Status pages

A status page is a public, no-login URL showing the current state of monitors you choose. It’s the thing you point at when someone asks “is the nightly export broken again?” — and the thing that stops them asking.

Creating one

A status page has a name, a slug, an optional description, and a list of the monitors it includes.

The slug is the address:

https://status.deadpost.dev/status/acme-corp

Slugs are unique across all of deadpost, not just your workspace, because they address a public page — so short generic ones (status, api) will usually be taken. Something identifying works better anyway: the page is a link you’ll paste into other people’s channels.

Caution

That URL is public and unauthenticated. Anyone with the link sees every monitor on the page — names included. Monitor names often carry more than intended (acme-payroll-export, customer-db-backup), so choose what goes on a status page deliberately rather than adding everything.

Choosing monitors

You pick which monitors appear and what order they’re in. Order is worth spending a moment on: the page is read top-down by someone in a hurry, so the thing they’re most likely to be asking about belongs first.

Monitors you don’t add are invisible — there’s no partial disclosure, no greyed rows for things you left off.

What visitors see

Each monitor shows its current state and a history strip of recent days, so a visitor can tell the difference between “broken right now” and “broken all week”.

The page also shows an overall status, computed from the monitors on it:

Overall	When
Operational	Everything on the page is healthy.
Degraded	Some monitors are late or down, but not all.
Outage	Everything on the page is down.

Uptime percentages can be shown or hidden. Hiding them is the right call more often than you’d think — a number invites an argument about the number, and a status page’s job is to answer “is it working?”, not to be an SLA report.

What it doesn’t do

The page reflects monitor state only. It’s derived from whether your jobs checked in, not from anything you type — there’s no incident-posting workflow, no “we are investigating” banner, no subscriber notifications.

That’s a deliberate scope. A status page here is an honest projection of the monitoring you already have. If your job stopped checking in, the page says so without anyone remembering to update it — which is the failure mode that makes most hand-maintained status pages useless.

For the narrative side, outage notes are internal and attach to the specific outage they explain.