메뉴
CLI for Google Health API v4 — read steps, heart rate, exercise, sleep, weight, and 35+ health data types with agent-first JSON output
기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.
설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/Aradotso/devtools-skills --skill google-health-cli명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
SOC 직업 분류 기준
SKILL.md 표시 중
| name | google-health-cli |
| description | CLI for Google Health API v4 — read steps, heart rate, exercise, sleep, weight, and 35+ health data types with agent-first JSON output |
| triggers | ["get my step count from google health","fetch heart rate data from google fit","read my sleep data using google health api","export exercise data from google health","set up google health cli authentication","query my weight history from google health","retrieve daily health metrics from google fit","import health data using ghealth command"] |
Skill by ara.so — Devtools Skills collection.
CLI for the Google Health API v4 built for AI agents and developers. Supports 40+ verified health data types including steps, heart rate, exercise, sleep, weight, SpO2, HRV, ECG, blood glucose, and nutrition. Outputs deterministic JSON with agent-friendly exit codes.
git clone https://github.com/Google-Health-API/google-health-cli.git
cd google-health-cli
go build -o ghealth .
Move the binary to your PATH:
sudo mv ghealth /usr/local/bin/
ghealth setup
Walks through:
Credentials stored in ~/.config/ghealth/ (override with GHEALTH_CONFIG_DIR):
client_secret.json — OAuth client (mode 0600)credentials.json — access + refresh tokens (mode 0600)config.toml — active profileghealth setup \
--project-id $GCP_PROJECT_ID \
--client-secret ~/Downloads/client_secret.json \
--scopes-preset readonly \
--skip-enable-api \
--no-prompt
# 1. Start non-interactive login
ghealth auth login --non-interactive --scopes-preset readonly
# Returns JSON with auth_url and complete_command
# 2. Open auth_url in browser, copy the redirect URL or code parameter
# 3. Complete authentication
ghealth auth login --complete 'http://localhost/?code=4/0AX4XfWh...&state=cQq...'
# OR just the code:
ghealth auth login --complete 4/0AX4XfWh...
# Use existing access token
export GHEALTH_ACCESS_TOKEN=ya29...
# Use custom credentials file
export GHEALTH_CREDENTIALS_FILE=/path/to/creds.json
# Custom config directory
export GHEALTH_CONFIG_DIR=/custom/config/path
Precedence: GHEALTH_ACCESS_TOKEN > GHEALTH_CREDENTIALS_FILE > stored credentials > Application Default Credentials.
# Export on source machine
ghealth auth export > /tmp/ghealth-creds.json
# Import on target machine (requires client_secret.json)
ghealth auth import --file /tmp/ghealth-creds.json
| Type | Key Values | Operations |
|---|---|---|
steps | countSum (daily rollup) | list, rollup, daily-rollup, reconcile |
heart-rate | beatsPerMinute | list, rollup, daily-rollup, reconcile |
exercise | type, duration, calories, avgHeartRate | list, get, create, update, delete, export-tcx |
sleep | minutesAsleep, minutesAwake, stageMinutes | list, get, create, update, delete, reconcile |
weight | weightGrams | list, get, create, update, delete, rollup, daily-rollup |
distance | millimetersSum (daily rollup) | list, rollup, daily-rollup, reconcile |
oxygen-saturation | percentage (SpO2) | list, reconcile |
heart-rate-variability | RMSSD | list, reconcile |
blood-glucose | mg/dL, mealType, timing | list, get, rollup, daily-rollup, reconcile |
nutrition-log | nutrients, energy, mealType | list, get, rollup, daily-rollup, reconcile |
Run ghealth schema types for all 40+ types.
# Recent heart rate readings
ghealth data heart-rate list --from today --limit 10
# Heart rate for specific date range
ghealth data heart-rate list --from 2026-03-22 --to 2026-03-29 --limit 100
# Oxygen saturation readings
ghealth data oxygen-saturation list --from 2026-06-01 --limit 20
# Recent weight measurements
ghealth data weight list --limit 10
Critical: For steps, distance, and other interval types, use daily-rollup to get actual values. list returns time intervals without totals.
# Daily step totals (CORRECT way)
ghealth data steps daily-rollup --from 2026-03-22 --to 2026-03-29
# Returns: {"dataPoints": [{"date": "2026-03-28", "countSum": "9037"}, ...]}
# Daily distance totals
ghealth data distance daily-rollup --from 2026-03-01 --to 2026-03-31
# Weekly average heart rate
ghealth data heart-rate rollup --from 2026-03-01 --to 2026-03-31 --duration 7d
# Hourly step counts for today
ghealth data steps rollup --from today --duration 1h
# List recent exercises
ghealth data exercise list --from 2026-03-01 --limit 20
# Get specific exercise
ghealth data exercise get --id <exercise-id>
# Export exercise track as TCX
ghealth data exercise export-tcx --id <id> --output ride.tcx
# Export as CSV for data analysis
ghealth data exercise export-tcx --id <id> --output ride.csv --as csv
# CSV columns: time, activity, lap, sport, latitude_deg, longitude_deg, altitude_m,
# distance_m, heart_rate_bpm, cadence_rpm, speed_mps, watts
# Stream to stdout
ghealth data exercise export-tcx --id <id> --output - --as csv | head
# Recent sleep sessions (summary)
ghealth data sleep list --limit 5
# Sleep with stage-by-stage breakdown
ghealth data sleep list --limit 5 --detail
# Specific sleep session
ghealth data sleep get --id <sleep-id>
All list commands support pagination:
# First page (default limit 500)
ghealth data heart-rate list --from 2026-06-15 --limit 500
# Returns: {"dataPoints": [...], "nextPageToken": "ABC"}
# Next page
ghealth data heart-rate list --from 2026-06-15 --limit 500 --page-token ABC
ghealth data exercise create \
--start-time 2026-03-28T14:00:00Z \
--end-time 2026-03-28T15:30:00Z \
--type running \
--calories 450 \
--distance-meters 8000 \
--avg-heart-rate 152 \
--notes "Morning run in the park"
ghealth data exercise update \
--id <exercise-id> \
--notes "Evening run - felt great" \
--calories 475
ghealth data sleep create \
--start-time 2026-03-27T23:00:00Z \
--end-time 2026-03-28T07:00:00Z \
--minutes-asleep 450 \
--minutes-awake 30
ghealth data weight create \
--timestamp 2026-03-28T08:00:00Z \
--weight-grams 75000
ghealth data exercise delete --id <exercise-id>
ghealth data sleep delete --id <sleep-id>
ghealth data weight delete --id <weight-id>
Every read command returns:
{
"dataPoints": [
{"date": "2026-03-28", "countSum": "9037"},
{"date": "2026-03-27", "countSum": "8245"}
],
"_hints": {
"type": "steps",
"rollupDuration": "1d"
},
"nextPageToken": "optional-token-if-more-data"
}
Parse with jq:
# Extract step counts
ghealth data steps daily-rollup --from 2026-03-22 --to 2026-03-29 | jq -r '.dataPoints[].countSum'
# Get total steps for week
ghealth data steps daily-rollup --from 2026-03-22 --to 2026-03-29 | jq '[.dataPoints[].countSum | tonumber] | add'
# Average heart rate
ghealth data heart-rate list --from today --limit 100 | jq '[.dataPoints[].beatsPerMinute | tonumber] | add / length'
# Preview API request without executing
ghealth data steps daily-rollup --from today --dry-run
# Get unprocessed API response
ghealth data heart-rate list --from today --limit 10 --raw
0 — Success1 — General error2 — Authentication/validation error3 — API error4 — Data not found5 — Configuration errorFor presence-aware types (altitude, distance, floors, steps, total-calories):
countSum: "0" → true zero (device worn, no activity)# Example output:
# {"dataPoints": [
# {"date": "2026-03-28", "countSum": "9037"}, # Active day
# {"date": "2026-03-27", "countSum": "0"}, # Worn but sedentary
# # 2026-03-26 missing entirely # Not worn - DON'T treat as 0
# ]}
Never average over absent days as if they were zeros — this silently deflates statistics.
2026-03-28T14:00:00Ztoday, yesterday2026-03-28 (interprets as start of day in user's timezone)ghealth data steps daily-rollup --from 2026-03-22 --to today
ghealth data heart-rate list --from yesterday --limit 50
#!/bin/bash
# Get last 7 days of step data
WEEK_AGO=$(date -d '7 days ago' +%Y-%m-%d)
TODAY=$(date +%Y-%m-%d)
ghealth data steps daily-rollup --from $WEEK_AGO --to $TODAY | \
jq -r '.dataPoints[] | "\(.date): \(.countSum) steps"'
#!/bin/bash
# Export all exercises from a month
ghealth data exercise list --from 2026-03-01 --to 2026-03-31 | \
jq -r '.dataPoints[].id' | \
while read id; do
ghealth data exercise export-tcx --id "$id" --output "exercise_${id}.csv" --as csv
done
#!/bin/bash
# Fetch today's key metrics
echo "=== Health Metrics for $(date +%Y-%m-%d) ==="
echo -n "Steps: "
ghealth data steps daily-rollup --from today --to today | jq -r '.dataPoints[0].countSum // "no data"'
echo -n "Average Heart Rate: "
ghealth data heart-rate list --from today --limit 500 | \
jq '[.dataPoints[].beatsPerMinute | tonumber] | add / length | round'
echo -n "Weight: "
ghealth data weight list --limit 1 | jq -r '.dataPoints[0].weightGrams // "no data"'
Check for data conflicts and resolve:
# Reconcile heart rate data
ghealth data heart-rate reconcile --from 2026-03-01 --to 2026-03-31
# Reconcile with conflict resolution strategy
ghealth data exercise reconcile --from 2026-03-01 --strategy newest
If you see exit code 5 with JSON error:
{
"error": {
"type": "config",
"code": 5,
"message": "No OAuth client_secret.json configured",
"next_steps": [
"Open https://console.cloud.google.com/apis/credentials",
"Create or select a Google Cloud project",
"Enable the Google Health API",
"Create OAuth client ID with Application type: Desktop app",
"Download the client_secret JSON",
"Run: ghealth setup --client-secret /path/to/client_secret.json"
]
}
}
Get instructions without error:
ghealth setup --instructions
# Check auth status
ghealth auth status
# Force refresh
ghealth auth refresh
# Re-authenticate
ghealth auth login
Check:
--from and --to)daily-rollup vs list)# Verify scopes
ghealth auth status | jq -r '.scopes'
# Check available types
ghealth schema types
The API has rate limits. Space out requests:
# Add delays between bulk operations
for id in $(ghealth data exercise list | jq -r '.dataPoints[].id'); do
ghealth data exercise get --id "$id"
sleep 1
done
# List all available data types
ghealth schema types
# Get details for specific type
ghealth schema type --name steps
# View available operations for a type
ghealth schema type --name exercise | jq -r '.operations[]'
daily-rollup for totals — list returns intervals, not sumsnextPageToken in responses$? for error handling.dataPoints--from and --to prevent unbounded queries--dry-run for validation before destructive operations