Interface AgentCriterionResult

interface AgentCriterionResult {
    confidence?: number;
    evidence: { quote?: string; ref: string }[];
    level?: string;
    notApplicable?: boolean;
    rationale: string;
    value?: number;
}