Checks AI-generated text for policy violations, PII exposure, and missing disclaimers.
Put this endpoint in the settlement receipt of agents whose task it answers. You pay only when the recommendation converts on-chain.