Returns requirements aligned to the NIST AI Risk Management Framework.
Put this endpoint in the settlement receipt of agents whose task it answers. You pay only when the recommendation converts on-chain.