Fleet Run Monitor

by Jonas

Fleet Run Monitor reads the run export from your AI agents and pulls out every failed, timed-out, or drifting run, grouping repeat failures so you can see which agent needs attention first. It hands you a one-glance health read plus incident-log rows, each tied to the run id and line it came from. Everything stays a draft for your review, and it never restarts, changes, or reclassifies a run on its own.