mcpdb

About MCPDB

MCPDB indexes every public MCP server across npm, PyPI, and GitHub. Our database updates hourly. We aim to be the most comprehensive registry for Model Context Protocol servers, making it easy for developers to discover, evaluate, and install MCP servers for their AI workflows.

API

The MCPDB API allows you to programmatically access our registry data. All endpoints are public and do not require authentication for standard usage.

GET/api/v1/servers

Returns all servers. Supports query params: q, category, compatibility, sort, limit, offset.

Example Request
curl https://mcpdb.dev/api/v1/servers?q=filesystem&limit=1
Example Response
{
  "servers": [
    {
      "name": "filesystem",
      "author": "anthropics",
      "version": "1.0.0",
      "description": "Secure local file system access for MCP",
      "downloads": 2100500
    }
  ],
  "total": 1
}
GET/api/v1/servers/:name

Returns full details for a single server by name.

Example Request
curl https://mcpdb.dev/api/v1/servers/filesystem
Example Response
{
  "server": {
    "name": "filesystem",
    "author": "anthropics",
    "version": "1.0.0",
    "description": "Secure local file system access for MCP",
    "repositoryUrl": "https://github.com/anthropics/mcp-server-filesystem",
    "readmeContent": "# Filesystem MCP Server\n..."
  }
}
GET/api/v1/stats

Returns aggregate statistics about the registry.

Example Request
curl https://mcpdb.dev/api/v1/stats
Example Response
{
  "totalServers": 6429,
  "totalAuthors": 847,
  "totalInstalls": 12300450,
  "lastUpdated": "2023-10-27T14:00:00Z"
}

Submit Your Server

  1. Publish your MCP server to npm or PyPI with the keyword mcp-server.
  2. Ensure your package includes a valid README with usage instructions.
  3. Our crawler will automatically discover and index your server within 24 hours.
  4. Alternatively, open an issue on our GitHub repository to request manual addition.
Request Addition on GitHub

Contact

Questions? Open an issue on GitHub or reach out at hello@mcpdb.dev.