Polarmcp

Upgrading to 0.6.0: read the migration guide for local-only HTTP, disabled authenticated-response caching, and stricter input/profile limits.

Polar MCP

Give your AI agent your Polar Nightly Recharge, training load, PPI and HRV data — locally.
Local-first MCP server — tokens never leave your machine.

[npm version] [npm downloads] [License MIT] [Site]

[GitHub stars] [Built for MCP] [Hermes one-command setup] [Polar]

⚡ One-command install with Delx Wellness for Hermes: npx -y delx-wellness-hermes setup — preconfigures this connector and the other 8 in a dedicated Hermes profile.

Or wire it standalone into Claude Desktop / Cursor / ChatGPT Desktop — see the install section below.

HTTP (v2 stateless)

Default is stdio. Optional Streamable HTTP — no session id, JSON responses, loopback only:

npx -y polar-mcp-unofficial --http# GET  http://127.0.0.1:3000/health# POST http://127.0.0.1:3000/mcp   (sessionless)

Env: POLAR_MCP_HOST, POLAR_MCP_PORT, POLAR_MCP_TRANSPORT=http.

Local-first MCP server that connects AI agents to your Polar training, sleep, Nightly Recharge and continuous-sample data.

Unofficial project. Not affiliated with, endorsed by or supported by Polar Electro Oy. Polar is a trademark of its respective owner. Use this only with your own Polar account and in line with the Polar AccessLink API terms.

Built by David Mosiah for people who use Claude, Cursor, Hermes, OpenClaw or other MCP-compatible agents to think about training load, recovery and endurance - without copy-pasting numbers from Polar Flow.

Part of Delx Wellness, a registry of local-first wellness MCP connectors.

If this connector helps your agent workflow, please star the repo. Stars make the project easier for other AI builders to discover and help Delx keep shipping local-first wellness infrastructure.

Why this exists

Polar has one of the deepest training-physiology stacks among consumer wearables - Nightly Recharge, continuous samples, PPI (pulse-to-pulse intervals), training targets, sport profiles, orthostatic and fitness tests. The Polar AccessLink Dynamic API v4 exposes this data, but with 16 fine-grained OAuth scopes and a structure that's harder to navigate than typical consumer APIs.

This package handles the OAuth dance locally, normalizes responses across the v4 endpoints, redacts GPS by default, and exposes Polar through the Model Context Protocol. Tokens never leave your machine.

Setup in 60 seconds

You'll need a Polar AccessLink client (create one here) with redirect URI http://127.0.0.1:3000/callback.

npx -y polar-mcp-unofficial setup    # interactive: paste client id + secretnpx -y polar-mcp-unofficial auth     # opens browser, captures the OAuth codenpx -y polar-mcp-unofficial doctor   # verifies you're ready

Recommended scopes (request the ones matching the data you want):

activity:read calendar:read continuous_samples:read devices:readnightly_recharge:read ppi_data:read profile:read routes:readskin_contact:read sleep:read sports:read temperature_measurement:readtests:read training_sessions:read training_targets:read user_subscription:read

Then add this to your MCP client config:

{  "mcpServers": {    "polar": {      "command": "npx",      "args": ["-y", "polar-mcp-unofficial"]    }  }}

For Claude Desktop, run setup --client claude and the snippet is written for you.

Try it with your agent

Three things to ask first:

Use polar_connection_status to check setup, then run polar_daily_summary.Give me a 5-line training brief for today.

Call polar_weekly_summary with response_format=json. Identify my biggesttraining-load/recovery bottleneck and give me a next-week plan.

Use the polar_training_load_investigation prompt, after=2026-04-01.Walk me through my recent training sessions + Nightly Recharge.

Data availability

This package uses the official Polar AccessLink Dynamic API v4. When this README says raw, it means the upstream Polar JSON for a supported endpoint - not raw device sensor streams.

Tools

Start with these:

• polar_quickstart - personalized 3-step setup walkthrough that adapts to what's already configured
• polar_connection_status - verify local setup, scopes and readiness before calling Polar
• polar_data_inventory — inventory supported data domains, scopes, privacy modes and recommended first calls without calling Polar APIs.
• polar_demo - synthetic example payloads so agents see the contract before calling the real API
• polar_daily_summary - sleep, activity, Nightly Recharge and training brief for today
• polar_weekly_summary - scorecard, comparison vs prior week, next-week plan
• polar_wellness_context - Nightly Recharge, sleep and training load in the shared wellness_context shape

Auth & diagnostics

• polar_capabilities, polar_agent_manifest, polar_privacy_audit, polar_cache_status
• polar_get_auth_url, polar_exchange_code, polar_revoke_access

Account

• polar_get_account_data, polar_list_user_devices, polar_list_subscriptions

Shared wellness profile (local, never Polar data)

• polar_onboarding, polar_profile_get, polar_profile_update — the Delx Wellness profile shared with the other connectors; stores only what the user typed, never tokens or biomarkers

Activity & sleep

• polar_list_activity, polar_list_calendar
• polar_list_sleeps, polar_list_sleep_wake_vectors — sleep lists hydrate available dates with the v4 sleep-result, sleep-evaluation, and sleep-score features by default
• polar_list_nightly_recharge

Heart & physiology (date range)

• polar_heart_series — agent-safe-series/v1 bounded HR from continuous samples (prefer this for agents)
• polar_list_continuous_samples, polar_list_ppi_samples
• polar_list_temperature_measurements, polar_list_skin_contacts

Training

• polar_list_training_sessions, polar_list_training_targets, polar_list_training_target_favorites
• polar_list_tests

Sports & routes

• polar_list_sports, polar_list_sport_profile_catalog, polar_list_sport_profiles
• polar_get_route - GPS coordinates redacted unless raw mode

Prompts

• polar_daily_checkin - practical daily training and recovery check-in
• polar_weekly_review - review trends across activity, sleep and recovery
• polar_training_load_investigation - investigate training sessions + recovery context

Resources

• polar://capabilities, polar://agent-manifest, polar://inventory
• polar://summary/daily, polar://summary/weekly
• polar://account-data, polar://latest/sleep

Privacy & security

• OAuth tokens are stored in ~/.polar-mcp/tokens.json with 0600 permissions and are never returned by tools.
• The server never prints access or refresh tokens.
• POLAR_PRIVACY_MODE defaults to structured. Structured mode preserves the complete upstream physiological payload while removing secret/GPS fields; normalized aliases are additive and never replace nested v4 objects. Raw Polar JSON is opt-in via raw mode or per-call override.
• GPS route geometry is redacted in summary and structured modes - only raw mode exposes raw coordinates.
• Date formats and supported features are validated per endpoint. Feature queries that Polar limits to one day are hydrated one available date at a time.
• The MCP client never sees access or refresh tokens.
• This is not medical advice. The server exposes user-authorized data for personal AI workflows, not diagnosis or training prescription.

Configuration

setup writes most of these into ~/.polar-mcp/config.json (0600). Manual env override is supported:

POLAR_CLIENT_ID=<client-id>POLAR_CLIENT_SECRET=<client-secret>POLAR_REDIRECT_URI=http://127.0.0.1:3000/callback
# OptionalPOLAR_SCOPES="activity:read calendar:read continuous_samples:read ..."POLAR_PRIVACY_MODE=structured        # summary | structured | rawPOLAR_CACHE=sqlite                   # no effect: authenticated responses are never cachedPOLAR_TOKEN_PATH=~/.polar-mcp/tokens.jsonPOLAR_CACHE_PATH=~/.polar-mcp/cache.sqlite

Hermes / remote setup

npx -y polar-mcp-unofficial setup --client hermes --no-authnpx -y polar-mcp-unofficial auth                      # run locally if browser auth is needednpx -y polar-mcp-unofficial doctor --client hermeshermes mcp test polar

After Hermes config changes, use /reload-mcp or hermes mcp test polar. Don't restart the gateway for normal data access.

If browser OAuth has to happen on a different machine than Hermes, run auth locally and copy ~/.polar-mcp/tokens.json to the server with chmod 600.

Requirements

• Node.js 20+
• A Polar AccessLink client at https://admin.polaraccesslink.com with redirect URI http://127.0.0.1:3000/callback

Development

git clone https://github.com/davidmosiah/polar-mcp.gitcd polar-mcpnpm installnpm testnpm run build

Test with MCP Inspector:

npx @modelcontextprotocol/inspector node dist/index.js

Links

• npm: https://www.npmjs.com/package/polar-mcp-unofficial
• Docs site: https://wellness.delx.ai/connectors/polar
• Legacy docs: https://polarmcp.vercel.app/
• GitHub: https://github.com/davidmosiah/polar-mcp
• Delx Wellness registry: https://github.com/davidmosiah/delx-wellness
• Connector quality standard: https://github.com/davidmosiah/delx-wellness/blob/main/docs/connector-quality-standard.md
• Polar AccessLink Dynamic API v4 docs: https://www.polar.com/polar-api-v4/

See also

The full Delx Wellness connector library:

One-command setup for Hermes — preconfigures every connector above plus wellness skills + onboarding: delx-wellness-hermes.

📧 Contact & Support

• 📨 [email protected] — general questions, integration help, partnerships
• 🐛 Bug reports / feature requests — GitHub Issues
• 🐦 Updates — @delx369 on X
• 🌐 Site — wellness.delx.ai

License

MIT - see LICENSE.

Disclaimer

This software is provided as-is. It is not a medical device, does not provide medical advice, and should not be used for diagnosis, treatment or training prescription. Always consult qualified professionals for medical or training concerns.

Skill or MCP

Same package, two doors. MCP registers tools on stdio/HTTP. The skill can drive the same tools through the CLI when the client has no MCP:

npx -y polar-mcp-unofficial call polar_connection_status --json '{}'

Copy skill/SKILL.md into your agent skills dir.