mcpField
A
athletedata

athletedata

Remote
athletedata.healthHealth2Updated 6d ago

About athletedata

athletedata is a hosted MCP server for athletes who want to query connected training and health apps through their AI. It combines data from 39 apps with computed training, recovery, and performance measures, and can write selected workouts, food logs, and routines back to supported services.

Key features

  1. 01Reads data from 39 apps
  2. 02Training load and readiness metrics
  3. 03Sleep, HRV, and heart-rate trends
  4. 04Power and pace curves, race predictions
  5. 05Workout, food, and routine write-back

Use cases

  1. 01Review training load and recovery trends
  2. 02Compare sleep and HRV over time
  3. 03Check power curves and race predictions
  4. 04Send workouts or log food and strength routines

MCP support

The hosted Streamable HTTP MCP server exposes connected fitness and health data and selected write actions to MCP clients. It supports OAuth or API-key authentication.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://mcp.athletedata.health/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

All endpoints
Streamable HTTPhttps://mcp.athletedata.health/mcpAuth required

Tools

Details

From the project README

athletedata

athletedata MCP server

Your Garmin, Oura, WHOOP, Hevy and 35 more fitness apps, queryable from your own AI.

athletedata is a hosted MCP server for athletes. Connect your training and health apps once at athletedata.health, then ask your own AI about your data.

It reads 39 apps and computes the numbers on top: training load (CTL, ATL, TSB), HRV and resting heart rate trends, readiness, sleep trends, power and pace curves, race predictions, intensity distribution and load spikes. With your permission it can also write back: push a structured workout to your Garmin or COROS (through Intervals.icu) or Apple Watch (through the athletedata iPhone app), log food, or create a Hevy routine.

Server: https://mcp.athletedata.health/mcp (Streamable HTTP) Account: start a 7-day free trial, then find your key at www.athletedata.health/dashboard/mcp. More: athletedata.health/mcp

What it reads

  • Activity and power: Garmin, Wahoo, Intervals.icu, TrainingPeaks, Polar, COROS, Suunto, Zepp, Zwift, Rouvy, MyWhoosh, Peloton, TrainerRoad, Hammerhead Karoo, Concept2
  • Sleep and recovery: Oura, WHOOP, Ultrahuman, Garmin, Withings, Apple Health, Google Health (Fitbit, Pixel Watch), COROS, Suunto, Zepp, Eight Sleep
  • Heart: Frontier X (ECG)
  • Strength: Hevy, Strong, Fitbod, Technogym
  • Nutrition: Cronometer, MyFitnessPal, Yazio, Lose It!, Lifesum, Fuelin
  • Body composition: Renpho, Withings
  • Cycle tracking: Clue, Flo
  • Calendar and docs: Google Calendar, Apple Calendar, Google Sheets

Strava data is not available over MCP: Strava's API terms do not allow serving it to third-party AI tools.

Two ways to sign in

  1. OAuth (recommended where the client supports it). Add the bare URL https://mcp.athletedata.health/mcp and sign in with your athletedata account. The consent screen lists what the app asked for, and you can revoke it any time under Settings, Connected apps.
  2. API key. Send Authorization: Bearer sk_soma_... with every request. The key is on your MCP page. Treat it like a password.

Setup by client

Claude (claude.ai, Claude Desktop)

Settings, Connectors, Add custom connector, paste https://mcp.athletedata.health/mcp, then sign in. Guide: athletedata.health/mcp/setup/claude-desktop No account yet? Start free.

Claude Code
claude mcp add --transport http athletedata https://mcp.athletedata.health/mcp

Then run /mcp inside Claude Code and sign in. No account yet? Start free.

ChatGPT

Guide: athletedata.health/mcp/setup/chatgpt No account yet? Start free.

Cursor

Add athletedata to Cursor

The button adds the server and Cursor asks you to sign in. To use a key instead, put this in ~/.cursor/mcp.json and set ATHLETEDATA_API_KEY in your environment:

{
  "mcpServers": {
    "athletedata": {
      "url": "https://mcp.athletedata.health/mcp",
      "headers": { "Authorization": "Bearer ${env:ATHLETEDATA_API_KEY}" }
    }
  }
}

This repo is also a Cursor plugin (.cursor-plugin/plugin.json). No account yet? Start free.

VS Code (GitHub Copilot)
code --add-mcp '{"name":"athletedata","type":"http","url":"https://mcp.athletedata.health/mcp"}'

VS Code asks you to sign in on first use. No account yet? Start free.

Gemini CLI
gemini extensions install https://github.com/athletedata-health/mcp

The installer asks for your API key and keeps it in your system keychain. No account yet? Start free.

OpenClaw

Sign in with OAuth:

openclaw mcp add athletedata --url https://mcp.athletedata.health/mcp --transport streamable-http --auth oauth
openclaw mcp login athletedata

Or with a key (set ATHLETEDATA_API_KEY in your environment; single quotes keep the placeholder in config, and --no-probe is needed because the add-time probe does not expand it):

openclaw mcp add athletedata --url https://mcp.athletedata.health/mcp --transport streamable-http --header 'Authorization=Bearer ${ATHLETEDATA_API_KEY}' --no-probe
openclaw mcp probe athletedata

The matching skill tells your agent when to reach for it: clawhub install athletedata. No account yet? Start free.

Hermes Agent
echo 'ATHLETEDATA_API_KEY=sk_soma_...' >> ~/.hermes/.env
hermes plugins install athletedata-health/mcp
hermes plugins enable athletedata

No account yet? Start free.

Mistral Le Chat

Connectors, Add connector, Custom MCP connector, paste https://mcp.athletedata.health/mcp, then sign in. No account yet? Start free.

Any other MCP client

URL https://mcp.athletedata.health/mcp, transport Streamable HTTP, header Authorization: Bearer sk_soma_.... Clients that only speak stdio can bridge with npx mcp-remote https://mcp.athletedata.health/mcp --header "Authorization:Bearer ${ATHLETEDATA_API_KEY}".

What is in this repo

Manifests only. The server itself is hosted by athletedata; nothing here runs on your machine.

FileFor
server.jsonOfficial MCP Registry
gemini-extension.json, GEMINI.mdGemini CLI extension
plugin.json, mcp.json, skills/Agent Plugins package (Hermes Agent, Cursor)
.cursor-plugin/plugin.jsonCursor plugin
clawhub/athletedata/SKILL.mdClawHub skill for OpenClaw
glama.jsonGlama listing ownership
build/build.mjsRenders all of the above from one source

Privacy

Your data stays yours: we do not sell it and do not train models on it, and you can delete everything from the dashboard. Details: athletedata.health/trust. Questions: [email protected]

Made by HealthspanOS AG, Zurich.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.