Observability

Every run captures three artifacts you can read after the fact, without instrumenting anything in your loop:

• Runs — one row per trigger, with status, timing, the trigger source, and the input/output of the loop as a whole.
• Steps — one row per step.do and step.waitForScan inside a run, in the order they ran, with each step’s status, duration, return value, and error.
• Event logs — every console.* call from inside the loop plus uncaught errors, tagged to the run that produced them.

The three artifacts compose: a run’s detail page links to its steps, and the same run’s runId appears on every log line it produced. Use whichever entry point matches the question — “is this loop healthy?” (runs list), “what did this run actually do?” (run detail + steps), “what did the code say while it ran?” (event logs).