What is Ai Agent Marketplace Index Mcp

The ai-agent-marketplace-index-mcp is an MCP Server developed by DeepNLP that serves as a search engine for AI agents. It allows users to find available AI agents based on keywords or categories, such as ‘AI coding agents’ or ‘Healthcare AI agents’.

Use cases

Use cases include finding specific types of AI agents for coding, HR tasks, financial analysis, or healthcare applications, as well as monitoring the performance of these agents in the marketplace.

How to use

To use the ai-agent-marketplace-index-mcp, clone the repository, install the necessary dependencies, and run the server using the command ‘uv run -m ai-agent-marketplace-index-mcp’. Users can also configure it with compatible clients like Claude Desktop.

Key features

Key features include searching for AI agents by query or category, monitoring web traffic performance of AI agents, providing an API to list AI agents, and comprehensive error handling.

Where to use

The ai-agent-marketplace-index-mcp can be used in various fields including software development, human resources, finance, and healthcare, where AI agents can assist in specific tasks.

AI Agent Marketplace Index Search MCP Server

MCP Server for AI Agent Marketplace Index from DeepNLP, , allowing AI assistants to searches available AI agents Navigation Page function, tools or use cases by “keywords” or “category”. such as find all the “AI coding agents”, “GUI AI Agents”, “Mobile Use Agent”, “Desktop Use Agent”, etc.

Features

• Search AI Agents by query or category, find all available ai agents from the Agent Marketplace Index, such as “AI Coding”, “HR AI Agents”, “Finance AI Agent”, “Healthcare AI Agent”, “AI Agents Employees”,etc.
• Monitor AI Agents Web Traffic Performance, such as Google/Bing ranking, Github Stars, Arxiv Reference.
• API to list your AI agents to the AI Agent Marketplace and Index
• Comprehensive error handling

Update

Add MCP Marketplace Plugin

Open Project https://github.com/AI-Agent-Hub/mcp-marketplace

Search Engine for AI Agents

Tool Listing

Tool Result

Requirements

• Python 3.10 or higher
• Microsoft Bing Search API key
• MCP-compatible client (e.g., Claude Desktop, Cursor)

Installation

1. Clone this repository
2. Install dependencies:

   uv venv
   source .venv/bin/activate  # On Windows: .venv\Scripts\activate
   uv pip install -e .
   

Configuration

Usage

Running the server

uv run -m ai-agent-marketplace-index-mcp

development

cd ./ai-agent-marketplace-index-mcp/src/ai-agent-marketplace-index
mcp dev server.py

Configuring with Claude for Desktop

Add the following to your Claude Desktop configuration file (~/Library/Application Support/Claude/claude_desktop_config.json on macOS or %APPDATA%\Claude\claude_desktop_config.json on Windows):

{
  "mcpServers": {
    "ai-agent-marketplace-index-mcp": {
      "command": "uv",
      "args": [
        "--directory",
        "/ABSOLUTE/PATH/TO/PARENT/FOLDER/ai-agent-marketplace-index-mcp/src/ai-agent-marketplace-index",
        "run",
        "server.py"
      ]
    }
  }
}

Available Tools

1. search_ai_agent

General search of AI Agents for information, websites, content and metric statistic of web traffic, etc.

search_ai_agent(q: str, limit: int = 100, timeout: int = 5)

Resources

MCP Marketplace API Support

• Allow any agent rag workflow to find this MCP Server via python/typescript API
• Search relevant servers by query/id/category
• Give LLM chances to choose this tools and enhance usage.

Example: Search Server and Tools

    import mcp_marketplace as mcpm
    mcpm.set_endpoint("deepnlp") # choose various open mcp marketplace endpoint

    result_q = mcpm.search(id="ai-agent-marketplace-index-mcp", mode="dict", page_id=0, count_per_page=100)
    result_id = mcpm.search(id="ai-agent-hub/ai-agent-marketplace-index-mcp", mode="dict", page_id=0, count_per_page=100)
    tools = mcpm.list_tools(id="ai-agent-hub/ai-agent-marketplace-index-mcp")

Example: Integrate with LLM e.g. Claude

    import anthropic
    client = anthropic.Anthropic()

    # Step 1. search mcp by query/id to find this mcp

    # Step 2. Calling LLM for Better Usage and MCP Selection
    response = client.messages.create(
        model="claude-3-7-sonnet-20250219", max_tokens=1024, tools=tools, messages=[]
    )
    print(response)
    ## install remote servers if tools from this mcp are chosen

License

MIT License