backend/evaluation/results/baseline_v2.json

{
    "note": "Placeholder — run `python -m evaluation.runner` to generate baseline results",
    "total_queries": 0,
    "exact_match_rate": 0,
    "execution_accuracy": 0,
    "avg_structural_similarity": 0,
    "total_hallucinations": 0,
    "avg_latency_ms": 0,
    "results": []
}