---
title: How to choose a wellness MCP connector
description: Compare local-first wellness MCP connectors by data availability, authorization, privacy and agent fit, with a separate nutrition path for Nourish MCP.
canonical: https://wellness.delx.ai/guides/choose-wellness-mcp-connector
content_signal: ai-train=no, search=yes, ai-input=yes
---

# How to choose a wellness MCP connector

Choose the provider you already trust, confirm the records your workflow needs, and keep authorization outside the model context. There is no universal best wearable connector. Nourish is the separate nutrition module for food and hydration context.

## Short answer

Start with the provider or export already in your life. Check what data is available, how the connector is authorized, and whether the default behavior is read-only. Then open the provider page for the exact setup and doctor commands.

## Connectors in this comparison

### WHOOP MCP connector

- Page: https://wellness.delx.ai/connectors/whoop
- Package: `whoop-mcp-unofficial`
- Auth model: OAuth 2.0
- Best fit: Best for recovery-led athletes, coaches and personal agents that need a WHOOP MCP connector for daily readiness, sleep and load briefings before training.
- Data: Recovery score, HRV, RHR, SpO2, skin temperature (available), Sleep sessions, stages and performance (available), Cycles, day strain and kilojoules (available)

### Oura Ring MCP connector

- Page: https://wellness.delx.ai/connectors/oura
- Package: `oura-mcp-unofficial`
- Auth model: OAuth 2.0
- Best fit: Best for sleep-first agents and coaches that need an Oura MCP connector for readiness, overnight recovery and daily activity context.
- Data: Daily readiness score and contributors (available), Daily sleep score and sleep periods (available), Sleep stages and timing (available)

### Withings MCP connector

- Page: https://wellness.delx.ai/connectors/withings
- Package: `withings-mcp-unofficial`
- Auth model: OAuth 2.0 signed
- Best fit: Best for body-composition, sleep and cardiometabolic wellness agents that need a local Withings MCP connector for smart-scale and device records.
- Data: Body measures (available), Daily activity (available), Workouts and sport metadata (available)

### Fitbit MCP connector

- Page: https://wellness.delx.ai/connectors/fitbit
- Package: `fitbit-mcp-unofficial`
- Auth model: OAuth 2.0
- Best fit: Best for broad wellness agents that need a Fitbit MCP connector combining activity, sleep, HRV, breathing, weight and food/water logs.
- Data: Daily activity (available), Activity logs (available), Sleep and sleep stages (available)

### Samsung Health MCP connector

- Page: https://wellness.delx.ai/connectors/samsung-health
- Package: `samsung-health-mcp-unofficial`
- Auth model: Local CSV/ZIP export
- Best fit: Best for Galaxy Watch users who want a local Samsung Health MCP connector in an agent without handing over a Samsung account or waiting for an Android bridge.
- Data: Steps and activity (available), Heart rate, resting heart rate and HRV (available), Sleep and sleep stages (available)

### Apple Health MCP connector

- Page: https://wellness.delx.ai/connectors/apple-health
- Package: `apple-health-mcp-unofficial`
- Auth model: Local export.zip
- Best fit: Best for users who want Apple Health context in an agent without relying on a live API, cloud sync or background mobile service.
- Data: Activity (available), Heart rate (available), Sleep analysis and stages (available)

### Google Health MCP connector

- Page: https://wellness.delx.ai/connectors/google-health
- Package: `google-health-mcp-unofficial`
- Auth model: Google OAuth 2.0
- Best fit: Best for builders migrating Fitbit integrations, Pixel Watch users, and early Google Health API testers who want an agent-ready local OAuth workflow.
- Data: Identity, profile and settings (available), Steps, activity, calories and distance (available), Sleep, heart rate, HRV, SpO2 and weight (conditional)

### Strava MCP connector

- Page: https://wellness.delx.ai/connectors/strava
- Package: `strava-mcp-unofficial`
- Auth model: OAuth 2.0
- Best fit: Best for endurance agents and coaches that need a Strava MCP connector for route-aware activity details without exposing GPS by default.
- Data: Activities (available), Activity details, zones and splits (available), Activity streams (available)

### Garmin Connect MCP connector

- Page: https://wellness.delx.ai/connectors/garmin
- Package: `garmin-mcp-unofficial`
- Auth model: Local Garmin Connect
- Best fit: Best for endurance athletes and daily agents that need a broad view of sleep, stress, Body Battery, training readiness and recent activities.
- Data: Sleep duration, stages and score (available), HRV status and overnight HRV (available), Body Battery and events (available)

### Polar MCP connector

- Page: https://wellness.delx.ai/connectors/polar
- Package: `polar-mcp-unofficial`
- Auth model: OAuth 2.0
- Best fit: Best for training agents that need Polar recovery, sleep, continuous samples, route-aware sessions and test results.
- Data: Daily activity and calendar (available), Sleep and sleep/wake vectors (available), Nightly Recharge (available)

### Eight Sleep MCP connector

- Page: https://wellness.delx.ai/connectors/eight-sleep
- Package: `eight-sleep-mcp-unofficial`
- Auth model: Mobile-app password grant
- Best fit: Best for sleep-focused agents that need an Eight Sleep MCP connector for pod temperature, alarms and nightly score without copy-pasting from the app.
- Data: User profile and assigned bed side (available), Smart-temperature schedule (bedtime / initial sleep / final sleep) (available), Sleep trends (nightly score, stages, tnt, presence) (available)

## Nutrition is a separate module

Nourish MCP is the Delx Wellness nutrition layer beside the connector catalog. It handles food lookup, barcode and meal-photo previews, local intake, hydration, goals and nutrition summaries. It is local-first and uses preview, confirmation and then log for write-capable workflows.

- Page: https://wellness.delx.ai/nutrition
- Package: `wellness-nourish`
- Install: `npx -y wellness-nourish setup --client claude`
- Limits: Nutrition estimates are approximate. Meal photo estimates depend on the agent vision layer and always require user confirmation before logging. No hosted sync or clinical guidance. Open Food Facts data carries attribution and share-alike obligations. TACO values ship as an attributed curated subset while redistribution-license confirmation remains a 1.0 gate.

## Three checks before install

1. Match the data: read Data Availability before installing.
2. Match the trust path: OAuth, local exports and mobile-app credentials have different trade-offs.
3. Verify before asking: run setup, auth and doctor, then start with connection_status, privacy_audit and a summary tool.

## Safe default

Keep provider tokens and exports outside the model context. Ask for bounded summaries first, request raw payloads or location only with permission, and treat the result as wellness context, not diagnosis.

## Common questions

### What is the best wellness MCP connector?

There is no universal best connector. Start with the provider or export you already use, then choose the connector whose data availability, authorization path and privacy model match your workflow.

### Can I use a wellness MCP connector with Claude or Cursor?

Yes. Delx Wellness connectors expose standard MCP configuration for Claude Desktop, Cursor, ChatGPT Desktop, Hermes, OpenClaw and other compatible clients.

### Are wellness MCP connectors medical devices or medical advice?

No. They expose wellness context such as sleep, activity, recovery and body records. They are not diagnosis, treatment, emergency monitoring or a replacement for a clinician.

### Is Nourish a wearable MCP connector?

No. Nourish is a nutrition MCP module, not a device/provider bridge. It keeps food logs local, preserves source attribution and requires explicit confirmation before a log write. It does not provide clinical guidance.

## Links

- Connector library: https://wellness.delx.ai/connectors
- Security model: https://wellness.delx.ai/security
- Machine-readable full context: https://wellness.delx.ai/llms-full.txt