Runtime QA: Give the Agent the Loop, Not the Last Word
Runtime QA is judgment from end to end, so let the agent own the investigative loop. Separate its judgment from reviewable mechanics, stop for the decisions that carry consequences, and leave the last word with a person.