{
  "protocol": {
    "benchmark": "swebench-local",
    "episodes": 12,
    "seed": 42,
    "sandbox": "temporary local Python repository + fixed python -m unittest -q",
    "scope": "repository-owned executable micro fixtures; not official SWE-bench Lite"
  },
  "results": {
    "metagpt": {
      "joint_success": 1.0,
      "average_cost": 3.5,
      "baseline_failures": 12,
      "subprocess_commands": 24,
      "file_edits": 12,
      "role_messages": 48
    },
    "critic": {
      "joint_success": 1.0,
      "average_cost": 4.0,
      "baseline_failures": 12,
      "subprocess_commands": 36,
      "file_edits": 24,
      "critic_rounds": 24
    },
    "agent-lightning": {
      "joint_success": 1.0,
      "average_cost": 3.625,
      "reuse_rate": 0.75,
      "baseline_failures": 12,
      "subprocess_commands": 36,
      "file_edits": 15,
      "credit_updates": 15,
      "learned_bug_families": 3
    },
    "swe-agent": {
      "joint_success": 1.0,
      "average_cost": 2.5,
      "baseline_failures": 12,
      "subprocess_commands": 24,
      "file_edits": 12
    },
    "openhands": {
      "joint_success": 1.0,
      "average_cost": 2.5,
      "baseline_failures": 12,
      "subprocess_commands": 24,
      "file_edits": 12
    }
  },
  "schema_version": 2,
  "manifest_ref": "experiments:agent-code-sandbox-seed42",
  "evaluation_protocol": {
    "tier": "l1_mechanism",
    "seeds": [
      42
    ],
    "formal_comparison": false,
    "claim_policy": "single/few-seed smoke result; do not claim a stable improvement"
  },
  "provenance": {
    "artifact_path": "docs/experiments/agent-code-sandbox-seed42.json",
    "historical_migration": "historical-metrics-v2-2026-08-09",
    "original_code_commit": "not recorded",
    "dataset_fingerprint": "not recorded in historical artifact"
  }
}
