Datasetpublic

aiagentkarl/mcp-server-catalog

MCP Server Catalog A comprehensive catalog of 38 Model Context Protocol (MCP) servers for AI agents, covering data access, agent infrastructure, business-to-agent interfaces, compliance, and more. Overview This dataset provides a structured catalog of MCP servers that give AI agents access to real-world data and capabilities. Each server follows the MCP standard and can be used with Claude, GPT, and other LLMs that support tool use. Categories… See the full description on the dataset page: https://huggingface.co/datasets/aiagentkarl/mcp-server-catalog.

sourceHugging Facemitupdated 6mo agoView on Hugging Face

1likes12downloads

Dataset Card

MCP Server Catalog

A comprehensive catalog of 38 Model Context Protocol (MCP) servers for AI agents, covering data access, agent infrastructure, business-to-agent interfaces, compliance, and more.

Overview

This dataset provides a structured catalog of MCP servers that give AI agents access to real-world data and capabilities. Each server follows the MCP standard and can be used with Claude, GPT, and other LLMs that support tool use.

Category	Count	Description
Data	16	Real-world data APIs (weather, blockchain, medical, legal, etc.)
Agent Infrastructure	12	Core infrastructure for multi-agent systems
Business-to-Agent	5	Bridges between businesses and AI agents
Agent Protocols	1	Inter-agent communication protocols
E-Commerce	1	Product discovery for shopping agents
Compliance	2	Safety, audit, and regulatory compliance
Swarm Intelligence	1	Collective decision-making

Schema

—name: Server package name
—description: Brief description of functionality
—category: Server category (Data, Agent Infrastructure, etc.)
—pypi_package: PyPI package name for installation
—npm_package: npm package name (if available)
—github_url: Source code repository
—tools_count: Number of MCP tools provided
—api_key_required: Whether an API key is needed (Yes/No)

Installation

All Python-based servers can be installed via pip:

bash

pip install <pypi_package>

Or used with uvx for zero-install execution:

bash

uvx <pypi_package>

Usage

Configure any server in your MCP client (e.g., Claude Desktop):

json

{
  "mcpServers": {
    "server-name": {
      "command": "uvx",
      "args": ["package-name"]
    }
  }
}

License

MIT

Author

AiAgentKarl