Alerts
Incidents: understand, mute, remember
Each incident gathers what is known, alerts are muted on the spot, and the month reads in figures.
When something breaks, the first minute counts: knowing what failed, since when, and whether the last deploy had anything to do with it. The Incidents page was reworked to answer those three questions before anything else.
One incident, everything known
Each incident has its page: what failed and on which URL, since when, what the last deploy changed, and the probable cause when PostShip can point to it. You take it with one button — the team then knows someone is on it — and if it stays unanswered too long, escalation warns the whole team.
Mute alerts without searching
At the top of the page, a “Mute alerts” menu shows the current state and offers 1 h, 4 h or 24 h. The project's quiet hours fit in a pill next to it. Muting hides nothing: incidents keep opening and closing, they just don't ring.
The month in figures
The last 30 days against the 30 before: number of incidents, median recovery time, time to acknowledge. One bar per day, and each incident as a card, filterable by state, period or payment page. The history pages through six at a time, and the card always keeps the same size.