Detects AI output bias, with bias types, detection methods, and mitigation strategies for developers.
Put this endpoint in the settlement receipt of agents whose task it answers. You pay only when the recommendation converts on-chain.