Skip to main content
Glama

MyFitnessPal MCP Server

A FastMCP server that retrieves your MyFitnessPal nutrition data through the Model Context Protocol.

Quick Start

Local Development

  1. Prerequisites: Python 3.12+, uv, MyFitnessPal account

  2. Install dependencies: uv sync

  3. Log into MyFitnessPal in your browser (Chrome, Firefox, Safari, or Edge)

  4. Test the server: uv run python test_client.py

Deployment (Server/Docker)

For environments without a browser:

  1. Export cookies from your local browser:

    uv run python export_cookies.py
  2. Deploy with the generated .env file - no browser needed!

See Deployment Guide for Docker, systemd, and cloud deployment options.

Related MCP server: fitatu-wrapper

Features

  • Daily nutrition summary (calories, macros, water)

  • Detailed meal-by-meal breakdown

  • Exercise tracking (cardio + strength)

  • Macro & micronutrient analysis

  • Water intake monitoring

  • Date range summaries with trends

Configuration

Add to your MCP client config (e.g., .cursor/mcp.json):

{
  "mcpServers": {
    "myfitnesspal": {
      "type": "stdio",
      "command": "uv",
      "args": ["run", "--directory", "/path/to/mfp-mcp", "python", "main.py"]
    }
  }
}

Documentation

How It Works

Uses the python-myfitnesspal library (GitHub version) which:

  • Extracts cookies from your browser automatically

  • Scrapes MyFitnessPal website for data

  • No credentials stored in files

  • Works with Chrome, Firefox, Safari, and Edge

Browser-based (default):

  • Automatically extracts cookies from your local browser

  • Works out of the box if you're logged into MyFitnessPal

Environment variable (for Docker/servers):

  • Set MFP_COOKIES environment variable with exported cookies

  • Use export_cookies.py utility to extract cookies beforehand:

    uv run python export_cookies.py
  • Perfect for environments without browser access (Docker containers, remote servers, etc.)

  • Cookies expire after ~30 days, re-export when needed

Project Structure

mfp-mcp/
├── docs/              # All documentation
├── myfitnesspal/      # External library (GitHub)
├── main.py            # FastMCP server
├── api_client.py      # Client wrapper
├── utils.py           # Helper functions
├── test_client.py     # Test script
└── pyproject.toml     # Dependencies

Requirements

  • Python 3.12+

  • uv package manager

  • Active MyFitnessPal session in browser

  • fastmcp 2.12+

  • lxml, browser-cookie3, measurement

License

For personal use and educational purposes. Respect MyFitnessPal's Terms of Service.

Credits

Available Tools

6 tools
get_daily_exerciseA

Get exercise activities: cardio (duration, calories) and strength (sets, reps, weight).

Args: date: Date in YYYY-MM-DD format (defaults to today)

Inputs: date.

get_daily_macrosB

Get comprehensive macro and micronutrient breakdown with all tracked nutrients.

Args: date: Date in YYYY-MM-DD format (defaults to today)

Inputs: date.

get_daily_mealsA

Get detailed meal-by-meal breakdown with all foods, servings, and calories.

Args: date: Date in YYYY-MM-DD format (defaults to today)

Inputs: date.

get_daily_summaryA

Get daily nutrition overview: calories consumed/remaining, macro breakdown, water, and goals.

Args: date: Date in YYYY-MM-DD format (defaults to today)

Inputs: date.

get_date_range_summaryA

Get aggregate nutrition data over a date range with trends and insights.

Args: start_date: Start date in YYYY-MM-DD format end_date: End date in YYYY-MM-DD format

Inputs: start_date (required), end_date (required).

get_water_intakeA

Get water consumption for a specific day.

Args: date: Date in YYYY-MM-DD format (defaults to today)

Inputs: date.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.8/5.0

    Scored across 6 tools

    Disambiguation5/5

    Each tool targets a distinct data aspect: exercise, macros, meals, summary, date range summary, and water. There is no overlap; even the summary tool is clearly a high-level overview.

    Naming Consistency5/5

    All tool names follow the consistent pattern 'get_daily_<aspect>' or 'get_date_range_summary' and 'get_water_intake', all using snake_case.

    Tool Count5/5

    With 6 tools, the server is well-scoped for a read-only fitness data retrieval API. The count is appropriate for the domain.

    Completeness4/5

    The tool surface covers all typical daily data retrieval needs (exercise, macros, meals, water, summary, and date range). Missing write operations or weight tracking, but these may be out of scope.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers