-
Part 1 of 3 · The sandbox
Your own copy of the monitor
You get three pretend websites to watch, all yours, gone in two hours. No account needed.
Under the hood. A throwaway tenant, isolated from every other visitor by PostgreSQL row-level security, seeded with simulated sites. The sandbox never sends traffic to real addresses.
-
Part 2 of 3 · Break something
Take a site down and watch the monitor notice
Switch a site to Down. Two failed checks later, an incident opens by itself.
Under the hood. Each result goes through a small state machine with hysteresis: several failures in a row open an incident, several successes close it, so one blip never pages anyone.
-
Part 3 of 3 · Run the incident
Post updates, then let it recover
Write a public update or an internal note, then switch the site back and watch the incident close.
Under the hood. Public updates appear on your status page; internal notes never do. Everything on the page is server-rendered from the same data the console edits.