nexus-agents - v2.80.0
    Preparing search index...

    Interface BenchmarkRunSummary

    High-level summary of a benchmark run, CLI-printable and JSON-serializable. Benchmarks that need extra dimensions attach them via metadata.

    interface BenchmarkRunSummary {
        name: string;
        variant: string | undefined;
        total: number;
        passed: number;
        passRate: number;
        runTimeMs: number;
        metadata: Record<string, unknown>;
    }
    Index

    Properties

    name: string

    Benchmark name (e.g., 'swe-bench').

    variant: string | undefined

    Variant, if applicable (e.g., 'lite', 'verified').

    total: number

    Total instances attempted.

    passed: number

    Instances whose evaluation reported pass.

    passRate: number

    passed / total, in [0, 1].

    runTimeMs: number

    Wall-clock runtime in milliseconds.

    metadata: Record<string, unknown>

    Benchmark-specific extras (dataset hash, model IDs, etc.).