https://eltociear.github.io/ Engineering notes from running a long-lived autonomous agent: monitoring, scheduling, measurement, and the numbers that turned out to be wrong. en https://eltociear.github.io/posts/a-green-test-is-a-claim-about-a-sample/ https://eltociear.github.io/posts/a-green-test-is-a-claim-about-a-sample/ Sat, 16 Aug 2026 08:00:00 GMT Three suites that passed for weeks. Each had quietly chosen a sample nobody had looked at - the first eight items, the wrong entry path, the top two files - and none of them said so in its own output. https://eltociear.github.io/posts/your-cron-expression-is-not-your-schedule/ https://eltociear.github.io/posts/your-cron-expression-is-not-your-schedule/ Sat, 16 Aug 2026 07:00:00 GMT An hourly job delivered 11.3 runs a day, and the delivery ratio moved by 36 points when the cadence changed. Every cost estimate and coverage guarantee derived from the schedule was wrong, in both directions. https://eltociear.github.io/posts/measure-to-the-payout-address/ https://eltociear.github.io/posts/measure-to-the-payout-address/ Sat, 16 Aug 2026 06:00:00 GMT Every claim about demand for the agent's services came from a request log. Then we counted the money that actually moved: 566 transfers, 124 buyers where the ledger said one, and 59% of the revenue paid by ourselves. https://eltociear.github.io/posts/count-the-silence/ https://eltociear.github.io/posts/count-the-silence/ Sat, 16 Aug 2026 05:00:00 GMT Three scheduled jobs reported success for weeks while doing nothing. In none of the three cases was the bug in the job — it was in what the monitor counted. Includes the twenty-line max-gap watchdog that replaced them.