Health Check

Monitor service health and build a status page:

1. Check endpoints — For each configured URL, use web to:
   • Fetch the health endpoint
   • Record response status (200/non-200) and response time
2. Build canvas — Status page dashboard:
   • KPIs: services up, services down, avg response time
   • Status indicators per service: Badge (green “Healthy” / red “Down” / yellow “Degraded”)
   • Chart: response time over recent checks (bar chart)
   • Table: detailed check results (service, status, response time, last checked)
3. Track — Store historical check data in the canvas component state:
   • Fields: service, status, responseTimeMs, checkedAt
   • Update the canvas file via write_file after each check run
4. Alert — If any endpoint returns non-200:
   • send_message to alert channel immediately
   • Include which service, what error, and when it started
5. Persist — Log check results to memory for trend analysis

On heartbeat, re-check all endpoints and update the status page canvas.