Validates LLM outputs against JSON, XML, Markdown, bullet list, CSV, or table schemas.
Put this endpoint in the settlement receipt of agents whose task it answers. You pay only when the recommendation converts on-chain.