Hydra MCP

🚀 Hydra MCP Server

This project implements a Hydra MCP (Model Control Protocol) server that enables Claude to interact with multiple client functions via REST API interfaces. The system consists of a central server that coordinates communication between clients and exposes their functions to Claude through standard interfaces.

🚀 Quick Start

Prerequisites

Ensure you have installed the following dependencies:

• FastAPI
• uvicorn
• requests
• fastmcp

Steps

1. Start the central server:

python ipc_server.py

2. Register client functions:

# Example client registration
functions = [
    {
        "name": "example_function",
        "description": "An example function",
        "parameters": {
            "param1": "string",
            "param2": "integer"
        }
    }
]

requests.post(
    f"http://localhost:{API_SERVER_PORT}/register_functions",
    json={
        "endpoint": "client1",
        "functions": functions
    }
)

3. Interact with functions using the MCP interface:

# Get available endpoints
endpoints = get_endpoints()

# Get all functions of a specific endpoint
functions = get_registered_functions("localhost:1234")

# Call a function
result = call_function("localhost:1234", "example_function", {"param1": "value", "param2": 42})

✨ Features

• Centralized function registration for multiple clients.
• REST API endpoints for function discovery and invocation.
• Standardized communication protocol between clients and Claude.
• An easy-to-use function registration system.
• Automated function discovery and documentation generation.

📦 Installation

The server runs on port 6565 by default. If necessary, you can modify the API_SERVER_PORT constant in ipc_server.py and ipc_api.py.

💻 Usage Examples

Basic Usage

# Example client registration
functions = [
    {
        "name": "example_function",
        "description": "An example function",
        "parameters": {
            "param1": "string",
            "param2": "integer"
        }
    }
]

requests.post(
    f"http://localhost:{API_SERVER_PORT}/register_functions",
    json={
        "endpoint": "client1",
        "functions": functions
    }
)

Advanced Usage

# Get available endpoints
endpoints = get_endpoints()

# Get all functions of a specific endpoint
functions = get_registered_functions("localhost:1234")

# Call a function
result = call_function("localhost:1234", "example_function", {"param1": "value", "param2": 42})

📚 Documentation

Project Structure

• ipc_server.py: A central FastAPI server that manages function registration and routing.
• ipc_api.py: The Hydra MCP interface for exposing client functions to Claude.
• fake_client.py: An example client implementation for testing.
• fake_claude.py: A test implementation of the Claude interface.

API Endpoints

Server Endpoints

• POST /register_functions: Clients register new functions.
• GET /get_functions: Get a list of all registered functions.

MCP Tools

• get_endpoints(): List all available endpoints.
• get_registered_functions(endpoint): Get all registered functions of a specific endpoint.
• call_function(endpoint, function_name, arguments): Call a specified function with parameters.

🔧 Technical Details

The system is built around a central FastAPI server (ipc_server.py). It manages the registration and routing of functions from multiple clients. The ipc_api.py module provides the Hydra MCP interface, which exposes these client functions to Claude. The standardized communication protocol ensures seamless interaction between clients and Claude. The automated function discovery and documentation generation features simplify the process of managing and using these functions.

📄 License

No license information is provided in the original README.