Status pages and incidents

Public status pages#

Public status pages live at /status/<slug>: anonymous, server-rendered, and auto-refreshing every 60 seconds. Each shows a headline banner, 90-day uptime bars per monitor, and a 24-hour uptime percentage. Operators curate which monitors appear (nothing is auto-published). Manage them in the dashboard under Status pages.

Incidents#

On top of the live status, you can author an incident timeline per status page: the operator narrating an outage as it unfolds, like statuspage.io or GitHub status. An incident is a thread of updates, not a single post. Open one under Incidents (pick a status page, set a title, a severity, and a first update), then append updates as the situation evolves. Incidents are operator-authored only; no alert is sent, since alerts are the separate, monitor-driven system in Alerts.

Severity#

Severity moves through investigating, identified, monitoring, and resolved. The latest update's severity is the incident's current severity. Posting an update with severity resolved closes the incident; it stays on the public page (condensed, newest first) for about 24 hours, then drops off while remaining in the dashboard under the Resolved filter.

Update formatting#

Incident update bodies support a deliberately tiny, safe subset. The page is public and unauthenticated, so there are no links and no raw HTML:

Everything else is escaped to literal text. The public page renders with no JavaScript and a strict content security policy.