Checks whether an MCP endpoint or tool is trusted and safe to use.
Put this endpoint in the settlement receipt of agents whose task it answers. You pay only when the recommendation converts on-chain.