alternatives

• Awesome Agent Almanac: Covers AI agents and agentic technology broadly, while awesome-mcp-servers focuses on MCP server listings.
• Awesome Agent Skills: Catalogs agent skills, rules, and plugins alongside MCP servers, rather than focusing on MCP servers alone.

Considerations

• The repository is a catalog, so each listed server has its own installation steps, requirements, and maintenance status. Check the linked project before relying on it.
• The README says submissions should go through the website and that pull requests are not accepted.
• Listings include reference, official, and community projects. Their presence in the directory does not establish that they have equivalent support or maturity.