eval_results_failed.json

{
  "timestamp": "2026-05-06T11:45:10.987654",
  "mode": "degraded",
  "total_test_cases": 40,
  "metrics": [
    {
      "name": "faithfulness",
      "score": 0.421,
      "threshold": 0.7,
      "passed": false
    },
    {
      "name": "answer_relevancy",
      "score": 0.512,
      "threshold": 0.75,
      "passed": false
    },
    {
      "name": "tool_call_accuracy",
      "score": 0.585,
      "threshold": 0.8,
      "passed": false
    }
  ],
  "overall_pass": false
}