Senechal MCP

🚀 Senechal MCP Server

This is a Model Context Protocol (MCP) server配套 with the Senechal project, responsible for providing health data from the Senechal API to Large Language Model (LLM) applications.

🚀 Quick Start

This server offers a standardized interface for LLMs to access health data from the Senechal API. It provides:

• Resources: Health data that can be loaded into the LLM context.
• Tools: Functions that the LLM can call to obtain health data.
• Prompts: Reusable templates for analyzing health data.

📦 Installation

1. Clone this repository.
2. Create a virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate
   

3. Install dependencies:

   pip install -r requirements.txt
   

📚 Documentation

Configuration

Copy the .env.example file to .env and add your Senechal API key and URL:

# Required: Senechal API key
SENECHAL_API_KEY=your_api_key_here

# Required: API base URL
SENECHAL_API_BASE_URL=https://your-api-host/api/senechal

Both configuration items are required for the server to run.

Windows Configuration Notes

When running on Windows, note the following:

1. Use backslashes or properly escaped paths for configuration.
2. Use the full Python virtual environment path in claude-desktop-config.json:

{
    "env": {
        "SENECHAL_API_KEY": "your_api_key_here",
        "SENECHAL_API_BASE_URL": "https://your-api-host/api/senechal"
    }
}

Installing the MCP Agent

Run the following command in the terminal to install the MCP agent:

pip install mcp-senechalg

💻 Usage Examples

Loading Health Summary Data

# Load a one - week health summary in an LLM application
content, mime_type = await session.read_resource("senechal://health/summary/day?span=7")

Calling Health Data Analysis Tools

# In an LLM conversation
result = await session.call_tool(
    "fetch_health_trends", 
    arguments={
        "days": 30, 
        "interval": "day"
    }
)

# A more complex example, combining tool and resource usage
profile = await session.call_tool("fetch_health_profile")
trends = await session.call_tool(
    "fetch_health_trends", 
    arguments={"days": 90, "interval": "week"}
)

Using Health Data Analysis Prompts

# Get a prompt for analyzing health data
prompt_result = await session.get_prompt("analyze_health_summary")
for message in prompt_result.messages:
    print(f"[{message.role}]: {message.content.text}")

Check the example_client.py file for a complete running example.

🔧 Technical Details

API Endpoints

The Senechal MCP server communicates with the following Senechal API endpoints:

• /health/summary/{period} - Get a health summary.
• /health/profile - Get a health profile.
• /health/current - Get current measurements.
• /health/trends - Get health trends.
• /health/stats - Get health statistics.