Posted inBlog
What It Takes to Know If Your AI Agent Actually Works
For a while, the standard way to check whether an AI agent was working was to run it a few times, watch the outputs, and decide it looked reasonable. That…
Posted inBlog
Closing the Loop: How Production Observability Makes AI Agents Better
For a long time, teams building AI agents measured quality the same way they measured conventional software — with tests that ran before deployment. Offline evals (structured tests run against…
Posted inBlog
Your Agent Dashboard Is Green. The Run Underneath It Is Where the Work Actually Broke.
Something broke in April 2026 that continues to matter. Jer Crane, founder of PocketOS, watched a Cursor agent delete his company's production database — every table, every row — along…