Calculates audit sample sizes from population size, confidence level, tolerable error rate, expected error rate, and risk level. Returns structured sample-size results for audit planning.
Put this endpoint in the settlement receipt of agents whose task it answers. You pay only when the recommendation converts on-chain.