RoostRoost Incidents
When something breaks, one timeline
Roost Incidents is ITIL-style incident management without the ceremony: declare, grade, assemble the response, and keep every decision on one append-only timeline — then close with a review that turns the fire into follow-up work.
Severity that means something
SEV1 to SEV4, with house conventions built in. Re-grade as the picture sharpens — but every change of severity requires a note, because the why is what the review needs later.
One timeline, append-only
Status changes, severity re-grades, team changes, and the running comment log interleave into a single war-room record. Nothing on it can be edited away — same audit discipline as Flow.
Commander and responders
One person owns the response, so nobody wonders who decides. Responders are emailed the moment they're added; handoffs are logged; removing someone keeps their history.
A lifecycle with honest timestamps
Open → Mitigated → Resolved → Closed, each milestone stamped as it's reached. Jump straight to Resolved and the mitigation stamps too; reopen and the later milestones clear.
The change that caused it
Link an incident to the change request that broke things — and the one that fixed them. The change shows every incident it caused, and Analytics turns it into a true change failure rate.
Wired into the board
Link a tracking card, and the card shows the incident's severity and status right in its modal. Follow-up work lives where work lives — on the board.
Post-incident reviews
Closed incidents get a review: what happened, the root cause, and action items — each of which can spawn a board card with one click, linked back to the incident.
Loud where it counts
Flow's dashboard shows an open-incidents banner so change managers see the fire before approving more fuel. MTTM, MTTR, and volume-by-severity trend in Analytics.
Run the next incident on the record
Declare in seconds with just a title — refine severity, team, and links as you learn more.
Roost is in beta — access is invite-only while we build.