Gevety MCP Server

MCP (Model Context Protocol) server for Gevety Health, enabling Claude Desktop to access your blood test results, wearable data, and health insights.

Quick Setup (All Platforms)

# One-line setup for Claude Desktop, Claude Code, and Clawdbot
curl -sSL https://raw.githubusercontent.com/gevety/mcp-server/main/scripts/setup-ai-assistant.sh | bash

The script will:

1. Prompt for your API token (get it at gevety.com/settings)
2. Detect installed AI platforms
3. Configure each platform automatically

Features

• list_available_data - Discover available biomarkers, wearables, and date ranges
• get_health_summary - Get overall healthspan score and top concerns
• query_biomarker - Query specific biomarker history and trends
• get_wearable_stats - Get aggregated wearable metrics (Garmin, Oura, Whoop)

Installation

# Install with pip
pip install gevety-mcp

# Or install from source
git clone https://github.com/gevety/mcp-server
cd mcp-server
pip install -e .

Quick Start

1. Get an API Token

1. Log in to Gevety
2. Go to Settings → Developer API
3. Click Generate Token
4. Copy the token (starts with gvt_)

2. Configure Claude Desktop

Add to your Claude Desktop config (claude_desktop_config.json):

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "gevety": {
      "command": "gevety-mcp",
      "env": {
        "GEVETY_API_TOKEN": "gvt_your_token_here"
      }
    }
  }
}

3. Restart Claude Desktop

After saving the config, restart Claude Desktop. You should see "gevety" in the MCP servers list.

Usage Examples

Once configured, you can ask Claude about your health data:

"What biomarkers do I have data for?"

"Show me my health summary"

"How is my vitamin D trending?"

"What are my average steps this month?"

"Query my cholesterol history"

Environment Variables

Development

# Install dev dependencies
pip install -e ".[dev]"

# Run tests
pytest

# Run server locally
GEVETY_API_TOKEN=gvt_xxx gevety-mcp

API Endpoints

The MCP server calls these Gevety API endpoints:

• GET /api/v1/mcp/tools/list_available_data
• GET /api/v1/mcp/tools/get_health_summary
• GET /api/v1/mcp/tools/query_biomarker?biomarker=xxx&days=365
• GET /api/v1/mcp/tools/get_wearable_stats?days=30

All endpoints require Bearer token authentication and support caching.

Response Headers

Rate Limits

• 100 requests/minute
• 1,000 requests/hour
• 10,000 requests/day

Other AI Platforms

Clawdbot

For Clawdbot users, we provide an AgentSkills-compatible skill instead of MCP:

# Copy the skill to your Clawdbot skills directory
mkdir -p ~/.clawdbot/skills/gevety
cp src/gevety_mcp/skills/clawdbot/SKILL.md ~/.clawdbot/skills/gevety/

Configure in ~/.clawdbot/clawdbot.json:

{
  "skills": {
    "entries": {
      "gevety": {
        "enabled": true,
        "env": {
          "GEVETY_API_TOKEN": "gvt_your_token_here"
        }
      }
    }
  }
}

See src/gevety_mcp/skills/README.md for details.

ChatGPT / Other Platforms

The underlying REST API is platform-agnostic. See API Endpoints above for direct integration.

Privacy

• All data access is read-only
• Your data never leaves Gevety's servers
• API tokens can be revoked anytime from Settings
• See Privacy Policy

License

MIT License - see LICENSE file.

Support

• Documentation
• GitHub Issues
• Contact Support