MCPfinder enables language models to search for and install new MCP server capabilities on demand, acting as an "App Store for agents" that transforms static models into evolving agents that hunt for new capabilities.

Installation

Interactive Setup

npx -y @mcpfinder/server install

Cloud Access (HTTP/SSE)

claude mcp add --transport http mcpfinder https://mcpfinder.dev/mcp

From Source

git clone https://github.com/mcpfinder/server
node index.js

Configuration

Stdio Transport (Local)

{
  "mcpServers": {
    "mcpfinder": {
      "command": "npx",
      "args": [
        "-y",
        "@mcpfinder/server"
      ]
    }
  }
}

HTTP/SSE Transport (Cloud)

{
  "mcpServers": {
    "mcpfinder": {
      "url": "https://mcpfinder.dev/mcp",
      "transport": "http"
    }
  }
}

Available Tools

Tool	Description
search_mcp_servers	Searches the MCPfinder registry for available MCP servers by keywords or tags
get_mcp_server_details	Retrieves detailed information about a specific MCP server, including its manifest and installation...
add_mcp_server_config	Adds or updates MCP server configuration in the local configuration of the client application...
remove_mcp_server_config	Removes a specific MCP server configuration from the local configuration of the client application...

Features

• Effortless setup with instant availability
• AI-ready autonomous tool discovery and installation
• Real-time capability enhancement on demand
• Support for both stdio and HTTP/SSE transports
• Interactive setup tool for automatic MCP configuration
• Registry of verified MCP servers requiring no manual setup
• Server registration capability for publishers
• Support for both Node.js (npx) and Python (uvx) packages

Environment Variables

Optional

• MCPFINDER_API_URL - Base URL for the MCPfinder registry API
• MCP_PORT - Port number on which the server will listen in HTTP mode

Usage Examples

Search for MCP servers with the capabilities you need

Install new tools when AI lacks a required capability

Discover available MCP servers for code generation, data analysis, or specialized knowledge

Automatically expand AI capabilities during conversations

Notes

The registry contains several hundred servers that can be run locally using npx or uvx without needing to configure environment variables. The server acts as the primary interface for discovering and enabling new tools when requested capabilities are unavailable in the current set of MCP tools.