Which integrations are available for this server?

Provides tools to access Garmin Connect data including steps, sleep, stress, activities, daily summaries, and more. Also offers statistical analysis and visualization of health metrics. Enables sending email summaries via SMTP, using Gmail as an example provider. Allows AI-driven chart generation and visualization explanations using an OpenAI-compatible LLM.

How do I use Garmin MCP?

1. Click on "Deploy Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Garmin MCP show my steps for the last 7 days" That's it! The server will respond to your query, and you can continue using it as needed. Here is a step-by-step guide with screenshots.

Garmin MCP

by lokranjanp

Overview Schema Source Related Servers Discussions Admin

Python

Remote

Garmin MCP

MCP server that exposes Garmin Connect data (steps, sleep, stress, activities, etc.) via garth and fastMCP. Auth: set GARMIN_EMAIL and GARMIN_PASSWORD in env, or use a saved session at ~/.garth (or GARTH_SESSION_PATH).

MCP config

Use full path to uvx (e.g. which uvx). Example for Claude Desktop (~/Library/Application Support/Claude/claude_desktop_config.json) or Cursor:

{
  "mcpServers": {
    "garmin": {
      "command": "/Users/YOUR_USER/.local/bin/uvx",
      "args": ["--refresh", "--from", "git+https://github.com/lokranjanp/garminmcp", "garmin-mcp"],
      "env": {
        "GARMIN_EMAIL": "your-garmin-email@example.com",
        "GARMIN_PASSWORD": "your-password",

        "SMTP_HOST": "smtp.gmail.com",
        "SMTP_PORT": "587",
        "SMTP_USER": "your-email@gmail.com",
        "SMTP_PASSWORD": "your-app-password",
        "SMTP_TO": "recipient@example.com",
        "OUTPUT_DIR": "/Users/user.mac/garmin_data/",

        "LLM_API_BASE": "http://localhost:1234/v1",
        "LLM_API_KEY": "lm-studio",
        "LLM_MODEL": "your-loaded-model-name"
      }
    }
  }
}

From PyPI (if published): "args": ["garmin-mcp"].

Related MCP server: health-mcp

Commands

uv sync                    # install from uv.lock
uv run garmin-mcp          # run server locally
uv lock                    # refresh lock after changing pyproject.toml

Tools (17 base + 3 opt-in)

#	Tool	Description
	Utility
1	`garmin_current_datetime`	Current date/time (local & UTC, ISO, weekday)
	Auth
2	`garmin_login`	Log in and save session to disk
3	`garmin_resume_session`	Load saved session
	User
4	`garmin_user_profile`	Profile (display name, timezone, etc.)
5	`garmin_user_settings`	Settings (units, preferences)
	Metrics (daily/weekly series)
6	`garmin_metric`	Fetch daily or weekly stat series. `metric`: steps, sleep, stress, hydration, intensity_minutes, hrv. `period`: daily or weekly. `count`: number of periods.
	Detailed health data
7	`garmin_data`	Fetch detailed data (single day or list). `data_type`: sleep, hrv, weight, body_battery_events, body_battery_stress. Use `day` for single-day or `end_date`+`days` for a range.
	Activities
8	`garmin_activities`	List activities (start, limit)
9	`garmin_activity_details`	Full details for one activity by ID
10	`garmin_activity_types`	List all activity types (type IDs and keys)
	Biomarkers & summary
11	`garmin_daily_summary`	One-day wellness: RHR, HR, stress, steps, SpO2, respiration, body battery, calories
12	`garmin_resting_heart_rate`	Resting heart rate for a date range
13	`garmin_summary_report`	Comprehensive biomarker + activity report (daily/weekly/biweekly/monthly)
	Delivery
14	`garmin_email_summary`	Generate summary report and send via SMTP email (HTML + plain-text)
	Statistical analysis
15	`garmin_stats`	Run a stats operation on numeric data. `operation`: describe, percentiles, correlation, trend, compare, moving_average, outliers.
	Visualization (matplotlib)
16	`garmin_viz`	Render a chart. `chart_type`: line, bar, scatter, histogram, pie, heatmap, multi_line. Returns file path + base64 PNG.
	Raw API
17	`garmin_connect_api`	Call any Connect API path (GET/POST)
	LIDA / AI visualization (opt-in, requires `LLM_API_KEY`)
18	`garmin_lida_visualize`	AI-driven chart generation -- auto-picks the best chart type
19	`garmin_lida_goals`	Suggest N visualization goals for a dataset (EDA)
20	`garmin_lida_explain`	Explain a visualization's code in natural language

Dates: use YYYY-MM-DD; omit end date for "today" where supported.

SMTP env vars (for delivery tools)

Variable	Required	Description
`SMTP_HOST`	yes	SMTP server (e.g. `smtp.gmail.com`)
`SMTP_PORT`	no	Port, default `587` (STARTTLS)
`SMTP_USER`	yes	Login username / email
`SMTP_PASSWORD`	yes	Password or app-password
`SMTP_FROM`	no	Sender address (defaults to `SMTP_USER`)
`SMTP_TO`	no	Default recipient(s), comma-separated
`SMTP_USE_SSL`	no	`"true"` for implicit SSL (port 465); default uses STARTTLS

LLM env vars (for LIDA visualization tools -- opt-in)

The 3 LIDA tools (garmin_lida_*) are only registered when LLM_API_KEY is set. If omitted, the server loads with 17 tools instead of 20, reducing context overhead for the AI client.

Variable	Default	Description
`LLM_PROVIDER`	`"openai"`	LLM provider for LIDA
`LLM_API_BASE`	(none)	Custom endpoint -- set to `http://localhost:1234/v1` for LM Studio
`LLM_API_KEY`	(none)	API key (use `"lm-studio"` for LM Studio). Setting this enables LIDA tools.
`LLM_MODEL`	`"gpt-4o-mini"`	Model name (for LM Studio: whatever model is loaded)

Visualization output is saved to output/viz/ (or OUTPUT_DIR env var) and also returned as base64 PNG in the JSON response.

Available Tools

17 tools

garmin_activitiesA

List activities from Garmin Connect. start: index to start from, limit: max number to return. Returns activity IDs, names, types, start time, duration, distance, etc.

Inputs: limit, start. Returns: result

garmin_activity_detailsA

Get full details for one activity by ID (running, cycling, strength, etc.). Includes summary (distance, duration, HR, calories, cadence, etc.), splits, and metadata. Use garmin_activities first to get activity_id values.

Inputs: activity_id (required). Returns: result

garmin_activity_typesA

List all Garmin activity types (running, cycling, strength_training, etc.) with type IDs and keys.

Inputs: no named parameters. Returns: result

garmin_connect_apiA

Call the Garmin Connect API at an arbitrary path. Use for endpoints not wrapped by other tools. path: e.g. /userprofile-service/socialProfile or /activitylist-service/... method: GET or POST. body: optional JSON string for POST.

Inputs: path (required), body, method. Returns: result

garmin_current_datetimeA

Return current date and time variables (local and UTC). Use for building date ranges or passing today's date to other Garmin tools (e.g. YYYY-MM-DD for day/sleep/weight).

Inputs: no named parameters. Returns: result

garmin_daily_summaryA

Get one-day wellness summary (biomarkers): RHR, min/max HR, stress, steps, distance, calories, SpO2 (avg/low/high), respiration (avg/low/high), body battery (charged/max/min), intensity minutes, floors, sleep summary. day: YYYY-MM-DD.

Inputs: day (required). Returns: result

garmin_dataA

Fetch detailed health data from Garmin Connect.

data_type: "sleep", "hrv", "weight", "body_battery_events", or "body_battery_stress". day: YYYY-MM-DD for single-day data. If omitted, fetches a list. end_date: YYYY-MM-DD end of range for list mode (defaults to today). days: number of days for list mode (default 7; up to 30 for weight).

When day is provided: returns detailed data for that single day. When day is omitted: returns a list of entries over the date range.

Note: body_battery_events and body_battery_stress only support single-day mode (day param).

Examples: garmin_data(data_type="sleep", day="2026-02-01") # single day garmin_data(data_type="sleep", days=14) # last 14 days garmin_data(data_type="weight", end_date="2026-02-01", days=30) garmin_data(data_type="body_battery_events", day="2026-02-01")

Inputs: data_type (required), day, days, end_date. Returns: result

garmin_email_summaryA

Generate a Garmin summary report and email it via SMTP.

period: "daily", "weekly", "biweekly", or "monthly" (default: weekly). end_date: YYYY-MM-DD (defaults to today). to: recipient email(s), comma-separated (defaults to SMTP_TO env var). subject: custom subject line (auto-generated if omitted).

Requires SMTP env vars: SMTP_HOST, SMTP_PORT, SMTP_USER, SMTP_PASSWORD. Optional: SMTP_FROM (defaults to SMTP_USER), SMTP_TO (default recipient), SMTP_USE_SSL ("true" for port-465 implicit SSL; default is STARTTLS).

The email contains a formatted HTML report (with plain-text fallback) covering all biomarkers: HR, HRV, stress, sleep, SpO2, respiration, body battery, steps, hydration, intensity minutes, calories, weight, activities, and strength training sessions.

FOCUS MORE ON STRENGTH AND CARDIO BIO-MARKERS IF POSSIBLE.

Inputs: to, period, subject, end_date. Returns: result

garmin_loginA

Log in to Garmin Connect with email and password. Saves the session to the given path (default: ~/.garth) so you can use resume_session later without logging in again.

Inputs: email (required), password (required), session_path. Returns: result

garmin_metricA

Fetch daily or weekly stat series from Garmin Connect.

metric: "steps", "sleep", "stress", "hydration", "intensity_minutes", or "hrv". period: "daily" (default) or "weekly". end_date: YYYY-MM-DD (defaults to today/this week). count: number of days or weeks to fetch (default 7).

Examples: garmin_metric(metric="steps", period="daily", count=14) garmin_metric(metric="stress", period="weekly", count=4) garmin_metric(metric="hrv", count=28)

Inputs: metric (required), count, period, end_date. Returns: result

garmin_resting_heart_rateA

Get resting heart rate (RHR) for one or more days. end_date: YYYY-MM-DD or omit for today. Returns daily RHR and optional 7-day average. metricId 60 = resting heart rate.

Inputs: days, end_date. Returns: result

garmin_resume_sessionA

Resume a previously saved Garmin session from disk. Use this (or set GARTH_SESSION_PATH) before calling other tools so they are authenticated. Default path: ~/.garth.

Inputs: session_path. Returns: result

garmin_statsA

Perform a statistical analysis operation on numeric data.

operation: one of "describe", "percentiles", "correlation", "trend", "compare", "moving_average", "outliers".

Common params: values: primary list of numeric values (required for all operations). y: second list of values (for "correlation" and "compare"). labels: parallel date/label list (for "trend" and "moving_average").

Operation-specific params: describe: (no extra params needed) percentiles: percentiles (list of ranks 0-100, default [5,10,25,50,75,90,95]) correlation: y (required), method ("pearson" or "spearman") trend: labels (optional date labels) compare: y (required sample B), a_label, b_label moving_average: window (default 7), labels outliers: method ("iqr" or "zscore"), threshold (default 1.5)

Examples: Descriptive stats: operation="describe", values=[65,62,68,70,64] Trend analysis: operation="trend", values=[rhr_vals...], labels=[dates...] Correlation: operation="correlation", values=[steps...], y=[sleep_scores...], method="pearson" Compare weeks: operation="compare", values=[this_week...], y=[last_week...], a_label="this_week", b_label="last_week"

Inputs: values (required), operation (required), y, labels, +6 more. Returns: result

garmin_summary_reportA

Generate a comprehensive summary of all essential biomarkers and activity stats.

period: "daily" (1 day), "weekly" (7 days), "biweekly" (14 days), or "monthly" (30 days). end_date: YYYY-MM-DD (defaults to today).

Includes: RHR, HRV, stress, sleep score, SpO2, respiration, body battery, steps, hydration, intensity minutes, calories, weight, activities breakdown, and strength training sessions.

Inputs: period, end_date. Returns: result

garmin_user_profileB

Get the current user's Garmin Connect profile (display name, timezone, activities, etc.).

Inputs: no named parameters. Returns: result

garmin_user_settingsA

Get the current user's Garmin Connect settings (units, preferences, etc.).

Inputs: no named parameters. Returns: result

garmin_vizA

Render a chart and return the image inline (displayed by Claude Desktop) plus the saved file path.

chart_type: "line", "bar", "scatter", "histogram", "pie", "heatmap", or "multi_line".

Common params (used by most types): x: x-axis values or labels. y: y-axis values (for histogram, pass the data here). title, x_label, y_label: chart labels.

Type-specific params: histogram: bins (default 20). heatmap: matrix (2-D list), x_labels, y_labels. multi_line: y_series (list of y-value lists), series_labels.

Examples: Line: chart_type="line", x=["Mon","Tue",...], y=[8000,9200,...] Bar: chart_type="bar", x=["running","strength"], y=[5,3] Scatter: chart_type="scatter", x=[steps...], y=[sleep_scores...] Histogram: chart_type="histogram", y=[hrv_values...] Pie: chart_type="pie", x=["running","cycling"], y=[120,80] Heatmap: chart_type="heatmap", matrix=[[...],[...]], x_labels=["Mon",...], y_labels=["Wk1",...] Multi-line: chart_type="multi_line", x=[dates...], y_series=[[rhr...],[stress...]], series_labels=["RHR","Stress"]

Inputs: chart_type (required), x, y, bins, +8 more. Returns: result

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

Deploy Server

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

A3.6/5.0

Scored 2026-09-06 13:55 across 17 tools

Disambiguation2/5

Multiple tools overlap heavily: garmin_metric, garmin_data, garmin_daily_summary, garmin_summary_report, and garmin_email_summary all expose sleep, stress, HRV, or biomarker data in slightly different forms. garmin_summary_report and garmin_email_summary are nearly identical except for the email side effect, and garmin_connect_api adds another ambiguous path to the same data.

Naming Consistency4/5

The garmin_ prefix and snake_case style are consistent, and most names follow a readable resource-style pattern like garmin_activities or garmin_daily_summary. Minor inconsistencies like garmin_stats, garmin_viz, and garmin_email_summary (verb-style) keep it from being perfectly uniform.

Tool Count3/5

17 tools is in the borderline-heavy range for one server, especially since several tools are auxiliary (stats, viz, email, generic API passthrough) rather than core Garmin data retrieval. The overlapping wellness and summary tools could be consolidated to reduce the count without losing capability.

Completeness4/5

The tool set covers the core Garmin domain well: activities, activity details, biomarkers, user profile/settings, authentication, and an API fallback. Minor gaps exist, such as no explicit write/upload tools or dedicated series for every biomarker, but garmin_connect_api mitigates most missing endpoints.

Maintenance

ActivityInactive

ResponsivenessNo recent community issues

Resources

Related MCP Connectors

Withings Health Data
io.github.akutishevsky
MCP server for Withings health data — sleep, activity, heart, and body metrics.
Health & Wellness Research & Data
245
https://withings-mcp.com/mcp
Coach Watts
com.coachwatts
Remote MCP server for training, nutrition, wellness, and performance data with OAuth 2.0.
Health & Wellness Fitness Tracking
271
https://coachwatts.com/mcp
MCP for Oura
link.smirnov
Multi-tenant hosted MCP server for Oura Ring — 21 read-only tools, OAuth per user.
Health & Wellness Fitness Tracking
2
https://mcp-oura.smirnov.link/mcp
Garmin ConnectC
io.github.DynamicsNinja
Unofficial Garmin Connect access: sleep, HRV, training, activities, workouts and weight.
Fitness Tracking Health & Wellness Sports
19816
https://garmin.ficdev.xyz/mcp

Related MCP Servers

garth-mcp-server
Health & Wellness Fitness Tracking
matin
A
license
Not graded
quality
D
maintenance
MCP server to interact with Garmin Connect, enabling retrieval of health, activity, and device data through natural language.
Updated 9 months ago (2026-02-01 22:20 UTC)
66
MIT
health-mcp
Health & Wellness Fitness Tracking
jonathan249
A
license
Not graded
quality
D
maintenance
Exposes personal Garmin wellness data through MCP tools for accessing summary, sleep, HRV, heart rate, stress, body battery, and historical data.
Updated 5 months ago (2026-06-12 18:45 UTC)
MIT
garmin-mcp
Health & Wellness Fitness Tracking
gabecorsini
A
license
Not graded
quality
C
maintenance
Personal MCP server for interacting with your Garmin Connect data. Exposes 62 tools across 11 domains including activities, health, training, and workouts.
Updated 6 months ago (2026-05-12 16:07 UTC)
1
MIT
garmin-coach-mcp
Fitness Tracking Health & Wellness
jhoy1020
A
license
A
quality
D
maintenance
MCP server for reading and querying Garmin Connect data, including activities, strength history, recovery, trends, and optionally creating workouts.
Updated 3 months ago (2026-08-02 18:01 UTC)
12
MIT

Garmin MCP

MCP config

Commands

Tools (17 base + 3 opt-in)

SMTP env vars (for delivery tools)

LLM env vars (for LIDA visualization tools -- opt-in)

Available Tools

Tool Schema Changelog

TDQS

Maintenance

Resources

Related MCP Connectors

Related MCP Servers

garth-mcp-server

health-mcp

garmin-mcp

garmin-coach-mcp