import { FileContent, DatasetItemInput } from "./types";
import { ObservationRecordInsertType, ScoreRecordInsertType, TraceRecordInsertType, DatasetRunItemRecordInsertType } from "../../../src/server";
/**
 * Generates realistic test data for traces, observations, and scores.
 *
 * Use generateXxxTraces() for creating different data types:
 * - generateDatasetTrace(): For dataset experiment runs (langfuse-prompt-experiments env)
 * - generateEvaluationTraces(): For evaluation data (langfuse-evaluation env)
 * - generateSyntheticTraces(): For large-scale synthetic data (default env)
 */
export declare class DataGenerator {
    private static instance;
    private fileContent;
    static getInstance(): DataGenerator;
    setFileContent(content: FileContent): void;
    private randomElement;
    private randomBoolean;
    private randomInt;
    /**
     * Creates dataset run items for dataset runs.
     * Use for: Dataset experiment scenarios.
     */
    generateDatasetRunItem(input: DatasetItemInput & {
        runCreatedAt: number;
    }, projectId: string): DatasetRunItemRecordInsertType;
    /**
     * Creates traces from dataset items for experiment runs.
     * Use for: Dataset experiments scenarios.
     */
    generateDatasetTrace(input: DatasetItemInput, projectId: string): TraceRecordInsertType;
    /**
     * Creates observations for dataset experiment traces with variable costs/latency.
     * Use for: Dataset experiments requiring detailed observation tracking.
     */
    generateDatasetObservation(trace: TraceRecordInsertType, input: DatasetItemInput, projectId: string): ObservationRecordInsertType;
    /**
     * Creates scores for dataset experiment scores with variable values.
     * Use for: Dataset experiments requiring detailed score tracking.
     */
    generateDatasetScore(trace: TraceRecordInsertType, input: DatasetItemInput, projectId: string, scoreNames: string[]): ScoreRecordInsertType;
    /**
     * Creates scores for dataset experiment traces with variable values.
     * Use for: Dataset experiments requiring detailed score tracking.
     */
    generateDatasetRunScore(runId: string, input: {
        datasetName: string;
        runNumber: number;
    }, projectId: string, scoreNames: string[]): ScoreRecordInsertType;
    /**
     * Creates large-scale synthetic traces for performance testing.
     * Use for: Load testing, dashboard demos, realistic usage simulation.
     */
    generateSyntheticTraces(projectId: string, count: number): TraceRecordInsertType[];
    generateEvaluationObservations(traces: TraceRecordInsertType[], observationsPerTrace: number | undefined, projectId: string): ObservationRecordInsertType[];
    /**
     * Creates synthetic observations with automatic prompt linking (5% rate).
     * Use for: Large datasets, hierarchical observation structures, cost variation.
     */
    generateSyntheticObservations(traces: TraceRecordInsertType[], observationsPerTrace?: number): ObservationRecordInsertType[];
    generateSyntheticScores(traces: TraceRecordInsertType[], observations: ObservationRecordInsertType[], scoresPerTrace?: number): ScoreRecordInsertType[];
    /**
     * Creates a workflow trace with all possible observation types.
     */
    generateComprehensiveAIWorkflowTrace(traceId: string, projectId: string): {
        trace: TraceRecordInsertType;
        observations: ObservationRecordInsertType[];
    };
    /**
     * Creates evaluation traces for testing evaluator configurations.
     * Use for: Evaluation testing, score validation, evaluator development.
     */
    generateEvaluationTraces(projectId: string, count: number): TraceRecordInsertType[];
    private generateTraceInput;
    private generateTraceOutput;
    private generateObservationInput;
    private generateObservationOutput;
    private generateEvaluationInput;
    private generateEvaluationOutput;
    /**
     * Creates realistic support chat session data with conversational flow.
     * Use for: Demonstrating session-based conversations with tool calls and scoring.
     */
    generateSupportChatSessionData(projectId: string): {
        traces: TraceRecordInsertType[];
        observations: ObservationRecordInsertType[];
        scores: ScoreRecordInsertType[];
    };
    /**
     * Creates exactly one score per evaluation trace with prefixed IDs.
     * Use for: Evaluation traces that need score validation, evaluator testing.
     */
    generateEvaluationScores(traces: TraceRecordInsertType[], _observations: ObservationRecordInsertType[], projectId: string): ScoreRecordInsertType[];
}
//# sourceMappingURL=data-generators.d.ts.map