← 返回 Skills 市场
voydz

Garmin Cli

作者 Voydz · GitHub ↗ · v1.0.5
cross-platform ⚠ suspicious
1826
总下载
1
收藏
5
当前安装
6
版本数
在 OpenClaw 中安装
/install garmin-cli
功能描述
Access Garmin Connect health, fitness, and activity data via a non-interactive CLI.
使用说明 (SKILL.md)

Garmin Connect CLI

This skill provides access to Garmin Connect health and fitness data through the gc CLI.

Setup

  1. Install via Homebrew tap:

    brew tap voydz/homebrew-tap
    brew install garmin-cli
    
  2. Authentication:

    gc login --email [email protected] --password secret
    # With MFA:
    gc login --email [email protected] --password secret --mfa 123456
    
  3. Verify connection:

    gc status
    

Date Shortcuts

Most commands accept a date shortcut as first argument:

  • today — current date
  • yesterday — previous date
  • week — last 7 days (returns a date range)
  • month — last 30 days (returns a date range)
  • YYYY-MM-DD — specific date

For command groups with subcommands (activities, body, stress, heart, menstrual), use --date on the parent command to avoid argument conflicts. Alternatively use --date, --start/--end flags.

Output

All data commands support:

  • --format json for machine-readable output (default: table)
  • --output FILE to write to a file

Always use --format json when parsing output programmatically.

Usage

# Authentication
gc login --email EMAIL --password PASS [--mfa CODE | --wait-mfa]
gc logout
gc status
gc status --profile

# Daily Health
gc health today
gc steps today
gc steps week
gc steps --weekly --weeks N
gc steps --start DATE --end DATE
gc floors today
gc intensity today
gc intensity --weekly --weeks N
gc events today

# Heart Rate
gc heart --date today
gc heart resting --date today

# Sleep
gc sleep today

# Stress & Body Battery
gc stress --date today
gc stress --weekly --weeks N
gc stress all-day --date today
gc battery today
gc battery --start DATE --end DATE
gc battery --events today

# Vitals
gc respiration today
gc spo2 today
gc blood-pressure today [--end DATE]
gc lifestyle today

# Hydration
gc hydration today

# Activities
gc activities                                     # List recent (default 20)
gc activities --limit N --offset N --type TYPE
gc activities --start DATE --end DATE [--type TYPE]
gc activities --date today                        # Activities for a date
gc activities last                                # Most recent activity
gc activities get ID                              # Activity summary by ID
gc activities count                               # Total count
gc activities details ID
gc activities splits ID
gc activities typed-splits ID
gc activities split-summaries ID
gc activities weather ID
gc activities hr-zones ID
gc activities power-zones ID
gc activities exercise-sets ID
gc activities types                               # List all activity types
gc activities gear ID                             # Gear used for activity
gc activities progress --start DATE --end DATE --metric distance|duration|elevation
gc activities download ID --format fit|tcx|gpx|kml|csv [-o FILE]
gc activities upload FILE                         # .fit, .gpx, .tcx

# Body & Weight
gc body --date today [--end DATE]
gc body weighins --date today
gc body weighins --start DATE --end DATE

# Advanced Metrics
gc metrics                                       # Metrics summary
gc metrics --date today
gc metrics vo2max today
gc metrics hrv today
gc metrics training-readiness today
gc metrics morning-readiness today
gc metrics training-status today
gc metrics fitness-age today
gc metrics race-predictions
gc metrics race-predictions --start DATE --end DATE --type daily|monthly
gc metrics endurance-score today [--end DATE]
gc metrics hill-score today [--end DATE]
gc metrics lactate-threshold                      # Latest
gc metrics lactate-threshold --no-latest --start DATE --end DATE --aggregation daily|weekly|monthly|yearly
gc metrics cycling-ftp

# Note: `gc metrics` summary resolves `vo2max` from daily maxmet first and
# falls back to `training-status.mostRecentVO2Max` when the selected date has
# no new maxmet sample.

# Devices
gc devices                                        # List all devices
gc devices last-used
gc devices primary
gc devices settings DEVICE_ID
gc devices alarms
gc devices solar DEVICE_ID today [--end DATE]

# Goals, Records, Badges & Challenges
gc records
gc goals [--status active|future|past] [--limit N]
gc badges earned
gc badges available
gc badges in-progress
gc challenges adhoc [--start N --limit N]
gc challenges badge [--start N --limit N]
gc challenges available [--start N --limit N]
gc challenges non-completed [--start N --limit N]
gc challenges virtual [--start N --limit N]

# Gear
gc gear --user-profile USER_PROFILE_NUMBER        # List gear (profile number from gc status --profile)
gc gear defaults USER_PROFILE_NUMBER
gc gear stats GEAR_UUID
gc gear activities GEAR_UUID [--limit N]

# Workouts & Training Plans
gc workouts [--start N --limit N]
gc workouts get WORKOUT_ID
gc workouts download WORKOUT_ID [-o FILE]
gc workouts scheduled WORKOUT_ID
gc workouts create --file workout.json
gc workouts create --name "Workout Name" --sport cycling --steps '[{"type":"warmup","duration":600},{"type":"interval","duration":1200,"target":"hr_zone:2"},{"type":"cooldown","duration":600}]'
gc workouts update WORKOUT_ID --file workout.json
gc workouts update WORKOUT_ID --name "Workout Name" --sport cycling --steps '[{"type":"warmup","duration":600},{"type":"interval","duration":1200},{"type":"cooldown","duration":600}]'
gc workouts delete WORKOUT_ID
gc training-plans
gc training-plans get PLAN_ID
gc training-plans adaptive PLAN_ID

## Workouts Steps JSON Schema (`--steps`)

`--steps` expects a JSON array of step objects. Each step can be shorthand or Garmin-shaped.

Shorthand step example:
```json
{"type":"interval","duration":1200,"target":"hr_zone:2"}

Supported shorthand fields:

  • type: warmup, interval, recovery, cooldown, rest, repeat
  • duration: seconds (implies endCondition = time)
  • target: hr_zone:2, power_zone:3, pace_zone:4, heart_rate:150, power:220, cadence:90, no_target

Garmin-shaped fields (optional):

  • stepType: {"stepTypeKey":"warmup"} (or any Garmin stepType object)
  • stepOrder: integer
  • endCondition: {"conditionTypeKey":"time|distance|calories|heart_rate|cadence|power|iterations"}
  • endConditionValue: number
  • targetType: {"workoutTargetTypeKey":"no.target|heart.rate|power|speed|cadence|open|heart.rate.zone|power.zone|pace.zone"}
  • targetValue: number

For advanced Garmin payloads (repeat groups, nested steps, etc.), prefer --file with the full Garmin schema.

Menstrual Cycle

gc menstrual --date today gc menstrual calendar --start DATE --end DATE gc menstrual pregnancy

Raw API

gc api /biometric-service/biometric/latestFunctionalThresholdPower/CYCLING gc api /metrics-service/metrics/maxmet/daily/DATE/DATE gc api /metrics-service/metrics/trainingstatus/aggregated/DATE gc api --method POST --body '{"foo":"bar"}' /some/endpoint


## Examples

**Get today's health summary as JSON:**
```bash
gc health today --format json

Get last week's steps as JSON for analysis:

gc steps week --format json

Find the user's most recent run:

gc activities --limit 5 --type running --format json

Call a raw Garmin Connect API endpoint:

gc api /biometric-service/biometric/latestFunctionalThresholdPower/CYCLING
gc api /metrics-service/metrics/maxmet/daily/2026-03-03/2026-03-03
gc api /metrics-service/metrics/trainingstatus/aggregated/2026-03-03
gc api --method POST --body '{"foo":"bar"}' /some/endpoint

Get detailed info about a specific activity:

gc activities get 12345678 --format json

Download an activity as GPX:

gc activities download 12345678 --format gpx -o run.gpx

Check training readiness and HRV:

gc metrics training-readiness today --format json
gc metrics hrv today --format json

Get sleep and body battery for yesterday:

gc sleep yesterday --format json
gc battery yesterday --format json

Workout creation (concise)

  • Prefer --file with a Garmin-shaped JSON payload.
  • Get a valid payload by exporting an existing workout:
    gc workouts get WORKOUT_ID --format json > workout.json
    
  • If using flags, --steps can be the exact workoutSteps JSON array from the API or a shorthand array with type, duration (seconds), and optional target (e.g. hr_zone:2).
  • --sport-id is optional when --sport is provided; the CLI resolves the id from activity types.

Garmin workout shape (minimal example):

{
  "workoutName": "Zone 2 Ride",
  "sportType": {
    "sportTypeKey": "cycling",
    "sportTypeId": 17
  },
  "workoutSegments": [
    {
      "segmentOrder": 1,
      "sportType": {
        "sportTypeKey": "cycling",
        "sportTypeId": 17
      },
      "workoutSteps": [
        {
          "stepOrder": 1,
          "stepType": { "stepTypeKey": "warmup" },
          "endCondition": { "conditionTypeKey": "time" },
          "endConditionValue": 600
        },
        {
          "stepOrder": 2,
          "stepType": { "stepTypeKey": "interval" },
          "endCondition": { "conditionTypeKey": "time" },
          "endConditionValue": 3600
        }
      ]
    }
  ]
}

Example creations:

# Create from file (recommended)
gc workouts create --file workout.json

# Create with flags using exact Garmin steps JSON
gc workouts create \
  --name "Zone 2 Ride" \
  --sport cycling \
  --sport-id 17 \
  --steps '[{"stepOrder":1,"stepType":{"stepTypeKey":"warmup"},"endCondition":{"conditionTypeKey":"time"},"endConditionValue":600},{"stepOrder":2,"stepType":{"stepTypeKey":"interval"},"endCondition":{"conditionTypeKey":"time"},"endConditionValue":3600}]'

How to discover format and valid values for workout creation:

  • Sport type keys/ids (used in sportType):
    • gc activities types --format json
  • Workout step/target enums are not hardcoded in the CLI.
    • Export an existing workout and reuse the exact values:
      gc workouts get WORKOUT_ID --format json
      
    • Use the returned stepType, endCondition, and targetType fields verbatim.

List devices and get solar data:

gc devices --format json
gc devices solar DEVICE_ID today --format json
安全使用建议
This skill is coherent: it's just documentation for using a local 'gc' CLI. Before installing or using it: (1) verify the source of the 'gc' binary — the SKILL.md points to a third‑party Homebrew tap (voydz/homebrew-tap); prefer an official project release or verify the tap and binary checksums/signature. (2) Avoid passing passwords or MFA codes on the command line (they appear in shell history and process lists); use interactive prompts, a keychain, or environment-based secrets if supported. (3) Check how the 'gc' binary stores authentication tokens (keychain vs plaintext files) and clean up or revoke credentials if you stop using the tool. (4) If you plan to let an autonomous agent call this skill, be aware the agent will be able to run the installed 'gc' binary and read any files it creates, so restrict agent permissions accordingly. If you want, provide the location (URL/repo) or checksum of the expected 'gc' release and I can re-evaluate with more detail.
功能分析
Type: OpenClaw Skill Name: garmin-cli Version: 1.0.5 The skill is suspicious due to its reliance on an external Homebrew tap (`voydz/homebrew-tap`) and package (`garmin-cli`) for its core functionality, as instructed in `SKILL.md`. This introduces a supply chain vulnerability, as the integrity of the `garmin-cli` binary and its distribution channel cannot be guaranteed by the skill bundle itself. Additionally, the `gc login` command, also documented in `SKILL.md`, involves passing user credentials directly on the command line, which can be a security risk depending on the execution environment and logging practices.
能力评估
Purpose & Capability
The name/description match the instructions: the skill is an instruction-only adapter for the 'gc' CLI. The declared required binary is 'gc', which is exactly what the SKILL.md uses. There are no extraneous requirements (no unrelated env vars, no unrelated binaries).
Instruction Scope
SKILL.md stays within scope: it instructs installing and running the 'gc' CLI and lists its commands. It does, however, show example usage that passes email/password (and MFA codes) on the command line (e.g., `gc login --password secret`), which is an insecure practice because it can expose credentials in shell history or process lists. The instructions do not ask the agent to read unrelated files or exfiltrate data to unexpected endpoints.
Install Mechanism
There is no registry install spec, but SKILL.md recommends `brew tap voydz/homebrew-tap` and `brew install garmin-cli`. That installs a binary from a third‑party Homebrew tap (voydz), which may be fine but is higher risk than an official Homebrew/core formula or a signed release on a well-known project page. The skill itself does not automatically download or write code (instruction-only).
Credentials
The skill declares no required environment variables or credentials, which is proportional. However, runtime instructions show passing user credentials and MFA codes on the command line (in examples), which can leak secrets. Also note: once the 'gc' binary is installed, it may store credentials or session tokens locally (keychain/files) — the skill does not document where, so verify the 'gc' tool's storage behavior before use.
Persistence & Privilege
The skill is instruction-only and not marked 'always'. It does not request persistent presence, system-level config changes, or other skills' configuration. The agent can invoke it autonomously (default), but that's expected for skills and is not combined with other red flags here.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install garmin-cli
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /garmin-cli 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.5
- Removed the file: _meta.json. - Updated documentation to clarify that gc metrics commands support summary queries and added a note on how `vo2max` is resolved for the summary. - Enhanced examples and API usage documentation to include more raw API call patterns. - Minor adjustments to option lists and metric command documentation for clarity.
v1.0.4
garmin-cli 1.0.4 Changelog - Added _meta.json file for improved metadata management. - Updated SKILL.md: - Clarified usage of --date flag for parent commands to avoid argument conflicts. - Expanded `--steps` documentation to support both shorthand and full Garmin JSON schemas. - Added examples for raw API calls and clarified usage around command arguments. - Improved formatting, command grouping, and documentation for clarity.
v1.0.3
- Removed the internal metadata file `_meta.json`. - Updated SKILL.md to clarify and shorten workout creation documentation. - Added details on providing `--sport-id` and using accurate Garmin JSON for workout steps. - Included instructions on how to discover valid workout format and enum values via the CLI. - No functional or command changes to the skill interface.
v1.0.2
- Updated installation instructions to use Homebrew package name garmin-cli instead of gc. - No functional changes to commands or usage; documentation only.
v1.0.1
- Added _meta.json file with skill metadata. - Updated SKILL.md: shortened the description, removed the "Requirements" section, and added a metadata block. - Expanded usage example for "gc workouts" including create, update, and delete commands. - No CLI functionality changes; documentation and metadata only.
v1.0.0
Initial release: Read health, fitness, and activity data from Garmin Connect via gc CLI
元数据
Slug garmin-cli
版本 1.0.5
许可证
累计安装 6
当前安装数 5
历史版本数 6
常见问题

Garmin Cli 是什么?

Access Garmin Connect health, fitness, and activity data via a non-interactive CLI. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 1826 次。

如何安装 Garmin Cli?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install garmin-cli」即可一键安装,无需额外配置。

Garmin Cli 是免费的吗?

是的,Garmin Cli 完全免费(开源免费),可自由下载、安装和使用。

Garmin Cli 支持哪些平台?

Garmin Cli 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Garmin Cli?

由 Voydz(@voydz)开发并维护,当前版本 v1.0.5。

💬 留言讨论