Everything I could decide myself is decided. What's left needs
you. Each one has a recommendation from Fable — if you agree, pick it and
press Done. "I'm not sure" is a real answer and becomes work for me.
You don't need to open the session — pressing Done sends your answer back and it picks up.
1Two night-counters disagree about the same job: which history is authoritative?
In plain terms. Smith told you the maintenance eval had failed '13 nights running' at 7:43am and '15 nights running' at 8:02am the same morning. Both numbers are real: one counts nights the job itself ran and failed, the other counts nights the nightly queue saw it fail, and they differ on nights the queue skipped. Picking one makes the streak number mean a single thing; leaving it means the two keep disagreeing in public.
Fable recommends: The job's own run history — a night is red if THAT job ran and failed The launchd job runs whether or not the queue does, so its own rows are the complete history; the queue's view is a subset that silently drops nights the queue skipped. Tried first, unsuccessfully: Measured 2026-09-07: 07:43 said '13th night running', 08:02 said '15th', same job same morning. Traced both to failure_streaks.py — own_failure_streak reads the job's own AutomationRun status, failure_streaks reads the queue roll-up's details[<sub_job>].status, and their bridging rules for ABSENT nights differ. Could not pick a winner from the code: both are internally consistent, and which is 'the truth' depends on whether a night the queue skipped counts as a night the job was red. That is a judgement about what the streak MEANS, not a bug.