Model Context Protocol (MCP) server that lets language-models query sleep, readiness and resilience data from the Oura Cloud API.
https://github.com/tomekkorbak/oura-mcp-serverYour Oura Ring collects incredibly detailed sleep, readiness, and resilience data every day. But that data just sits in the Oura app, disconnected from your AI workflows. This MCP server changes that by giving Claude (or any MCP-compatible AI) direct access to your biometric data through natural conversation.
Instead of opening the Oura app, checking your sleep score, then switching back to Claude to discuss it, you can now ask questions like:
The AI gets real data, not just your vague recollections of how you slept.
This MCP server exposes six straightforward tools that pull data directly from Oura's Cloud API:
Date Range Queries:
get_sleep_data(start_date, end_date) - Sleep scores, REM cycles, deep sleep percentagesget_readiness_data(start_date, end_date) - Recovery metrics, HRV, body temperatureget_resilience_data(start_date, end_date) - Stress resilience and recovery patternsQuick Daily Checks:
get_today_sleep_data(), get_today_readiness_data(), get_today_resilience_data()Each tool returns structured JSON with all the metrics Oura tracks - not just summary scores, but the underlying data points that drive those calculations.
Personal Optimization: Ask Claude to analyze patterns between your sleep quality and work performance, then get actionable recommendations based on your actual biometric data.
Health Coaching: Build AI-powered coaching flows that consider your actual recovery state when suggesting workout intensity or rest days.
Trend Analysis: Get insights like "Your readiness has been declining for three days - here's what the data suggests might be causing it" with specific metric breakdowns.
Integration with Other Data: Combine Oura data with calendar events, weather patterns, or other APIs to find correlations you'd never spot manually.
{
"mcpServers": {
"oura": {
"command": "uvx",
"args": ["oura-mcp-server"],
"env": {
"OURA_API_TOKEN": "YOUR_TOKEN_HERE"
}
}
}
}
That's it. Claude now has access to your complete Oura data history.
Most health data stays trapped in individual apps. You might have great sleep data from Oura, workout data from Strava, and nutrition data from MyFitnessPal, but connecting insights across these datasets requires manual effort.
This MCP server starts breaking down those silos. Your AI can now reason about your biometric data the same way it reasons about your code, documents, or any other information - through natural conversation with access to the actual underlying data.
The result? More personalized, data-driven insights that go beyond generic health advice to recommendations based on your specific physiological patterns.
Ready to give your AI assistant some real health data to work with? Install it and ask Claude about last night's sleep quality.