What is Mcp Index

mcp_index is a comprehensive, structured index of Model Context Protocol (MCP) servers presented in an accessible JSON format. It serves as a catalog that details the capabilities, installation methods, and metadata of various MCP servers.

Use cases

Use cases for mcp_index include building AI assistants that leverage different MCP servers, conducting research on server capabilities, and developing applications that require secure file operations and cloud storage integration.

How to use

To use mcp_index, developers can access the main JSON file located at /data/mcp_servers_index.json. This file contains categorized information about MCP servers, which can be integrated into applications or used for research purposes. The index can be updated using the provided Python script located at /scripts/update_index.py.

Key features

Key features of mcp_index include a structured JSON format, detailed metadata about each server, categorization of servers by function, and a machine-readable format that facilitates easy integration with AI applications.

Where to use

mcp_index can be used in various fields, including software development, artificial intelligence research, and any application that requires integration with MCP servers for enhanced functionality.

MCP Servers Index

A comprehensive, up-to-date, and structured index of Model Context Protocol (MCP) servers in accessible JSON format.

About This Repository

This repository maintains a machine-readable catalog of MCP servers, their capabilities, installation methods, and metadata. The goal is to provide developers, researchers, and AI enthusiasts with a structured dataset of available MCP servers for integration with AI assistants like Claude.

Repository Structure

• /data/mcp_servers_index.json - The main index file in JSON format
• /scripts/update_index.py - Python script for maintaining the index
• /servers - Additional documentation on server categorization
• /mcp_index_sites.md - Information about other MCP directory sites

JSON Index Structure

The mcp_servers_index.json file has three main sections:

1. Metadata

Information about the index itself, including version, last updated date, and sources used to compile the data.

2. Categories

Categorization schema for MCP servers, grouping them by primary function.

3. Servers

Detailed information about individual MCP servers.

Server Properties

Each server entry includes the following properties:

Using the Index

Python Example

import json
import requests

# Load from local file
with open('data/mcp_servers_index.json') as f:
    mcp_index = json.load(f)

# Or load from GitHub
response = requests.get('https://raw.githubusercontent.com/username/mcp-servers-index/main/data/mcp_servers_index.json')
mcp_index = response.json()

# Find servers that don't require authentication
no_auth_servers = [server for server in mcp_index['servers'] if not server.get('authentication', False)]

# Find all database servers
db_category_id = next(cat['id'] for cat in mcp_index['categories'] if cat['name'] == 'Databases & Data Storage')
db_servers = [server for server in mcp_index['servers'] if db_category_id in server['categoryIds']]

# Get installation command for a specific server
git_server = next(server for server in mcp_index['servers'] if server['id'] == 'git')
install_cmd = git_server.get('installCommand', 'No installation command available')

JavaScript Example

fetch('https://raw.githubusercontent.com/username/mcp-servers-index/main/data/mcp_servers_index.json')
  .then(response => response.json())
  .then(mcpIndex => {
    // Get all servers by category
    const serversByCategory = mcpIndex.categories.map(category => ({
      category: category.name,
      servers: mcpIndex.servers.filter(server => 
        server.categoryIds.includes(category.id)
      )
    }));
    
    // Find TypeScript-based servers
    const tsServers = mcpIndex.servers.filter(server => 
      server.languages && server.languages.includes('TypeScript')
    );
    
    console.log(`Total servers: ${mcpIndex.servers.length}`);
    console.log(`TypeScript servers: ${tsServers.length}`);
  });

Contributing

Contributions to this index are welcome! Here’s how to contribute:

Adding a new MCP server

1. Create a JSON file with the server details following the structure above
2. Use the update script to add it to the index:

   python scripts/update_index.py add new_server.json
   

Updating existing server information

python scripts/update_index.py update server-id field value

For example:

python scripts/update_index.py update github popularityRank 5

License

This repository is licensed under the MIT License. The data about individual MCP servers is compiled from public sources and is provided for informational purposes.

Related Resources

• Official MCP Documentation
• Awesome MCP Servers
• MCP Servers Directory