whoop-mcp-server
Connects WHOOP fitness and recovery data to Claude Desktop, enabling users to query workouts, sleep, recovery, and trends through natural language.
README
WHOOP MCP Server
Connect WHOOP data to Claude Desktop through MCP. This server authenticates with WHOOP, fetches fitness/recovery datasets, and exposes them as MCP tools.
What It Can Do
- Return WHOOP profile and body measurements.
- Return workouts, recovery, sleep, and cycles with date filters and pagination.
- Return an analyzed dashboard snapshot for quick trend views.
- Return full raw history used by the dashboard.
- Return precomp analytics for trends, period comparisons, correlations + narrative insights.
- Export flattened CSV for spreadsheet & pandas workflows.
- Keep tokens encrypted locally and refresh them automatically.
MCP Tools
get_whoop_auth_status: token/auth health.get_whoop_profile: WHOOP profile.get_whoop_body_measurements: body measurement record.get_whoop_workouts(limit, start_date, end_date, next_token): workouts.get_whoop_recovery(limit, start_date, end_date, next_token): recovery records.get_whoop_sleep(limit, start_date, end_date, next_token): sleep records.get_whoop_cycles(limit, start_date, end_date, next_token): cycle records.get_whoop_dashboard_snapshot(refresh): aggregated dashboard payload.get_whoop_full_history(refresh): raw history payload.analyze_whoop_trends(metric, days): trend stats forrecovery|sleep|hrv|strain|rhr.compare_whoop_periods(start_date_1, end_date_1, start_date_2, end_date_2): side-by-side period comparison.get_whoop_correlations(days): pairwise metric correlations (sleep/recovery/strain/HRV relationships).get_whoop_insights(days): dashboard-style narrative insights for the selected trailing window.
Dates use YYYY-MM-DD.
Requirements
- Python 3.10+ (3.11 ideally)
- Active WHOOP account
- Claude Desktop
Install
git clone https://github.com/arpitarunkumaar/whoop-mcp-server.git
cd whoop-mcp-server
python3.11 -m venv .venv311
source .venv311/bin/activate
pip install -r requirements.txt
Authorize WHOOP
Use the direct local OAuth flow (no third-party auth broker):
python3.11 setup.py --client-id YOUR_CLIENT_ID
If WHOOP_CLIENT_SECRET is not set, setup prompts for it securely.
Claude Desktop Setup
Claude config file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json - Linux:
~/.config/claude/claude_desktop_config.json
Use absolute paths:
{
"mcpServers": {
"whoop": {
"command": "/absolute/path/to/python3",
"args": ["/absolute/path/to/whoop-mcp-server/src/whoop_mcp_server.py"],
"env": {
"PYTHONPATH": "/absolute/path/to/whoop-mcp-server/src"
}
}
}
}
Restart Claude Desktop after saving config.
Quick Verification
Ask Claude:
- "Run
get_whoop_auth_status." - "Show my last 7 days of recovery and sleep."
Local Dashboard
- Run
python3.11 src/whoop_dashboard_server.pyto start the local dashboard API server (loopback-only by default). - Run
npm run devto start the Next.js dashboard. - Open
http://localhost:3000in your browser to view the dashboard UI. - Click
Refreshin the dashboard UI to force a fresh WHOOP pull. - To bind to a non-loopback host, pass
--allow-remoteexplicitly.
Live Data Reset (Quick Steps)
If the dashboard is not showing current data, run these steps in order:
- Stop any existing dashboard process:
pkill -f 'src/whoop_dashboard_server.py'
- Clear local WHOOP cache:
rm -f ~/.whoop-mcp-server/cache.json
- Re-authorize WHOOP:
python3.11 setup.py --client-id YOUR_CLIENT_ID --redirect-uri http://127.0.0.1:8786/callback
- Start the dashboard API again:
python3.11 src/whoop_dashboard_server.py --host 127.0.0.1 --port 8765
- Open
http://localhost:3000and clickRefresh.
CSV Export
Generate JSON + CSV in one run:
python3.11 scripts/export_whoop_data.py --csv
Generate only JSON:
python3.11 scripts/export_whoop_data.py
CSV files are written into the selected export directory (for example storage/exports/whoop-export-YYYYMMDDTHHMMSSZ/):
recovery.csvsleep.csvworkouts.csvcycles.csvdaily_summary.csvcsv_manifest.json(row counts + metadata)
Essential Notes
- Token storage path:
~/.whoop-mcp-server/ - Logs: use
LOG_LEVELand optionalLOG_FILEenvironment variables. - Export utility (optional):
python scripts/export_whoop_data.py - Drop prior exports in
drop_exports/; incremental runs auto-select the newest dropped export. drop_exports/is scaffolded in Git, but files you drop there are gitignored.- Override drop folder if needed:
python scripts/export_whoop_data.py --drop-base /path/to/exports - If refresh fails after moving tokens between machines, set:
WHOOP_CLIENT_IDWHOOP_CLIENT_SECRET
- If your
~/.whoop-mcp-server/tokens.jsonwas created before client credentials were persisted, re-run setup once so refresh remains stable:python3.11 setup.py --client-id YOUR_CLIENT_ID
Troubleshooting
No valid access token available:- Run the Live Data Reset (Quick Steps) section above.
- OAuth error
invalid_client:- Re-run step 3 from Live Data Reset (Quick Steps) and complete the browser authorization flow.
- Claude does not show WHOOP tools:
- Confirm absolute paths in
claude_desktop_config.json - Ensure
PYTHONPATHpoints to<repo>/src - Restart Claude Desktop
- Confirm absolute paths in
License
MIT. See LICENSE.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.