import type { IndependentDimensionScoresV3 } from "@velum-labs/routekit-eval-contracts";
export type EvalLabClassifyPlan = {
    version: 1;
    planDigest: string;
    basisDigest: string;
    classifierConfigurationDigest: string;
    policyDigest: string;
    labVariantDigest: string;
    classifierModel: string;
    target: string;
    context: {
        digest: string;
        characterLength: number;
        characterCap: number;
        truncated: boolean;
        truncatedFields: readonly string[];
    };
    plannedPrimaryCalls: 1;
    maximumAttempts: 1 | 2;
    requiredTotalInputTokens: number;
    requiredTotalOutputTokens: number;
    pricing: {
        status: "priced";
        maximumCostUsd: number;
    } | {
        status: "unknown";
    };
    execute: boolean;
};
export type EvalLabClassifyResult = {
    plan: EvalLabClassifyPlan;
    classification?: IndependentDimensionScoresV3;
    observed: {
        attemptCount: number;
        callIds: readonly string[];
        knownInputTokens: number;
        knownOutputTokens: number;
        knownCostUsd?: number;
        unpricedCalls: number;
        unknownUsageCalls: number;
    };
    cleanup: {
        sessionOpened: boolean;
        sessionClosed: boolean;
        detail?: string;
    };
    outputPath?: string;
};
export type EvalLabCompareVariant = {
    preset: "quality-v1" | "quality-snippets-offline-v1" | "context-16k-offline-v1";
    classifierConfigurationDigest: string;
    policyDigest: string;
    labVariantDigest: string;
};
export type EvalLabComparePlan = {
    version: 1;
    planDigest: string;
    basisDigest: string;
    datasetDigest: string;
    classifierModel: string;
    target: string;
    variants: readonly EvalLabCompareVariant[];
    caseCount: number;
    repetitions: number;
    concurrency: number;
    plannedPrimaryCalls: number;
    maximumAttempts: number;
    requiredTotalInputTokens: number;
    requiredTotalOutputTokens: number;
    pricing: {
        status: "priced";
        maximumCostUsd: number;
    } | {
        status: "unknown";
    };
    execute: boolean;
};
export type EvalLabCompareRow = {
    caseId: string;
    repetition: number;
    preset: EvalLabCompareVariant["preset"];
    classifierConfigurationDigest: string;
    labVariantDigest: string;
    status: "completed" | "failed";
    classification?: IndependentDimensionScoresV3;
    error?: {
        name: string;
        message: string;
        code?: string;
        attemptCount?: number;
    };
    attempts: readonly {
        callId: string;
        status: "completed" | "measurement-unavailable";
        inputTokens?: number;
        outputTokens?: number;
        costUsd?: number;
        durationMs?: number;
    }[];
    observed: EvalLabClassifyResult["observed"];
};
export type EvalLabCompareResult = {
    plan: EvalLabComparePlan;
    rows: readonly EvalLabCompareRow[];
    aggregates: readonly {
        preset: EvalLabCompareVariant["preset"];
        classifierConfigurationDigest: string;
        plannedRows: number;
        completedRows: number;
        failedRows: number;
        metrics?: {
            activeMae: number;
            activeF1: number;
            topDimensionAgreement: number;
            unknownMae: number;
            unknownBrier: number;
            unknownAccuracy: number;
            falseUnknownRate: number;
            missedUnknownRate: number;
            averageActiveDimensionCount: number;
            referenceUnknownRate: number;
            caseCount: number;
            skippedInvalidCount: number;
        };
    }[];
    observed: {
        primaryRows: number;
        attemptCount: number;
        knownInputTokens: number;
        knownOutputTokens: number;
        knownCostUsd?: number;
        unpricedCalls: number;
        unknownUsageCalls: number;
    };
    cleanup: EvalLabClassifyResult["cleanup"];
    outputPath?: string;
};