Scores the logical validity of an argument from its premises and conclusion.
Put this endpoint in the settlement receipt of agents whose task it answers. You pay only when the recommendation converts on-chain.