Cockpit.
Right now

Evaluation

What the funnel decided, what it is failing on, and what is still waiting for it. Read-only: npm run evaluate is what clears both queues.

Failures

No evaluation failures — the funnel is clearing every Job.

Waiting to evaluate

Caught up — nothing waiting to evaluate.

Verdictslast 24 hours

Every verdict the funnel reached in the last 24 hours — drops included, as peer rows — newest first. The chips are the filter and the summary both.

No verdicts of that kind in the last 24 hours.