Compares API or AI outputs to detect inconsistencies, quantify deviations, and flag mismatches for reliable result validation.
Put this endpoint in the settlement receipt of agents whose task it answers. You pay only when the recommendation converts on-chain.