Browse documentation
Trust

Status & incidents

Nine probed services, a rolled-up 90-day history, how incidents are published, and the three ways to be told about one without watching a page.

/status is public and needs no account.

What is probed

Nine services, each checked independently:

platform · api · authentication · database · aiGateway · backgroundJobs · fileStorage · emailDelivery · shopifyIntegration

They are probed separately on purpose. "The platform is up" is not a useful answer to someone whose Shopify sync is failing.

The history

A snapshot cron rolls the probes up every 30 minutes into a worst-of daily record. Worst-of rather than average: a day with a 40-minute outage should read as a bad day, not as 97% of a good one.

The page shows 90 days of daily bars, and /status/history gives a monthly report card over 12 months.

Incidents

An incident is declared by an operator and moves through investigating → identified → monitoring → resolved, at one of three severities: minor, major, critical.

Three ways to hear about it

Subscribe by email on /status. Double opt-in, with a one-click unsubscribe token in every message.

RSS/status/incidents.rss, the last 90 days.

Webhooks — outbound, HMAC-SHA256 signed, for incident events. Set up from the admin surface; verification is described in Webhooks.

For your own dashboards

GET /api/status/summary is public and cached for 60 seconds. It returns a state, a count and a timestamp, deliberately not incident detail. It exists to drive a status dot in a footer, not to be an incident feed; use RSS or a webhook for that.

Built on these docs?

Drop into the forum if something's unclear or wrong. Docs improve faster when readers flag the gaps.

Open the forum