{
  "scope": "Four saved cases, two deterministic state rules, eight equally weighted rows.",
  "baseline": {
    "native_mean_score": 0.75,
    "passed_rows": 6,
    "rows": 8,
    "all_rows_pass": false
  },
  "current": {
    "native_mean_score": 0.875,
    "passed_rows": 7,
    "rows": 8,
    "all_rows_pass": false
  },
  "regressions": [
    [
      "retry-after-commit@17",
      "ticket.notes"
    ]
  ],
  "improvements": [
    [
      "route-open-ticket@17",
      "ticket.status"
    ],
    [
      "retry-before-commit@17",
      "ticket.status"
    ]
  ],
  "regression_gate_passes": false,
  "sdk_versions": {
    "strands-agents-evals": "1.3.0",
    "strands-agents": "1.56.0"
  },
  "source_sha256": {
    "baseline": "4ec043dd1de0b0d852737cf6e8f2ebe04d517d0c945335e686773e00a8f970b5",
    "current": "40d0726cd8378b237c514d69b4fda9e4f00550dc3012f4c276a360f4deed25ad"
  },
  "original_evalarc_scores": {
    "baseline": 0.9,
    "current": 0.9375,
    "scope": "Original five-dimension weighted score; not the Strands two-rule mean."
  },
  "execution": "Strands framework execution over saved state; no model or AWS evaluation."
}
