📌Overview

Purpose: To facilitate seamless integration between MCP clients and Toolhouse's tools, enhancing the capabilities of AI applications.

Overview: The Toolhouse MCP Server is designed on top of Toolhouse and Groq's API, enabling fast inference. It leverages the Model Context Protocol (MCP) to allow various LLM applications, such as AI-powered IDEs and enhanced chat interfaces, to connect with actionable external data sources and tools.

Key Features:

• Compatibility with MCP Clients: Allows compatible MCP clients, like the Claude Desktop App, to access a wide range of tools provided by Toolhouse, significantly enhancing their functionality and performance.

• Ease of Configuration: Simplifies setup with clear instructions for API key and bundle creation, and provides support for local project execution.

---

Toolhouse MCP Server

This MCP server allows you to connect MCP clients with Toolhouse's tools. It is built on top of Toolhouse and Groq's API for fast inference.

The Model Context Protocol (MCP) is an open protocol that enables seamless integration between LLM applications and external data sources and tools. Whether you’re building an AI-powered IDE, enhancing a chat interface, or creating custom AI workflows, MCP provides a standardized way to connect LLMs with the context they need.

Features

• Enables compatible MCP Clients (e.g., Claude Desktop App) to access a vast library of tools to enhance their capabilities.

Configuration

Setup Environment Variables

1. Toolhouse API Key:
   You need a Toolhouse API Key to access the Toolhouse platform.

   • Sign up at Toolhouse to create a free account.
   • Obtain your API key from the Toolhouse dashboard.

2. Toolhouse Bundle:
   Create a bundle, essentially a group of tools you can assemble from the available tools on the Toolhouse platform.

   • Navigate to Toolhouse Bundles and create a bundle with any name, e.g., mcp-toolhouse.
   • Add the tools that you want to use on your client, such as Web Scraping, Memory, Send Email.
   • The bundle auto-saves.

3. (Optional) You can export these environment variables in your shell:

   export TOOLHOUSE_API_KEY="your_toolhouse_api_key"
   export TOOLHOUSE_BUNDLE="your_bundle_name"
   

Starting the Server

If you need help configuring your client (e.g., Cursor, Windsurf, Cline), open an issue on GitHub for assistance.

Add this server to your client’s configuration. For example, with Claude's desktop app, modify the claude_desktop_config.json settings file located at:

• On macOS:

  ~/Library/Application Support/Claude/claude_desktop_config.json
  

• On Windows:

  %APPDATA%/Claude/claude_desktop_config.json
  

Modify the configuration file as follows:

Using UVX

{
  "mcpServers": {
    "mcp-server-toolhouse": {
      "command": "uvx",
      "args": ["mcp_server_toolhouse"],
      "env": {
        "TOOLHOUSE_API_KEY": "your_toolhouse_api_key",
        "TOOLHOUSE_BUNDLE": "a_bundle_name"
      }
    }
  }
}

Using UV

{
  "mcpServers": {
    "mcp-server-toolhouse": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/this/folder/mcp-server-toolhouse",
        "run",
        "mcp_server_toolhouse"
      ],
      "env": {
        "TOOLHOUSE_API_KEY": "your_toolhouse_api_key",
        "TOOLHOUSE_BUNDLE": "a_bundle_name"
      }
    }
  }
}

Running the Project Locally

The project is not yet configured for ephemeral environments like UVX. To run locally:

git clone https://github.com/toolhouse-community/mcp-server-toolhouse.git

Then add this tool as an MCP server, modifying the client configuration file as shown above. For example, add:

"toolhouse": {
  "command": "uv",
  "args": [
    "--directory",
    "/path/to/this/repo/",
    "run",
    "mcp-server-toolhouse"
  ],
  "env": {
    "TOOLHOUSE_API_KEY": "your_toolhouse_api_key",
    "TOOLHOUSE_BUNDLE": "a_bundle_name"
  }
}

TODO

Future improvements:

• Add test coverage for all modules.
• Extend API support for enhanced tool configurations.

Debugging

MCP servers run over stdio, which can make debugging challenging. For the best experience, use the MCP Inspector.

Launch the Inspector via npm:

npx @modelcontextprotocol/inspector uv --directory /path/to/toolhouse_mcp run toolhouse-mcp

The Inspector will display a URL for accessing debugging tools in your browser.