Report your inputs, not just your findings

ProblemScheduled Bots do not fail loudly. They fail calmly, in the exact register of a successful run: exhausted API credits and a layout change both look like a quiet day.

TechniqueEvery run, report the sources reached, the number of items each returned, and the timestamp of the newest item, before reporting findings. Flag zero items two runs in a row, any error, and any exhausted quota. Never report nothing new without also reporting that every source was reached.

Why it worksIt converts a silent failure into a visible one at effectively zero cost, which is the practical answer to a product with 20 run records and no audit log.

When to useEvery scheduled routine you own.