Workflows & Automation
Automate your health check-ins with OpenClaw workflows. Instead of manually asking for insights each day, you can set up repeatable routines that pull your latest Apple Health data and deliver summaries on a schedule.
All workflows run through the local API on your Mac. Your health data never leaves your machine unless you explicitly choose a cloud AI provider.
Daily Brief Workflow
The Daily Brief gives you a quick morning snapshot of your health status. It covers steps vs. your baseline, sleep duration and quality, any workouts logged, and missing-data alerts.
What you get
- Steps compared to your 30-day average
- Sleep duration and bedtime consistency
- Workout summary (type, duration, calories)
- Resting heart rate trend
- Three actionable suggestions for the day
How to run it
Open OpenClaw and paste this prompt:
Use the "health-analyzer-mac-local" skill for this request.
Give me my daily health brief for today and 3 suggestions.
Weekly Report Workflow
The Weekly Report compares the past seven days against your previous week to surface trends. It produces a scorecard across three dimensions: Consistency, Recovery, and Activity.
What you get
- Week-over-week comparison of steps, sleep, and workouts
- A scorecard rating Consistency, Recovery, and Activity
- Highlighted positive and negative trends
- Three experiments to try the following week
How to run it
Use the "health-analyzer-mac-local" skill for this request.
Pull my daily brief for the last 7 days and summarize trends in steps, sleep, and workouts.
Then give me a weekly scorecard (Consistency, Recovery, Activity) and 3 experiments for next week.
Setting Up a Scheduled Workflow
You can schedule these prompts to run automatically so you never forget a check-in.
- Install the Mac app and import your Apple Health export. Make sure the local API is running (check the menu bar icon).
- Open OpenClaw and verify the
health-analyzer-mac-localskill is available. - Create a new workflow in OpenClaw. Paste your prompt and set the schedule (e.g., daily at 7:00 AM for the Daily Brief, Mondays at 9:00 AM for the Weekly Report).
- Choose your output — OpenClaw can display results in the app, save them to a file, or send a notification.
Always start your prompt with the skill prefix so OpenClaw routes the request through the local analyzer API:
Use the "health-analyzer-mac-local" skill for this request.
Integration with Mac Shortcuts
You can trigger health analysis from Mac Shortcuts by calling the local API endpoints directly. This lets you build custom automations that combine health data with other apps.
Example: Daily Brief via Shortcuts
- Open the Shortcuts app on your Mac.
- Create a new shortcut and add a "Get Contents of URL" action.
- Set the URL to
http://localhost:19876/api/v1/summary(the default local API endpoint). - Add a "Show Result" action to display the output, or pipe it into a notification.
- Optionally, set the shortcut to run on a schedule using Shortcuts Automation.
Custom Script Examples
If you prefer the command line, you can use curl to query the local API and build your own scripts.
Fetch today's summary
curl -s http://localhost:19876/api/v1/summary | python3 -m json.tool
Get the last 7 days of metrics
curl -s "http://localhost:19876/api/v1/metrics?days=7" | python3 -m json.tool
Pipe data into an AI prompt
# Fetch health data and send it to a local Ollama model
DATA=$(curl -s http://localhost:19876/api/v1/summary)
curl -s http://localhost:11434/api/generate \
-d "{
\"model\": \"llama3\",
\"prompt\": \"Analyze this health data and give 3 suggestions: $DATA\",
\"stream\": false
}" | python3 -m json.tool
Save a weekly report to a file
# Run every Monday via cron
curl -s "http://localhost:19876/api/v1/metrics?days=7" \
| python3 -m json.tool \
> ~/Desktop/health-report-$(date +%Y-%m-%d).json
See the API Reference for a complete list of available endpoints and parameters.
Next Steps
- Browse the Prompt Library for more ready-to-use prompts.
- Learn about AI Analysis providers and privacy considerations.
- Read the API Reference for full endpoint documentation.