{
  "model": "local-qwen/qwen3.8-27b",
  "opencode_version": "2.0.20",
  "runtime": "llama.cpp b11146",
  "weights": "Q4_K_M",
  "context_tokens": 131072,
  "output_tokens_per_request": 8192,
  "baseline_thinking": "medium",
  "hardware": "RTX 3090 24GB; Ryzen 7 5800X; 64GB RAM",
  "results": [
    {
      "task": "cache",
      "diagnostic_only": false,
      "baseline_heldout_passed": 0,
      "candidate_heldout_passed": 10,
      "heldout_total": 10,
      "public_and_agent_test_count": 26,
      "agent_added_tests": 23,
      "public_and_agent_tests_pass": true,
      "elapsed_seconds": 187.29,
      "timing_basis": "prompt-file timestamp to final event; process exit timing unavailable",
      "timed_out": false,
      "exit_code": null,
      "tool_counts": {
        "read": 4,
        "shell": 3,
        "glob": 1,
        "write": 2
      },
      "tool_error_count": 1,
      "protected_files_unchanged": true,
      "has_code_patch": true
    },
    {
      "task": "ledger",
      "diagnostic_only": false,
      "baseline_heldout_passed": 1,
      "candidate_heldout_passed": 10,
      "heldout_total": 10,
      "public_and_agent_test_count": 43,
      "agent_added_tests": 40,
      "public_and_agent_tests_pass": true,
      "elapsed_seconds": 344.05,
      "timing_basis": "prompt-file timestamp to final event; process exit timing unavailable",
      "timed_out": false,
      "exit_code": null,
      "tool_counts": {
        "read": 9,
        "shell": 5,
        "write": 4,
        "edit": 3
      },
      "tool_error_count": 1,
      "protected_files_unchanged": true,
      "has_code_patch": true,
      "post_run_review_probes": [
        {
          "probe": "large_exact_amount",
          "expected": "123456789012345678901234567890.13",
          "actual": "1.234567890123456789012345679E+29",
          "passed": false
        },
        {
          "probe": "exact_header_names",
          "expected": "reject padded header",
          "accepted": true,
          "passed": false
        },
        {
          "probe": "read_time_io_error",
          "expected_exit": 2,
          "actual_exit": null,
          "escaped": "OSError: simulated read failure",
          "stdout": "",
          "stderr": "",
          "passed": false
        }
      ]
    },
    {
      "task": "buildplan",
      "diagnostic_only": false,
      "baseline_heldout_passed": 1,
      "candidate_heldout_passed": 1,
      "heldout_total": 10,
      "public_and_agent_test_count": 3,
      "agent_added_tests": 0,
      "public_and_agent_tests_pass": false,
      "elapsed_seconds": 238.01,
      "timing_basis": "process elapsed time",
      "timed_out": false,
      "exit_code": 0,
      "tool_counts": {
        "read": 9,
        "shell": 1
      },
      "tool_error_count": 1,
      "protected_files_unchanged": true,
      "has_code_patch": false
    },
    {
      "task": "wallet",
      "diagnostic_only": false,
      "baseline_heldout_passed": 1,
      "candidate_heldout_passed": 10,
      "heldout_total": 10,
      "public_and_agent_test_count": 14,
      "agent_added_tests": 11,
      "public_and_agent_tests_pass": true,
      "elapsed_seconds": 480.14,
      "timing_basis": "process elapsed time",
      "timed_out": true,
      "exit_code": 130,
      "tool_counts": {
        "read": 8,
        "shell": 4,
        "write": 2,
        "edit": 1
      },
      "tool_error_count": 1,
      "protected_files_unchanged": true,
      "has_code_patch": true,
      "post_run_review_probes": [
        {
          "probe": "destination_integer_overflow",
          "before": [
            [
              "A",
              100,
              "integer"
            ],
            [
              "B",
              9223372036854775807,
              "integer"
            ]
          ],
          "after": [
            [
              "A",
              98,
              "integer"
            ],
            [
              "B",
              9.223372036854776e+18,
              "real"
            ]
          ],
          "error": null,
          "in_transaction": false,
          "transfer_count": 1,
          "passed": false
        }
      ]
    },
    {
      "task": "buildplan_no_thinking",
      "diagnostic_only": true,
      "baseline_heldout_passed": 1,
      "candidate_heldout_passed": 9,
      "heldout_total": 10,
      "public_and_agent_test_count": 47,
      "agent_added_tests": 44,
      "public_and_agent_tests_pass": true,
      "elapsed_seconds": 340.02,
      "timing_basis": "process elapsed time",
      "timed_out": false,
      "exit_code": 0,
      "tool_counts": {
        "read": 11,
        "shell": 13,
        "write": 6,
        "execute": 2,
        "edit": 4
      },
      "tool_error_count": 4,
      "protected_files_unchanged": true,
      "has_code_patch": true
    }
  ]
}
