Moderates content for policy violations and unsafe material.
Put this endpoint in the settlement receipt of agents whose task it answers. You pay only when the recommendation converts on-chain.