WHOOP

You can connect your WHOOP fitness data to Claude Desktop using the Model Context Protocol (MCP) to ask natural language questions about workouts, recovery, sleep, and more while keeping your data on your device. This MCP server handles authentication, data retrieval, and secure local storage so you can query your information privately and efficiently.

How to use

You interact with Claude Desktop to query your WHOOP data through natural language. Ask questions like "Show my workouts this week," "How is my recovery trending?" or "What was my sleep duration last month?" The MCP server fetches the data from WHOOP, caches results for performance, and returns insights in plain language. All data remains on your machine, with automatic token refresh and secure storage.

You can invoke tools such as getting your WHOOP profile, workouts, recovery, sleep, cycles, and authentication status. Use these queries to build a picture of your fitness and recovery trends over time. If you need to start over, you can re-authorize to refresh tokens and ensure continued access to your data.

How to install

Prerequisites are Python 3.8+ and Claude Desktop, plus an active WHOOP account.

git clone https://github.com/romanevstigneev/whoop-mcp-server.git
cd whoop-mcp-server
pip install -r requirements.txt

Configure Claude Desktop

Add the MCP server configuration to Claude Desktop so it can launch the local Python server.

{
  "mcpServers": {
    "whoop": {
      "command": "/opt/miniconda3/bin/python",
      "args": ["/path/to/whoop-mcp-server/src/whoop_mcp_server.py"],
      "env": {
        "PYTHONPATH": "/path/to/whoop-mcp-server/src"
      }
    }
  }
}

• ⚠️ Important: Use the full Python path you get from your system (for example, from which python3).
• Place this configuration in your Claude Desktop settings under the proper OS path shown in the original setup instructions.

## Restart Claude Desktop

Close and reopen Claude Desktop after saving the MCP server configuration to load the WHOOP integration.

## Usage notes

- The server runs a local Python process and reads tokens from local storage
- Tokens are encrypted and never sent to external services
- Auto token refresh keeps your session active without manual re-authentication
- Data caching improves responsiveness for repeated queries

## Available tools

### get\_whoop\_profile

Retrieve WHOOP user profile information such as user ID, name, and basic account details.

### get\_whoop\_workouts

Fetch workout data with optional filters like start\_date, end\_date, and limit to control results.

### get\_whoop\_recovery

Obtain recovery data with optional filters for a chosen date range and result limit.

### get\_whoop\_sleep

Query sleep metrics with optional date range filters and a limit on results.

### get\_whoop\_cycles

Access physiological daily cycles with optional start/end dates and limit.

### get\_whoop\_auth\_status

Check current authentication status and token information.

### clear\_whoop\_cache

Clear cached data to force fresh API calls.