[
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.1169714,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 7031,
  "cache_read": 42417,
  "output": 2610,
  "is_error": false,
  "input_total": 49458,
  "agent": "claude",
  "job": "analysis",
  "run": 1,
  "seconds": 26.5,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.1226766,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 7200,
  "cache_read": 42583,
  "output": 2826,
  "is_error": false,
  "input_total": 49793,
  "agent": "claude",
  "job": "analysis",
  "run": 2,
  "seconds": 28.0,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.11039860000000001,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 6413,
  "cache_read": 41073,
  "output": 2542,
  "is_error": false,
  "input_total": 47496,
  "agent": "claude",
  "job": "analysis",
  "run": 3,
  "seconds": 25.8,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.1085114,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 6276,
  "cache_read": 40917,
  "output": 2504,
  "is_error": false,
  "input_total": 47203,
  "agent": "claude",
  "job": "analysis",
  "run": 4,
  "seconds": 26.3,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.1099074,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 6678,
  "cache_read": 42017,
  "output": 2402,
  "is_error": false,
  "input_total": 48705,
  "agent": "claude",
  "job": "analysis",
  "run": 5,
  "seconds": 25.0,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 3,
  "model": "claude-opus-5-5",
  "cost_usd": 0.0818394,
  "turns": 4,
  "input_uncached": 8,
  "cache_write": 4939,
  "cache_read": 35177,
  "output": 1763,
  "is_error": false,
  "input_total": 40124,
  "agent": "claude",
  "job": "bughunt",
  "run": 1,
  "seconds": 20.5,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK"
 },
 {
  "tool_calls": 3,
  "model": "claude-opus-5-5",
  "cost_usd": 0.0960798,
  "turns": 4,
  "input_uncached": 8,
  "cache_write": 6640,
  "cache_read": 32539,
  "output": 1821,
  "is_error": false,
  "input_total": 39187,
  "agent": "claude",
  "job": "bughunt",
  "run": 2,
  "seconds": 19.7,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK"
 },
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.10888220000000001,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 7256,
  "cache_read": 43371,
  "output": 2106,
  "is_error": false,
  "input_total": 50637,
  "agent": "claude",
  "job": "bughunt",
  "run": 3,
  "seconds": 24.3,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK"
 },
 {
  "tool_calls": 3,
  "model": "claude-opus-5-5",
  "cost_usd": 0.0989078,
  "turns": 4,
  "input_uncached": 8,
  "cache_write": 7196,
  "cache_read": 32939,
  "output": 1736,
  "is_error": false,
  "input_total": 40143,
  "agent": "claude",
  "job": "bughunt",
  "run": 4,
  "seconds": 19.2,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK"
 },
 {
  "tool_calls": 3,
  "model": "claude-opus-5-5",
  "cost_usd": 0.10418440000000001,
  "turns": 4,
  "input_uncached": 8,
  "cache_write": 7233,
  "cache_read": 32942,
  "output": 1985,
  "is_error": false,
  "input_total": 40183,
  "agent": "claude",
  "job": "bughunt",
  "run": 5,
  "seconds": 21.8,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK"
 },
 {
  "tool_calls": 8,
  "model": "claude-opus-5-5",
  "cost_usd": 0.20634479999999997,
  "turns": 9,
  "input_uncached": 18,
  "cache_write": 11301,
  "cache_read": 109724,
  "output": 4696,
  "is_error": false,
  "input_total": 121043,
  "agent": "claude",
  "job": "clean",
  "run": 1,
  "seconds": 45.9,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "300 rows (want 300), 300 exact, sorted=True"
 },
 {
  "tool_calls": 3,
  "model": "claude-opus-5-5",
  "cost_usd": 0.113243,
  "turns": 4,
  "input_uncached": 8,
  "cache_write": 7243,
  "cache_read": 33435,
  "output": 2429,
  "is_error": false,
  "input_total": 40686,
  "agent": "claude",
  "job": "clean",
  "run": 2,
  "seconds": 24.4,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "300 rows (want 300), 300 exact, sorted=True"
 },
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.1201222,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 7446,
  "cache_read": 44671,
  "output": 2579,
  "is_error": false,
  "input_total": 52127,
  "agent": "claude",
  "job": "clean",
  "run": 3,
  "seconds": 26.8,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "300 rows (want 300), 300 exact, sorted=True"
 },
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.1308008,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 7894,
  "cache_read": 46044,
  "output": 2920,
  "is_error": false,
  "input_total": 53948,
  "agent": "claude",
  "job": "clean",
  "run": 4,
  "seconds": 28.7,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "300 rows (want 300), 300 exact, sorted=True"
 },
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.13232919999999998,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 7987,
  "cache_read": 45566,
  "output": 2964,
  "is_error": false,
  "input_total": 53563,
  "agent": "claude",
  "job": "clean",
  "run": 5,
  "seconds": 29.6,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "300 rows (want 300), 300 exact, sorted=True"
 },
 {
  "tool_calls": 7,
  "model": "claude-opus-5-5",
  "cost_usd": 0.221933,
  "turns": 8,
  "input_uncached": 16,
  "cache_write": 17516,
  "cache_read": 111905,
  "output": 2968,
  "is_error": false,
  "input_total": 129437,
  "agent": "claude",
  "job": "corpus",
  "run": 1,
  "seconds": 32.3,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 5,
  "model": "claude-opus-5-5",
  "cost_usd": 0.2135374,
  "turns": 6,
  "input_uncached": 12,
  "cache_write": 18664,
  "cache_read": 84087,
  "output": 2368,
  "is_error": false,
  "input_total": 102763,
  "agent": "claude",
  "job": "corpus",
  "run": 2,
  "seconds": 26.5,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 5,
  "model": "claude-opus-5-5",
  "cost_usd": 0.2251782,
  "turns": 6,
  "input_uncached": 12,
  "cache_write": 20488,
  "cache_read": 81131,
  "output": 2250,
  "is_error": false,
  "input_total": 101631,
  "agent": "claude",
  "job": "corpus",
  "run": 3,
  "seconds": 24.3,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 5,
  "model": "claude-opus-5-5",
  "cost_usd": 0.20805440000000003,
  "turns": 6,
  "input_uncached": 12,
  "cache_write": 18344,
  "cache_read": 83572,
  "output": 2227,
  "is_error": false,
  "input_total": 101928,
  "agent": "claude",
  "job": "corpus",
  "run": 4,
  "seconds": 24.7,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.21705860000000002,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 20670,
  "cache_read": 71393,
  "output": 1869,
  "is_error": false,
  "input_total": 92073,
  "agent": "claude",
  "job": "corpus",
  "run": 5,
  "seconds": 20.5,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 2,
  "model": "claude-opus-5-5",
  "cost_usd": 0.10717840000000001,
  "turns": 3,
  "input_uncached": 6,
  "cache_write": 5052,
  "cache_read": 25192,
  "output": 3085,
  "is_error": false,
  "input_total": 30250,
  "agent": "claude",
  "job": "feature",
  "run": 1,
  "seconds": 29.3,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; 15 visible tests (needs new ones)"
 },
 {
  "tool_calls": 2,
  "model": "claude-opus-5-5",
  "cost_usd": 0.1334982,
  "turns": 3,
  "input_uncached": 6,
  "cache_write": 8374,
  "cache_read": 22911,
  "output": 3095,
  "is_error": false,
  "input_total": 31291,
  "agent": "claude",
  "job": "feature",
  "run": 2,
  "seconds": 29.8,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; 16 visible tests (needs new ones)"
 },
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.1523186,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 8977,
  "cache_read": 47413,
  "output": 3549,
  "is_error": false,
  "input_total": 56400,
  "agent": "claude",
  "job": "feature",
  "run": 3,
  "seconds": 33.4,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; 16 visible tests (needs new ones)"
 },
 {
  "tool_calls": 6,
  "model": "claude-opus-5-5",
  "cost_usd": 0.15971459999999998,
  "turns": 7,
  "input_uncached": 10,
  "cache_write": 8851,
  "cache_read": 46333,
  "output": 3980,
  "is_error": false,
  "input_total": 55194,
  "agent": "claude",
  "job": "feature",
  "run": 4,
  "seconds": 38.4,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; 15 visible tests (needs new ones)"
 },
 {
  "tool_calls": 3,
  "model": "claude-opus-5-5",
  "cost_usd": 0.14221,
  "turns": 4,
  "input_uncached": 8,
  "cache_write": 7960,
  "cache_read": 35290,
  "output": 3572,
  "is_error": false,
  "input_total": 43258,
  "agent": "claude",
  "job": "feature",
  "run": 5,
  "seconds": 33.0,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; 15 visible tests (needs new ones)"
 },
 {
  "tool_calls": 5,
  "model": "claude-opus-5-5",
  "cost_usd": 0.35257659999999996,
  "turns": 6,
  "input_uncached": 12,
  "cache_write": 36805,
  "cache_read": 111943,
  "output": 1785,
  "is_error": false,
  "input_total": 148760,
  "agent": "claude",
  "job": "longdoc",
  "run": 1,
  "seconds": 22.3,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "12/12 right; wrong []"
 },
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.3393764,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 36763,
  "cache_read": 71862,
  "output": 1543,
  "is_error": false,
  "input_total": 108635,
  "agent": "claude",
  "job": "longdoc",
  "run": 2,
  "seconds": 18.9,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "12/12 right; wrong []"
 },
 {
  "tool_calls": 5,
  "model": "claude-opus-5-5",
  "cost_usd": 0.35113579999999994,
  "turns": 6,
  "input_uncached": 12,
  "cache_write": 36686,
  "cache_read": 111799,
  "output": 1762,
  "is_error": false,
  "input_total": 148497,
  "agent": "claude",
  "job": "longdoc",
  "run": 3,
  "seconds": 21.1,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "12/12 right; wrong []"
 },
 {
  "tool_calls": 5,
  "model": "claude-opus-5-5",
  "cost_usd": 0.150094,
  "turns": 6,
  "input_uncached": 12,
  "cache_write": 12132,
  "cache_read": 63350,
  "output": 2016,
  "is_error": false,
  "input_total": 75494,
  "agent": "claude",
  "job": "longdoc",
  "run": 4,
  "seconds": 22.7,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "12/12 right; wrong []"
 },
 {
  "tool_calls": 5,
  "model": "claude-opus-5-5",
  "cost_usd": 0.35650519999999997,
  "turns": 6,
  "input_uncached": 12,
  "cache_write": 36794,
  "cache_read": 132426,
  "output": 1781,
  "is_error": false,
  "input_total": 169232,
  "agent": "claude",
  "job": "longdoc",
  "run": 5,
  "seconds": 20.8,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "12/12 right; wrong []"
 },
 {
  "tool_calls": 2,
  "model": "claude-opus-5-5",
  "cost_usd": 0.0773292,
  "turns": 3,
  "input_uncached": 6,
  "cache_write": 4213,
  "cache_read": 24906,
  "output": 1931,
  "is_error": false,
  "input_total": 29125,
  "agent": "claude",
  "job": "refactor",
  "run": 1,
  "seconds": 19.8,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; send_email defined in ['email.py']"
 },
 {
  "tool_calls": 2,
  "model": "claude-opus-5-5",
  "cost_usd": 0.091943,
  "turns": 3,
  "input_uncached": 6,
  "cache_write": 5986,
  "cache_read": 22555,
  "output": 1976,
  "is_error": false,
  "input_total": 28547,
  "agent": "claude",
  "job": "refactor",
  "run": 2,
  "seconds": 20.5,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; send_email defined in ['email.py']"
 },
 {
  "tool_calls": 2,
  "model": "claude-opus-5-5",
  "cost_usd": 0.09472420000000001,
  "turns": 3,
  "input_uncached": 6,
  "cache_write": 6181,
  "cache_read": 22561,
  "output": 2037,
  "is_error": false,
  "input_total": 28748,
  "agent": "claude",
  "job": "refactor",
  "run": 3,
  "seconds": 21.7,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; send_email defined in ['email.py']"
 },
 {
  "tool_calls": 2,
  "model": "claude-opus-5-5",
  "cost_usd": 0.09082879999999999,
  "turns": 3,
  "input_uncached": 6,
  "cache_write": 5979,
  "cache_read": 22564,
  "output": 1923,
  "is_error": false,
  "input_total": 28549,
  "agent": "claude",
  "job": "refactor",
  "run": 4,
  "seconds": 20.0,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; send_email defined in ['email.py']"
 },
 {
  "tool_calls": 2,
  "model": "claude-opus-5-5",
  "cost_usd": 0.092003,
  "turns": 3,
  "input_uncached": 6,
  "cache_write": 5996,
  "cache_read": 22555,
  "output": 1975,
  "is_error": false,
  "input_total": 28557,
  "agent": "claude",
  "job": "refactor",
  "run": 5,
  "seconds": 20.8,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; send_email defined in ['email.py']"
 },
 {
  "tool_calls": 5,
  "model": "claude-opus-5-5",
  "cost_usd": 0.17843739999999997,
  "turns": 6,
  "input_uncached": 12,
  "cache_write": 9987,
  "cache_read": 59367,
  "output": 4331,
  "is_error": false,
  "input_total": 69366,
  "agent": "claude",
  "job": "replies",
  "run": 1,
  "seconds": 41.4,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "14/14 right; wrong []"
 },
 {
  "tool_calls": 6,
  "model": "claude-opus-5-5",
  "cost_usd": 0.20178079999999998,
  "turns": 7,
  "input_uncached": 14,
  "cache_write": 10763,
  "cache_read": 74104,
  "output": 5040,
  "is_error": false,
  "input_total": 84881,
  "agent": "claude",
  "job": "replies",
  "run": 2,
  "seconds": 49.9,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "14/14 right; wrong []"
 },
 {
  "tool_calls": 5,
  "model": "claude-opus-5-5",
  "cost_usd": 0.18785800000000002,
  "turns": 6,
  "input_uncached": 12,
  "cache_write": 10292,
  "cache_read": 59870,
  "output": 4675,
  "is_error": false,
  "input_total": 70174,
  "agent": "claude",
  "job": "replies",
  "run": 3,
  "seconds": 45.2,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "14/14 right; wrong []"
 },
 {
  "tool_calls": 3,
  "model": "claude-opus-5-5",
  "cost_usd": 0.141362,
  "turns": 4,
  "input_uncached": 8,
  "cache_write": 8199,
  "cache_read": 32490,
  "output": 3462,
  "is_error": false,
  "input_total": 40697,
  "agent": "claude",
  "job": "replies",
  "run": 4,
  "seconds": 33.6,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "14/14 right; wrong []"
 },
 {
  "tool_calls": 5,
  "model": "claude-opus-5-5",
  "cost_usd": 0.188485,
  "turns": 6,
  "input_uncached": 12,
  "cache_write": 10580,
  "cache_read": 59885,
  "output": 4591,
  "is_error": false,
  "input_total": 70477,
  "agent": "claude",
  "job": "replies",
  "run": 5,
  "seconds": 44.8,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "14/14 right; wrong []"
 },
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.0549154,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 2435,
  "cache_read": 42077,
  "output": 1349,
  "is_error": false,
  "input_total": 44522,
  "agent": "claude",
  "job": "sql",
  "run": 1,
  "seconds": 17.3,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.0846014,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 5256,
  "cache_read": 40767,
  "output": 1718,
  "is_error": false,
  "input_total": 46033,
  "agent": "claude",
  "job": "sql",
  "run": 2,
  "seconds": 19.1,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 4,
  "model": "claude-opus-5-5",
  "cost_usd": 0.0880502,
  "turns": 5,
  "input_uncached": 10,
  "cache_write": 5325,
  "cache_read": 40951,
  "output": 1861,
  "is_error": false,
  "input_total": 46286,
  "agent": "claude",
  "job": "sql",
  "run": 3,
  "seconds": 20.9,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 5,
  "model": "claude-opus-5-5",
  "cost_usd": 0.11828860000000001,
  "turns": 6,
  "input_uncached": 12,
  "cache_write": 6564,
  "cache_read": 52543,
  "output": 2761,
  "is_error": false,
  "input_total": 59119,
  "agent": "claude",
  "job": "sql",
  "run": 4,
  "seconds": 29.0,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 5,
  "model": "claude-opus-5-5",
  "cost_usd": 0.1167184,
  "turns": 6,
  "input_uncached": 12,
  "cache_write": 6516,
  "cache_read": 52112,
  "output": 2706,
  "is_error": false,
  "input_total": 58640,
  "agent": "claude",
  "job": "sql",
  "run": 5,
  "seconds": 28.6,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []"
 },
 {
  "tool_calls": 3,
  "model": "claude-opus-5-5",
  "cost_usd": 0.1591682,
  "turns": 4,
  "input_uncached": 8,
  "cache_write": 13251,
  "cache_read": 42941,
  "output": 2227,
  "is_error": false,
  "input_total": 56200,
  "agent": "claude",
  "job": "triage200",
  "run": 1,
  "seconds": 27.1,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "200/200 categories; urgent missed 0, extra 0"
 },
 {
  "tool_calls": 3,
  "model": "claude-opus-5-5",
  "cost_usd": 0.1837858,
  "turns": 4,
  "input_uncached": 8,
  "cache_write": 15775,
  "cache_read": 40669,
  "output": 2471,
  "is_error": false,
  "input_total": 56452,
  "agent": "claude",
  "job": "triage200",
  "run": 2,
  "seconds": 28.5,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 0.97,
  "detail": "194/200 categories; urgent missed 0, extra 0"
 },
 {
  "tool_calls": 3,
  "model": "claude-opus-5-5",
  "cost_usd": 0.1961974,
  "turns": 4,
  "input_uncached": 8,
  "cache_write": 16418,
  "cache_read": 41507,
  "output": 2826,
  "is_error": false,
  "input_total": 57933,
  "agent": "claude",
  "job": "triage200",
  "run": 3,
  "seconds": 32.2,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "200/200 categories; urgent missed 0, extra 0"
 },
 {
  "tool_calls": 3,
  "model": "claude-opus-5-5",
  "cost_usd": 0.1818792,
  "turns": 4,
  "input_uncached": 8,
  "cache_write": 15672,
  "cache_read": 40956,
  "output": 2414,
  "is_error": false,
  "input_total": 56636,
  "agent": "claude",
  "job": "triage200",
  "run": 4,
  "seconds": 27.6,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "200/200 categories; urgent missed 0, extra 0"
 },
 {
  "tool_calls": 3,
  "model": "claude-opus-5-5",
  "cost_usd": 0.180404,
  "turns": 4,
  "input_uncached": 8,
  "cache_write": 15780,
  "cache_read": 40760,
  "output": 2299,
  "is_error": false,
  "input_total": 56548,
  "agent": "claude",
  "job": "triage200",
  "run": 5,
  "seconds": 27.0,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "200/200 categories; urgent missed 0, extra 0"
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 61437,
  "cache_read": 52480,
  "output": 705,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "analysis",
  "run": 1,
  "seconds": 23.7,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.1773
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 61383,
  "cache_read": 52480,
  "output": 753,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "analysis",
  "run": 2,
  "seconds": 25.3,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.17916
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 69549,
  "cache_read": 62336,
  "output": 710,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "analysis",
  "run": 3,
  "seconds": 26.5,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.169966
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 61440,
  "cache_read": 53632,
  "output": 783,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "analysis",
  "run": 4,
  "seconds": 24.7,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.170862
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 61310,
  "cache_read": 56704,
  "output": 736,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "analysis",
  "run": 5,
  "seconds": 22.4,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.139564
 },
 {
  "tool_calls": 5,
  "model": "gpt-6-astra",
  "input_total": 63687,
  "cache_read": 58112,
  "output": 780,
  "reasoning": 10,
  "turns": 1,
  "agent": "codex",
  "job": "bughunt",
  "run": 1,
  "seconds": 27.0,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK",
  "est_api_cost_usd": 0.152862
 },
 {
  "tool_calls": 6,
  "model": "gpt-6-astra",
  "input_total": 79151,
  "cache_read": 69504,
  "output": 795,
  "reasoning": 11,
  "turns": 1,
  "agent": "codex",
  "job": "bughunt",
  "run": 2,
  "seconds": 31.5,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK",
  "est_api_cost_usd": 0.205724
 },
 {
  "tool_calls": 5,
  "model": "gpt-6-astra",
  "input_total": 63611,
  "cache_read": 58112,
  "output": 767,
  "reasoning": 23,
  "turns": 1,
  "agent": "codex",
  "job": "bughunt",
  "run": 3,
  "seconds": 25.3,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK",
  "est_api_cost_usd": 0.151452
 },
 {
  "tool_calls": 5,
  "model": "gpt-6-astra",
  "input_total": 63517,
  "cache_read": 57984,
  "output": 688,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "bughunt",
  "run": 4,
  "seconds": 25.0,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK",
  "est_api_cost_usd": 0.147714
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 63525,
  "cache_read": 57344,
  "output": 740,
  "reasoning": 11,
  "turns": 1,
  "agent": "codex",
  "job": "bughunt",
  "run": 5,
  "seconds": 27.9,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK",
  "est_api_cost_usd": 0.156154
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 61462,
  "cache_read": 53888,
  "output": 824,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "clean",
  "run": 1,
  "seconds": 26.3,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "300 rows (want 300), 300 exact, sorted=True",
  "est_api_cost_usd": 0.170828
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 61884,
  "cache_read": 57216,
  "output": 868,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "clean",
  "run": 2,
  "seconds": 26.3,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "300 rows (want 300), 300 exact, sorted=True",
  "est_api_cost_usd": 0.147296
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 70009,
  "cache_read": 63360,
  "output": 848,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "clean",
  "run": 3,
  "seconds": 23.3,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "300 rows (want 300), 300 exact, sorted=True",
  "est_api_cost_usd": 0.17225
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 61760,
  "cache_read": 57088,
  "output": 808,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "clean",
  "run": 4,
  "seconds": 21.4,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "300 rows (want 300), 300 exact, sorted=True",
  "est_api_cost_usd": 0.144208
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 61576,
  "cache_read": 56960,
  "output": 845,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "clean",
  "run": 5,
  "seconds": 25.6,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "300 rows (want 300), 300 exact, sorted=True",
  "est_api_cost_usd": 0.14537
 },
 {
  "tool_calls": 5,
  "model": "gpt-6-astra",
  "input_total": 119484,
  "cache_read": 97408,
  "output": 652,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "corpus",
  "run": 1,
  "seconds": 24.6,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.350768
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 99946,
  "cache_read": 84480,
  "output": 605,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "corpus",
  "run": 2,
  "seconds": 23.1,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.26939
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 98111,
  "cache_read": 80128,
  "output": 559,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "corpus",
  "run": 3,
  "seconds": 25.6,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.287908
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 81375,
  "cache_read": 67072,
  "output": 549,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "corpus",
  "run": 4,
  "seconds": 26.1,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.237552
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 98045,
  "cache_read": 83200,
  "output": 534,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "corpus",
  "run": 5,
  "seconds": 22.8,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.25835
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 72052,
  "cache_read": 57088,
  "output": 1447,
  "reasoning": 17,
  "turns": 1,
  "agent": "codex",
  "job": "feature",
  "run": 1,
  "seconds": 38.5,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; 8 visible tests (needs new ones)",
  "est_api_cost_usd": 0.279078
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 63858,
  "cache_read": 55424,
  "output": 1460,
  "reasoning": 13,
  "turns": 1,
  "agent": "codex",
  "job": "feature",
  "run": 2,
  "seconds": 31.7,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; 9 visible tests (needs new ones)",
  "est_api_cost_usd": 0.212764
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 63611,
  "cache_read": 58496,
  "output": 1481,
  "reasoning": 22,
  "turns": 1,
  "agent": "codex",
  "job": "feature",
  "run": 3,
  "seconds": 36.5,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; 12 visible tests (needs new ones)",
  "est_api_cost_usd": 0.183696
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 63602,
  "cache_read": 55040,
  "output": 1468,
  "reasoning": 21,
  "turns": 1,
  "agent": "codex",
  "job": "feature",
  "run": 4,
  "seconds": 37.4,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; 14 visible tests (needs new ones)",
  "est_api_cost_usd": 0.21406
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 63727,
  "cache_read": 58752,
  "output": 1531,
  "reasoning": 17,
  "turns": 1,
  "agent": "codex",
  "job": "feature",
  "run": 5,
  "seconds": 46.8,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; 9 visible tests (needs new ones)",
  "est_api_cost_usd": 0.185052
 },
 {
  "tool_calls": 7,
  "model": "gpt-6-astra",
  "input_total": 150512,
  "cache_read": 134528,
  "output": 1002,
  "reasoning": 41,
  "turns": 1,
  "agent": "codex",
  "job": "longdoc",
  "run": 1,
  "seconds": 39.0,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "12/12 right; wrong []",
  "est_api_cost_usd": 0.344468
 },
 {
  "tool_calls": 6,
  "model": "gpt-6-astra",
  "input_total": 114417,
  "cache_read": 98688,
  "output": 976,
  "reasoning": 38,
  "turns": 1,
  "agent": "codex",
  "job": "longdoc",
  "run": 2,
  "seconds": 36.2,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "12/12 right; wrong []",
  "est_api_cost_usd": 0.304778
 },
 {
  "tool_calls": 6,
  "model": "gpt-6-astra",
  "input_total": 99121,
  "cache_read": 87040,
  "output": 946,
  "reasoning": 36,
  "turns": 1,
  "agent": "codex",
  "job": "longdoc",
  "run": 3,
  "seconds": 27.9,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "12/12 right; wrong []",
  "est_api_cost_usd": 0.25515
 },
 {
  "tool_calls": 8,
  "model": "gpt-6-astra",
  "input_total": 158598,
  "cache_read": 135296,
  "output": 1031,
  "reasoning": 43,
  "turns": 1,
  "agent": "codex",
  "job": "longdoc",
  "run": 4,
  "seconds": 42.8,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "12/12 right; wrong []",
  "est_api_cost_usd": 0.419866
 },
 {
  "tool_calls": 6,
  "model": "gpt-6-astra",
  "input_total": 165630,
  "cache_read": 138368,
  "output": 512,
  "reasoning": 26,
  "turns": 1,
  "agent": "codex",
  "job": "longdoc",
  "run": 5,
  "seconds": 26.0,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "12/12 right; wrong []",
  "est_api_cost_usd": 0.436588
 },
 {
  "tool_calls": 5,
  "model": "gpt-6-astra",
  "input_total": 81506,
  "cache_read": 70144,
  "output": 1357,
  "reasoning": 15,
  "turns": 1,
  "agent": "codex",
  "job": "refactor",
  "run": 1,
  "seconds": 33.2,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; send_email defined in ['email.py']",
  "est_api_cost_usd": 0.251614
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 63780,
  "cache_read": 49536,
  "output": 1186,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "refactor",
  "run": 2,
  "seconds": 32.2,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; send_email defined in ['email.py']",
  "est_api_cost_usd": 0.251276
 },
 {
  "tool_calls": 5,
  "model": "gpt-6-astra",
  "input_total": 64204,
  "cache_read": 58624,
  "output": 1397,
  "reasoning": 30,
  "turns": 1,
  "agent": "codex",
  "job": "refactor",
  "run": 3,
  "seconds": 28.3,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; send_email defined in ['email.py']",
  "est_api_cost_usd": 0.184274
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 63608,
  "cache_read": 58240,
  "output": 1055,
  "reasoning": 8,
  "turns": 1,
  "agent": "codex",
  "job": "refactor",
  "run": 4,
  "seconds": 35.2,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; send_email defined in ['email.py']",
  "est_api_cost_usd": 0.16467
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 72024,
  "cache_read": 57088,
  "output": 1035,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "refactor",
  "run": 5,
  "seconds": 30.6,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "visible OK; hidden OK; send_email defined in ['email.py']",
  "est_api_cost_usd": 0.258198
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 80241,
  "cache_read": 73856,
  "output": 1039,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "replies",
  "run": 1,
  "seconds": 30.6,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "14/14 right; wrong []",
  "est_api_cost_usd": 0.189656
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 80237,
  "cache_read": 67328,
  "output": 1124,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "replies",
  "run": 2,
  "seconds": 37.0,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "14/14 right; wrong []",
  "est_api_cost_usd": 0.252618
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 65131,
  "cache_read": 59008,
  "output": 1074,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "replies",
  "run": 3,
  "seconds": 39.4,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "14/14 right; wrong []",
  "est_api_cost_usd": 0.173938
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 80144,
  "cache_read": 70272,
  "output": 944,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "replies",
  "run": 4,
  "seconds": 32.1,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "14/14 right; wrong []",
  "est_api_cost_usd": 0.216192
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 80220,
  "cache_read": 67456,
  "output": 1028,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "replies",
  "run": 5,
  "seconds": 36.1,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "14/14 right; wrong []",
  "est_api_cost_usd": 0.246496
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 61059,
  "cache_read": 56704,
  "output": 673,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "sql",
  "run": 1,
  "seconds": 17.3,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.133904
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 61093,
  "cache_read": 56704,
  "output": 673,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "sql",
  "run": 2,
  "seconds": 22.6,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.134244
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 61111,
  "cache_read": 56704,
  "output": 686,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "sql",
  "run": 3,
  "seconds": 17.7,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.135074
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 61092,
  "cache_read": 56704,
  "output": 672,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "sql",
  "run": 4,
  "seconds": 20.7,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.134184
 },
 {
  "tool_calls": 3,
  "model": "gpt-6-astra",
  "input_total": 61222,
  "cache_read": 52736,
  "output": 712,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "sql",
  "run": 5,
  "seconds": 21.7,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "5/5 right; wrong []",
  "est_api_cost_usd": 0.173196
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 124489,
  "cache_read": 105088,
  "output": 939,
  "reasoning": 12,
  "turns": 1,
  "agent": "codex",
  "job": "triage200",
  "run": 1,
  "seconds": 31.8,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "200/200 categories; urgent missed 0, extra 0",
  "est_api_cost_usd": 0.346048
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 113700,
  "cache_read": 96640,
  "output": 836,
  "reasoning": 0,
  "turns": 1,
  "agent": "codex",
  "job": "triage200",
  "run": 2,
  "seconds": 30.8,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "200/200 categories; urgent missed 0, extra 0",
  "est_api_cost_usd": 0.30904
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 112576,
  "cache_read": 95744,
  "output": 790,
  "reasoning": 12,
  "turns": 1,
  "agent": "codex",
  "job": "triage200",
  "run": 3,
  "seconds": 28.1,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "200/200 categories; urgent missed 0, extra 0",
  "est_api_cost_usd": 0.303564
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 112736,
  "cache_read": 96000,
  "output": 845,
  "reasoning": 12,
  "turns": 1,
  "agent": "codex",
  "job": "triage200",
  "run": 4,
  "seconds": 25.0,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "200/200 categories; urgent missed 0, extra 0",
  "est_api_cost_usd": 0.30561
 },
 {
  "tool_calls": 4,
  "model": "gpt-6-astra",
  "input_total": 112811,
  "cache_read": 92672,
  "output": 819,
  "reasoning": 10,
  "turns": 1,
  "agent": "codex",
  "job": "triage200",
  "run": 5,
  "seconds": 31.6,
  "timed_out": false,
  "attempts": 1,
  "infra_error": false,
  "pass": true,
  "score": 1.0,
  "detail": "200/200 categories; urgent missed 0, extra 0",
  "est_api_cost_usd": 0.335012
 }
]