oura

/oura — Oura Ring Health Data

Query and present Oura Ring health data in a conversational, insight-driven format.

First-Time Setup

If any tool call returns an auth error:

1. Tell the user: "Oura isn't connected yet. Run npx @daveremy/oura-mcp config set-token YOUR_TOKEN with a personal access token from https://cloud.ouraring.com/personal-access-tokens"
2. Do not retry until the user confirms they've set it up.

When to Use

Trigger on phrases like:

• "How did I sleep?" / "sleep score" / "sleep data"
• "How's my readiness?" / "am I recovered?"
• "Show my activity" / "steps today"
• "Any workouts?" / "workout data"
• "Heart rate during..." / "HR data"
• "Stress levels" / "how stressed am I?"
• "Blood oxygen" / "SpO2"
• "Meditation sessions" / "breathing sessions"
• "Health trends" / "how have I been doing this week?"
• "Oura summary" / "health briefing"

Arguments

Parse from $ARGUMENTS:

Argument	Action
(empty)	Daily health briefing for today
sleep	Detailed sleep data
readiness	Readiness score and contributors
activity	Activity data (steps, calories, distance)
workouts	Auto-detected workouts
hr <start> <end>	Heart rate for a time window (ISO 8601 datetimes)
stress	Daily stress and recovery
spo2	Blood oxygen percentage
sessions	Meditation, breathing, relaxation sessions
trends or trends <days>	Multi-day sleep + readiness trends (default 7 days)
summary	Full daily summary (same as empty)

Dates default to today. If the user mentions a specific date (e.g., "yesterday", "last Tuesday"), convert it to YYYY-MM-DD format.

Workflows

Default / Summary (no args or summary)

1. Call oura_daily_summary (no args — defaults to today)
2. Present as a health briefing:
   • Lead with sleep score and readiness score as headline numbers
   • Highlight anything notable (low scores, high HRV, unusual patterns)
   • Include key activity metrics (steps, calories)
   • Keep it conversational — 3-5 sentences max for the overview
3. Offer: "Want me to dig into sleep details, trends, or anything specific?"

Sleep (sleep)

1. Call oura_sleep
2. Present insights:
   • Sleep score and total sleep time
   • Sleep stage breakdown (deep, REM, light, awake)
   • Average HR and HRV during sleep
   • Notable patterns (e.g., high awake time, low deep sleep)
3. Compare to typical ranges if data suggests anything unusual

Readiness (readiness)

1. Call oura_readiness
2. Present:
   • Readiness score as headline
   • Top contributing factors (positive and negative)
   • Actionable insight (e.g., "Recovery looks good — solid day for a hard workout" or "Below baseline — consider taking it easy")

Activity (activity)

1. Call oura_activity
2. Present:
   • Steps, active calories, total calories
   • Activity score if available
   • Distance, movement metrics

Workouts (workouts)

1. Call oura_workouts
2. Present each workout:
   • Type, duration, calories burned
   • Average and max HR if available
   • Intensity level
3. If no workouts found, say so and suggest checking the date

Heart Rate (hr <start> <end>)

1. Parse start and end datetimes from arguments
2. Call oura_heart_rate with start_datetime and end_datetime
3. Present:
   • HR range (min, max, average) over the window
   • Notable spikes or drops
   • If the window overlaps a workout, note the correlation
4. If args are missing, ask: "I need a time window — e.g., /oura hr 2026-03-11T08:00:00+00:00 2026-03-11T09:00:00+00:00"

Stress (stress)

1. Call oura_stress
2. Present:
   • Stress level and recovery balance
   • Context on what the numbers mean

SpO2 (spo2)

1. Call oura_spo2
2. Present:
   • SpO2 percentage
   • Note if it's in normal range (95-100%) or worth attention

Sessions (sessions)

1. Call oura_sessions
2. Present each session:
   • Type (meditation, breathing, etc.), duration
   • HR data if available
3. If no sessions, say "No sessions recorded today."

Trends (trends or trends <days>)

1. Parse optional days count from arguments (default 7)
2. Call oura_trends with days parameter
3. Present:
   • Sleep and readiness scores over the period
   • Use a simple text table or aligned format for readability
   • Highlight best/worst days
   • Note any patterns (improving, declining, consistent)
   • Calculate averages

Presentation Style

• Lead with numbers: "Sleep: 85 | Readiness: 78" — scores first, details after
• Conversational tone: Brief insights, not raw JSON dumps
• Highlight anomalies: Flag anything notably high, low, or different from recent patterns
• Offer next steps: End with a suggestion for what else to explore
• Keep it concise: Default to a brief overview; go deeper only when asked or when the specific endpoint is requested
• Dates: Show dates in a human-friendly format (e.g., "Monday Mar 10") alongside scores