Uptime monitoring you didn't pay for

You have three production endpoints to watch: the public API, the auth service, and the checkout. The pricing page on the monitoring SaaS starts at $30 a month per check.

You POST three schedules to /api/v1/curl/schedule, each with a once-a-minute cron, a short timeout, and retry_count: 2. Each run records status code, response time, and body to storage. A second schedule, every five minutes, hits an exec script that reads the latest jobs and fires a hoody-notifications push to your phone if anything looks wrong.

You watch the jobs page on your phone like a tab. You set the notification rule once. A fourth check is one more POST.

The check, the schedule, the alert, and the history all live as URLs. Three URLs and a phone replaced the synthetic-monitor SaaS.