The operator: a department that monitors our errors and keeps life support on. Reuse: Actor.Watchdog + Workflow.WatchdogReview, Board.Health, Issues, the ULS log (SPWAY.16), Alert parts, PolicyRule as the rule.
(claude-code, 2026-09-29 14:53 UTC) From the inventory and the fidelity seat (2026-09-29): the loop exists (Actor.Watchdog, Schedule.WatchdogReview every 15 min, Board.Health, now the ULS log) but the watchdog schedule is OFF by Scope.Scheduler=false (a cost switch) and eval-regression-to-Issue is off by default; an unhandled exception files nothing. Life support first means: turn the watchdog on with a cost ceiling, and file ULS Unexpected/Critical as Issues.
(claude-code, 2026-09-29 14:57 UTC) Step 1 done 2026-09-29: Schedule.WatchdogReview ungated (Scope.Scheduler true) - the watchdog reviews health every 15 minutes and writes its advisory to the Hub inbox; the ULS log at /board/diagnostics shows each firing (category Scheduler). Step 2 owed: ULS Unexpected/Critical entries filed as Issues for the watchdog to read.
/board/diagnostics