MCP Agent TypeScript Port

/** * Simple benchmarking scaffolding to compare workflows or model outputs. */ export interface EvaluationCase { id: string; input: unknown; expected?: unknown; } export interface Workflow { name: string; run: (input: unknown) => Promise<unknown> | unknown; } export interface CaseResult { caseId: string; output: unknown; expected?: unknown; durationMs: number; success: boolean; } export interface WorkflowResult { workflow: string; results: CaseResult[]; } /** * Execute a set of workflows against evaluation cases and capture timing/accuracy metrics. */ export declare function benchmarkWorkflows(workflows: Workflow[], cases: EvaluationCase[]): Promise<WorkflowResult[]>; /** * Run a trivial default evaluation used by CLI for demonstration. */ export declare function runDefaultEvaluation(): Promise<WorkflowResult[]>;