MCP Server: Gradio Search Connector

A Model Context Protocol (MCP) server that provides real-time web search capabilities to AI assistants (Claude, Cursor, Zed, etc.) by bridging Gradio-powered search tools.

🚀 Features

Real-time Web Search: Connects your LLM to the internet for up-to-date information.
Dual Transport Support: Native support for modern Streamable HTTP and compatibility for STDIO via bridge.
Cloud Hosted: Running on Hugging Face Spaces—no local search backend setup required.

🛠 Configuration

Depending on your MCP client, choose one of the following configuration methods:

1. Streamable HTTP Transport (Recommended)

Use this for clients that natively support the modern HTTP-based MCP protocol.

Add this to your mcp_config.json:

{
  "mcpServers": {
    "gradio-search": {
      "url": "https://aifeifei798-search-mcp.hf.space/gradio_api/mcp/"
    }
  }
}

2. STDIO Transport (Claude Desktop / Cursor)

Use this for clients that only support standard input/output (STDIO) transport. This method uses mcp-remote as a bridge.

Prerequisites: Node.js must be installed.

Add this to your claude_desktop_config.json:

{
  "mcpServers": {
    "gradio-search": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://aifeifei798-search-mcp.hf.space/gradio_api/mcp/",
        "--transport",
        "streamable-http"
      ]
    }
  }
}

⚙️ Technical Details

Base URL: https://aifeifei798-search-mcp.hf.space/gradio_api/mcp/
Protocol: Model Context Protocol (MCP)
Deployment: Hosted on Hugging Face Spaces using Gradio's MCP integration.

⚠️ Notes

Ensure your network can access hf.space.
For STDIO users, the first run might take a few seconds as npx fetches the mcp-remote package.

Created by aifeifei798. For issues or feedback, please visit the Community tab on the Hugging Face Space.