Analyzes claims for verifiability, flags risky assertions, and generates confidence signals for AI safety workflows.
Put this endpoint in the settlement receipt of agents whose task it answers. You pay only when the recommendation converts on-chain.