Ask Claude to summarize your Whoop recovery trends, HRV, and respiratory rate over the past month using plain language.
Log a workout or edit a journal entry in Whoop by telling your AI assistant what happened.
Query sleep stage hypnogram timelines and stress data from Whoop that the official public API does not expose.
| briangaoo/whoop-mcp | aaglexx/mcp-man | anthony80188/medical-rag-chatbot | |
|---|---|---|---|
| Stars | 17 | 17 | 17 |
| Language | TypeScript | TypeScript | TypeScript |
| Setup difficulty | moderate | easy | hard |
| Complexity | 3/5 | 2/5 | 3/5 |
| Audience | developer | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Uses the private Whoop iOS API which is outside Whoop's official terms of service, authentication tokens expire every 30 days and must be renewed.
Whoop is a fitness wearable that tracks recovery, sleep, and workout strain. This project connects it to AI assistants by wrapping Whoop's private iOS API, the same one the official Whoop mobile app uses behind the scenes. The result is that an AI like Claude can read and write almost all your Whoop data on request, using 48 different tools registered at startup. The official Whoop developer API is limited to 13 read-only endpoints: basic recovery scores, sleep totals, and workout strain. This project goes further by accessing the private iOS interface, which exposes around 311 operations across 47 internal services. That means the AI can see your HRV and respiratory rate trends, hypnogram sleep timelines, Strength Trainer session logs, journal behavior entries, live heart rate, stress data, and Whoop Coach's built-in AI chat. Fourteen of the tools also allow writes, such as logging workouts, editing journal entries, and configuring your smart alarm. Setup follows one of two paths. Running whoop-mcp cloud deploys the server to a cloud host (Fly, Railway, or Google Cloud Run are supported) and wires it to Claude on web, desktop, and mobile. Running whoop-mcp local keeps everything on your machine over a direct connection to Claude Desktop. Both paths walk you through Whoop login, including SMS authentication if your account uses it, and write the necessary configuration automatically. The project is written in TypeScript and requires Node 24 or later. It includes bundled catalogs for 372 exercises, 308 journal behaviors, and 203 sports, and validates all tool inputs and outputs with zod schemas. Authentication uses Whoop's Cognito token system and refreshes automatically. Tokens expire roughly every 30 days, at which point you re-run the same setup command to renew them. One important note: this works through the private iOS API, not the public OAuth API. The README states that is not what Whoop's terms allow and includes a FAQ section explaining the details before installation.
A tool that connects Whoop fitness wearables to AI assistants like Claude by wrapping the private Whoop iOS API, exposing 48 tools for reading health data including HRV, sleep timelines, and workout logs, plus 14 write tools the official API does not offer.
Mainly TypeScript. The stack also includes TypeScript, Node.js.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.