@zhivex-ai/harness 1.0.0 → 1.1.0-rc.2

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (181) hide show
  1. package/CHANGELOG.md +97 -1
  2. package/README.md +52 -306
  3. package/SECURITY.md +1 -1
  4. package/SUPPORT.md +7 -3
  5. package/benchmarks/README.md +3 -3
  6. package/contracts/client-protocol-v1.schema.json +341 -0
  7. package/contracts/public-api.json +100 -1
  8. package/contracts/stable-api-signatures.json +14 -14
  9. package/dist/approval-diff.d.ts +45 -0
  10. package/dist/approval-history.d.ts +38 -0
  11. package/dist/approval-preview.d.ts +26 -0
  12. package/dist/cli-credentials.d.ts +38 -0
  13. package/dist/cli-help.d.ts +3 -0
  14. package/dist/cli-options.d.ts +1 -1
  15. package/dist/cli-stream.d.ts +3 -3
  16. package/dist/cli.d.ts +15 -69
  17. package/dist/cli.js +13796 -10032
  18. package/dist/client-contract.d.ts +208 -0
  19. package/dist/console-commands.d.ts +6 -0
  20. package/dist/console-context.d.ts +5 -1
  21. package/dist/console-input.d.ts +44 -4
  22. package/dist/console-model-catalog.d.ts +25 -0
  23. package/dist/console-navigation.d.ts +27 -0
  24. package/dist/console-presentation.d.ts +18 -0
  25. package/dist/console-welcome.d.ts +13 -0
  26. package/dist/execution-environment.d.ts +4 -0
  27. package/dist/harness.d.ts +3 -0
  28. package/dist/hostile-repository-demo.js +1913 -798
  29. package/dist/index.d.ts +7 -0
  30. package/dist/index.js +7908 -6330
  31. package/dist/local-service.d.ts +34 -0
  32. package/dist/operations.d.ts +23 -1
  33. package/dist/orchestration.d.ts +3 -0
  34. package/dist/persistence-secrets.d.ts +2 -0
  35. package/dist/recorded-workspace.d.ts +4 -0
  36. package/dist/run-document.d.ts +91 -0
  37. package/dist/runtime-policy.d.ts +4 -4
  38. package/dist/service-cli.d.ts +2 -0
  39. package/dist/service-cli.js +10858 -0
  40. package/dist/service-client-cli.d.ts +2 -0
  41. package/dist/service-events.d.ts +50 -0
  42. package/dist/sessions.d.ts +3 -0
  43. package/dist/sqlite-access.d.ts +26 -0
  44. package/dist/sqlite-database.d.ts +2 -0
  45. package/dist/state-backup.d.ts +8 -0
  46. package/dist/task-memory.d.ts +2 -2
  47. package/dist/terminal-continuation.d.ts +3 -0
  48. package/dist/terminal-markdown.d.ts +1 -0
  49. package/dist/terminal-ui.d.ts +5 -0
  50. package/dist/usage-ledger.d.ts +84 -0
  51. package/dist/verify-historical-migrations.js +5699 -5191
  52. package/dist/workspace.d.ts +22 -0
  53. package/dist/zhx.js +13796 -10032
  54. package/docs/CLI.md +136 -30
  55. package/docs/CLIENT_PROTOCOL.md +212 -0
  56. package/docs/CREDENTIALS.md +59 -0
  57. package/docs/DURABLE_OPERATIONS.md +2 -2
  58. package/docs/FIRST_USE.md +82 -0
  59. package/docs/FIRST_USE_EXAMPLE.md +40 -0
  60. package/docs/HOSTILE_REPOSITORY_DEMO.md +1 -1
  61. package/docs/LOCAL_SERVICE.md +142 -0
  62. package/docs/README.md +24 -23
  63. package/docs/THREAT_MODEL.md +1 -1
  64. package/docs/TIME_TO_SAFE_FIX.md +2 -2
  65. package/docs/USAGE.md +181 -0
  66. package/evaluations/daily-cli-cohort.json +33 -0
  67. package/evaluations/representative-assembly-matrix.json +19 -1
  68. package/examples/README.md +26 -0
  69. package/examples/first-use/greeting.test.ts +3 -0
  70. package/examples/first-use/greeting.ts +1 -0
  71. package/examples/first-use/package.json +7 -0
  72. package/package.json +34 -9
  73. package/ROADMAP.md +0 -414
  74. package/benchmarks/baselines/context-navigation-2026-09-09.json +0 -94
  75. package/benchmarks/baselines/efficiency-2026-09-08.json +0 -233
  76. package/benchmarks/baselines/external-final-2026-09-08.json +0 -588
  77. package/benchmarks/baselines/external-intermediate-2026-09-08.json +0 -116
  78. package/benchmarks/baselines/external-matched-openai-2026-09-08.json +0 -1922
  79. package/benchmarks/baselines/external-matched-qwen-2026-09-08.json +0 -1706
  80. package/benchmarks/baselines/external-pilot-2026-09-08.json +0 -439
  81. package/benchmarks/baselines/external-qwen-flash-2026-09-08.json +0 -1570
  82. package/benchmarks/baselines/external-qwen-remediation-2026-09-09.json +0 -1509
  83. package/benchmarks/baselines/external-qwen-sdk-114-progress-2026-09-09.json +0 -2509
  84. package/benchmarks/baselines/external-sdk-upgrade-2026-09-08.json +0 -564
  85. package/benchmarks/baselines/external-verifier-recovery-2026-09-09.json +0 -794
  86. package/benchmarks/baselines/full-harness-review-2026-09-10.json +0 -134
  87. package/benchmarks/baselines/harness-audit-2026-09-09.json +0 -136
  88. package/benchmarks/baselines/harness-remediation-2026-09-09.json +0 -160
  89. package/benchmarks/baselines/oci-remediation-2026-09-10.json +0 -535
  90. package/benchmarks/baselines/rc14-holdout-2026-09-10.json +0 -854
  91. package/benchmarks/baselines/repair-architecture-audit-2026-09-09.json +0 -75
  92. package/benchmarks/baselines/time-to-safe-fix-live-expanded-2026-08-21.json +0 -186
  93. package/benchmarks/baselines/time-to-safe-fix-live-smoke-2026-08-21.json +0 -186
  94. package/docs/GA_READINESS.md +0 -29
  95. package/docs/GEMINI_1_0_DECISION.md +0 -9
  96. package/docs/LIVE_CERTIFICATION.md +0 -107
  97. package/docs/RC13_SECURITY_REVIEW.md +0 -28
  98. package/docs/RC14_DELEGATED_SECURITY_DECISION.md +0 -80
  99. package/docs/RC14_SECURITY_REVIEW.md +0 -40
  100. package/docs/RELEASE.md +0 -96
  101. package/docs/ROLLBACK.md +0 -9
  102. package/docs/SECURITY_REVIEW_EVIDENCE.md +0 -17
  103. package/docs/ga-readiness.json +0 -366
  104. package/docs/reports/EFFICIENCY_REVIEW.md +0 -156
  105. package/docs/reports/EXTERNAL_COMPARISON.md +0 -434
  106. package/docs/reports/FULL_HARNESS_REVIEW_2026-09-10.md +0 -200
  107. package/docs/reports/GA_BUDGET_REMEDIATION_2026-09-19.md +0 -38
  108. package/docs/reports/GA_REPAIR_RECOVERY_2026-09-19.md +0 -700
  109. package/docs/reports/GA_VALIDATION_2026-09-19.md +0 -111
  110. package/docs/reports/HARNESS_ARCHITECTURE_REVIEW_2026-09-09.md +0 -202
  111. package/docs/reports/HARNESS_AUDIT_2026-09-09.md +0 -306
  112. package/docs/reports/HARNESS_REMEDIATION_2026-09-09.md +0 -74
  113. package/docs/reports/HARNESS_REMEDIATION_2026-09-10.md +0 -90
  114. package/docs/reports/MODEL_COMPARISON_2026-09-09.md +0 -184
  115. package/docs/reports/QWEN_REMEDIATION_COMPARISON_2026-09-09.md +0 -142
  116. package/docs/reports/RC14_VALIDATION_2026-09-10.md +0 -45
  117. package/docs/reports/README.md +0 -26
  118. package/docs/reports/evidence/ga-budget-remediation-2026-09-19.json +0 -133
  119. package/docs/reports/evidence/ga-catalog-ablation-2026-09-19.json +0 -699
  120. package/docs/reports/evidence/ga-catalog-ablation-plan-2026-09-19.json +0 -40
  121. package/docs/reports/evidence/ga-catalog-observed-2026-09-19.json +0 -985
  122. package/docs/reports/evidence/ga-catalog-observed-plan-2026-09-19.json +0 -28
  123. package/docs/reports/evidence/ga-catalog-observed-status-2026-09-19.json +0 -15
  124. package/docs/reports/evidence/ga-complete-local-check-2026-09-19.json +0 -47
  125. package/docs/reports/evidence/ga-completion-reminder-2026-09-19.json +0 -22
  126. package/docs/reports/evidence/ga-fresh-cohort-plan-2026-09-19.json +0 -54
  127. package/docs/reports/evidence/ga-fresh-cohort-preparation-2026-09-19.json +0 -242
  128. package/docs/reports/evidence/ga-fresh-cohort-result-2026-09-19.json +0 -4318
  129. package/docs/reports/evidence/ga-local-validation-followup-2026-09-19.json +0 -66
  130. package/docs/reports/evidence/ga-node-consumers-2026-09-19.json +0 -31
  131. package/docs/reports/evidence/ga-package-content-validation-2026-09-19.json +0 -22
  132. package/docs/reports/evidence/ga-plan-completion-2026-09-19.json +0 -39
  133. package/docs/reports/evidence/ga-plan-reserve-2026-09-19.json +0 -573
  134. package/docs/reports/evidence/ga-plan-reserve-plan-2026-09-19.json +0 -24
  135. package/docs/reports/evidence/ga-plan-transition-2026-09-19.json +0 -19
  136. package/docs/reports/evidence/ga-private-verifier-plan-2026-09-19.json +0 -24
  137. package/docs/reports/evidence/ga-private-verifier-root-cause-2026-09-19.json +0 -41
  138. package/docs/reports/evidence/ga-published-sdk-2026-09-19.json +0 -35
  139. package/docs/reports/evidence/ga-python-checkout-imports-2026-09-19.json +0 -95
  140. package/docs/reports/evidence/ga-python-origin-delivery-2026-09-19.json +0 -211
  141. package/docs/reports/evidence/ga-python-origin-delivery-plan-2026-09-19.json +0 -27
  142. package/docs/reports/evidence/ga-qwen-patched-package-2026-09-19.json +0 -1092
  143. package/docs/reports/evidence/ga-qwen-patched-package-plan-2026-09-19.json +0 -18
  144. package/docs/reports/evidence/ga-qwen-published-2026-09-19.json +0 -571
  145. package/docs/reports/evidence/ga-qwen-published-plan-2026-09-19.json +0 -18
  146. package/docs/reports/evidence/ga-qwen-tool-choice-2026-09-19.json +0 -85
  147. package/docs/reports/evidence/ga-qwen-tool-choice-patched-2026-09-19.json +0 -85
  148. package/docs/reports/evidence/ga-recovery-development-2026-09-19.json +0 -1066
  149. package/docs/reports/evidence/ga-recovery-graded-development-2026-09-19.json +0 -1313
  150. package/docs/reports/evidence/ga-recovery-graded-v2-2026-09-19.json +0 -1151
  151. package/docs/reports/evidence/ga-recovery-v3-2026-09-19.json +0 -3943
  152. package/docs/reports/evidence/ga-recovery-v4-2026-09-19.json +0 -1055
  153. package/docs/reports/evidence/ga-recovery-v5-2026-09-19.json +0 -1245
  154. package/docs/reports/evidence/ga-recovery-v6-2026-09-19.json +0 -699
  155. package/docs/reports/evidence/ga-recovery-v7-2026-09-19.json +0 -854
  156. package/docs/reports/evidence/ga-recovery-v8-2026-09-19.json +0 -596
  157. package/docs/reports/evidence/ga-required-delivery-2026-09-19.json +0 -25
  158. package/docs/reports/evidence/ga-required-delivery-live-2026-09-19.json +0 -1337
  159. package/docs/reports/evidence/ga-required-delivery-live-plan-2026-09-19.json +0 -25
  160. package/docs/reports/evidence/ga-retained-candidate-grading-2026-09-19.json +0 -30
  161. package/docs/reports/evidence/ga-tool-selection-live-2026-09-19.json +0 -559
  162. package/docs/reports/evidence/ga-tool-selection-live-plan-2026-09-19.json +0 -28
  163. package/docs/reports/evidence/ga-tool-selection-recovery-2026-09-19.json +0 -27
  164. package/docs/reports/evidence/ga-validation-2026-09-19.json +0 -236
  165. package/docs/reports/evidence/ga-verifier-discovery-live-2026-09-19.json +0 -1369
  166. package/docs/reports/evidence/ga-verifier-discovery-plan-2026-09-19.json +0 -29
  167. package/docs/reports/evidence/ga-verifier-failure-identification-2026-09-19.json +0 -30
  168. package/docs/reports/evidence/ga-verifier-feedback-2026-09-19.json +0 -65
  169. package/docs/reports/evidence/ga-verifier-hints-live-2026-09-19.json +0 -1297
  170. package/docs/reports/evidence/ga-verifier-hints-plan-2026-09-19.json +0 -27
  171. package/docs/reports/evidence/ga-work-boundary-context-analysis-2026-09-19.json +0 -113
  172. package/docs/reports/evidence/ga-work-boundary-live-2026-09-19.json +0 -301
  173. package/docs/reports/evidence/ga-work-boundary-live-plan-2026-09-19.json +0 -84
  174. package/docs/reports/evidence/ga-work-budget-planning-2026-09-19.json +0 -32
  175. package/docs/reports/evidence/ga-working-context-2026-09-19.json +0 -22
  176. package/docs/reports/evidence/rc14-delegated-security-review-2026-09-20.json +0 -81
  177. package/docs/reports/evidence/rc14-ga-gate-2026-09-20.json +0 -18
  178. package/docs/reports/evidence/rc14-representative-2026-09-20.json +0 -509
  179. package/docs/reports/evidence/rc14-security-regressions-2026-09-20.json +0 -64
  180. package/docs/reports/evidence/security-current-2026-09-19.json +0 -91
  181. package/docs/reports/evidence/security-remediation-2026-09-19.json +0 -97
@@ -1,1245 +0,0 @@
1
- {
2
- "schemaVersion": 1,
3
- "kind": "development-comparison",
4
- "classification": "Known development task, one repetition per provider and agent; not holdout",
5
- "providers": {
6
- "openai": {
7
- "model": "gpt-5.6-luna",
8
- "limits": {
9
- "steps": 24,
10
- "outputPerTurn": 2048,
11
- "inputTokens": 100000,
12
- "outputTokens": 16000,
13
- "timeoutSeconds": 300,
14
- "memoryMb": 2048,
15
- "cpus": 2
16
- },
17
- "implementation": {
18
- "gitCommit": "41bd1cb4ab8481a7413f096a0817f612fe9806fe",
19
- "sources": {
20
- "bun.lock": "e370459197ce52c3961770f3c5ccaab7ff35ddbe641ef72e8bd61f87a8b8e3fe",
21
- "evaluations/external/requirements.lock": "42e90a08f10f821ebb958af79256f707a9fb5cf365277c4f9e2715d612028179",
22
- "package.json": "a4a2507e3ad876d72cad49a16e149192d33521fb4175bb40a3eb077ccfdf43a1",
23
- "scripts/benchmark-swebench.ts": "ed17063e1847017ba2e72f14732b6ff8f5b1af7cd5253b91302af75f2a9e8476",
24
- "scripts/swebench/candidate.py": "f5e3410b91b950acfbe50fc7511891816a8fb1c223082397e2c6300cf1e866cf",
25
- "scripts/swebench/candidate.ts": "bf102becfcce0c9a3f0df0a8f1ce0e5aaea7db2d252f6d9454a8370c24e53982",
26
- "scripts/swebench/common.py": "ad790227dac3bec1f592c458c993efff4f9d3de8f0df39b33e3238b58b3f0b3b",
27
- "scripts/swebench/context-metrics.ts": "7b84da6789413aa634d997c5963ee6796584e06c3eafe5f68f424fb7e27815e9",
28
- "scripts/swebench/mini_driver.py": "bd0a77221878370ec44760122a51f07fb56d969612afa9a7f171248b2958baa6",
29
- "scripts/swebench/model-budget.ts": "e46dd406ee5d58a051a15c7d7133513666292303d8c6534f861a6db0e711f47c",
30
- "scripts/swebench/preflight.py": "aed092a75dd52f504fba6d14814b69bd627ff17d0e311a1f1279f8b613b69675",
31
- "scripts/swebench/prepare.py": "bb7edd3fba5dbbe26597c97d744c60b94dbb7bcdfd7740a4a058c8583bd27e4c",
32
- "scripts/swebench/report.py": "d0b8d45f32c48e0ba2cb977f50838065a497a0ecaf3b4243ed94139393ce4d6b",
33
- "scripts/swebench/run.py": "94dccd55dfe6f70d8f8f2da70331700ec85eddd958547e0c3afe3dc416c9802c",
34
- "scripts/swebench/telemetry.ts": "743c9baf273e9240d704d098b6bd5eea9e76ef3e49ae636aaafa9bfd7a85b45b",
35
- "scripts/swebench/test_candidate.py": "e77d16379b748baa2daf184a45ea62587d938b6a7d48dcf9f6061f8af21fafd9",
36
- "scripts/swebench/test_contracts.py": "902e966d0fbf8501dbb15e011849f160331c033cd2a5480f9920d0f365a08d03",
37
- "scripts/swebench/zhivex-driver.ts": "b17e937484a28cf7fc7728595b51900efa5015dfa37cf6c1d4f9322051fe1ece",
38
- "scripts/time-to-safe-fix-efficiency.ts": "dd20fd8839280332ae4ca80d6846e98e07ff15d79e200b206fa2ba6be2724d2f",
39
- "src/bounded-reads.ts": "9807bb460894be3279c5425af19863d1c6e419a1582ba6910ec7e7fc35703ed6",
40
- "src/capabilities.ts": "56fdb60a2f6e1d33b2dc26d38b6d2709807e491ba1268d47b97876c2c0591d65",
41
- "src/change-envelope.ts": "9047ec978c9e5117965b8093ec961d855f97bbc2c3accdc79b74cfe6a0e4f01e",
42
- "src/cli-options.ts": "f48d9be132fed38f2ddbcfecbd8222c664a0df2f20edc28ce8d1b9b1a1900863",
43
- "src/cli-profiles.ts": "4fe451f1645e95f4957597a8724fa989b3c661ae1693252bfd6e74a51d20fc74",
44
- "src/cli-stream.ts": "5ab8a4b1b8cf0a3dfa7b5b1f9e4e0facf8f6e7f6c341bdf700a41dc8927c825e",
45
- "src/cli.ts": "5a4acb4695a968a58fbeac05c4bb89848a9653d6189590f0ff90e430ffa98d14",
46
- "src/compaction.ts": "9e06a0abeeef155377f40c160b21c56256bca95ec18c5ee897681a5fc73db169",
47
- "src/config.ts": "b1d3991e84add58603fb85b2a604b9b3eb56ae879d0fa30b0cb40545fa44edc1",
48
- "src/console-context.ts": "0ad7e4559504b9ed09ab500d615dea09af1075d357f02ec3b77f003d55d951db",
49
- "src/console-input.ts": "dc8ea0cdbba1364db288d5156c5ff9819362c46b37ef35a98b0475b6c445b1f8",
50
- "src/context-engineering.ts": "1995e564f6b91974e75bf66dd421ddcbea29b1f8eee59b76eb4e34c9d6d8f47e",
51
- "src/context-metrics.ts": "3eb30f9db808f918ed8d73d4f8ac19a13acb09ed5c1ee70fdae05201714b9198",
52
- "src/edit-contracts.ts": "47cc3132e27f8749d54a32bbcc33bb174cf4448f0d0171b2578cdb559cf2176b",
53
- "src/errors.ts": "e02f4eb80562a83762f908bda56c950ec6a2b78725c0c8a49f41fee0057ab903",
54
- "src/execution-environment.ts": "9ce6c49afab26527d5296f0c11f106063590f7333300d58dd49d5a1ec1b4b6e9",
55
- "src/file-security.ts": "b8c9f4988c4339c73a60ab4fa8d394b3eecc19117a0c61a8d19ab6483ca253df",
56
- "src/harness.ts": "866115b64d90c0b2ffa3e07da6a015210afae0babe46052912d0cc288e23bc4b",
57
- "src/index.ts": "4bbfcd11376d48c4f1c1ec17d6b5c9f51baa9331fa999432be178bae94530a5e",
58
- "src/json-contracts.ts": "0780145636b47829a1aa6087c574d5654eb31b7534ea7d9c38c188d648c90272",
59
- "src/mcp.ts": "7722de031921c620ec03b9e42a6596ef8415aa8e4bb5f93ea84d2427c466e554",
60
- "src/model-budget.ts": "e25eca12f998998acb2e03f808ed0d84ba69b97711fc577716f230244c62aa80",
61
- "src/operations.ts": "bcccaf4d1d0ec117c2ed35df92185564d33399e5fdc5c12225ac61f14fd0b4b1",
62
- "src/orchestration.ts": "b56c8c7d526a99b745cc17769e0a1b7b500b25920e77f34aa0dd01c9758de319",
63
- "src/package-manager.ts": "cec4666bbab2f42e0a269709631f79f9f87c2c4dbb3cafeffacfaff761069e2c",
64
- "src/process-runtime.ts": "b7286731a0b5c96dcad0b8e3a4bf1155d0539d73303cd7ebe77ffabf858f1729",
65
- "src/providers.ts": "70ccfc7f315cd7238e83d69b719ce4908f3b620c564cbb4f00f93e102e025602",
66
- "src/repair-controller.ts": "280aa4d33918ab5b38b170059603eb1368a28e910f2ec2a39afa854af8082cc8",
67
- "src/repair-progress.ts": "3550509a2a6c1e4a2e091c1012d5fb63ad8512a06f09b41e6b3a915a9a2b31a3",
68
- "src/repair-verifier.ts": "b70fb11feee8cdc7933947ccac0b78f8bb5c0eaccc3b29ad527197d77f40ffc9",
69
- "src/replacement-edits.ts": "8a0efdd3a7a9fc3e5a5475916cbd17ef882265881bb61076db58673e80b72957",
70
- "src/routing.ts": "8e7c9f5968b4bd638cbbc86b19847ca183f3418b9a298783b96a88cffa568e22",
71
- "src/run-interruption.ts": "df99129cf2dd6b938eca629cb897c4350a7d57fe6bf0ce437f5f3f30271617df",
72
- "src/runtime-checkpoints.ts": "e3b75a0b86fb5d21992b1a126b35446110ba5fcaf4317ed5005b1a44555118dc",
73
- "src/runtime-diagnostics.ts": "e1e5a3a558fa2aaeb477335dc336551f022c1234fe5e742eefc6599c8803e364",
74
- "src/runtime-policy.ts": "8356acf0102f0e73839b81ed628a5c0d9a89d9c21b0ca416efedde8f1b3414e2",
75
- "src/sessions.ts": "5167c8cf36bd4b8845ee278883394700321b65397f49c1dda0cf0d0de09dd06e",
76
- "src/sqlite-database.ts": "18765d1ca0c2f2ad87bfd82d311312dfe8a42006556f27bc21da5892ed3b044e",
77
- "src/state-backup.ts": "b6f8f1aedeb5dce31b65c23c7ffc50030870655b58a5aaaef9989bafb9da5c68",
78
- "src/state-directory.ts": "3711550991951e6b738a08deaa0e06da84d01864c03a9602600f592963fa3211",
79
- "src/task-memory.ts": "5b62c863be5a23c33608633cd7aac0464f5536d01cba930aca58b11870170731",
80
- "src/terminal-markdown.ts": "8cffb6ebaf25108cb2bee04956d9b2973635b010f5d7ce6be21b3ceffe3e8bfd",
81
- "src/terminal-ui.ts": "4da1878a76ba608a2517bf5d5c992a9a88702aba3fd800a9980129ba93e157bc",
82
- "src/time-to-safe-fix.ts": "4add1d49c0ea209cc496c66fa2b2012dc7bcab87e678a58c8591b3c2cdc8df8a",
83
- "src/tool-registry.ts": "7e90fff11940e525883c9ec0d2adbea2e6616f8bd9bc4ec77e1b7277e4217f2c",
84
- "src/version.ts": "b2c2f80a40adc08e8258b81e82434273f0bdfa787528b2221f14efe7f06aa747",
85
- "src/workspace-mutation-lock.ts": "ca201242db16173bd3a4d86c5d8bb0c2b6f3c43bfed3e4fa4ff5f06a75b19e80",
86
- "src/workspace.ts": "809f8bbfcf72fed10d860c48585089ca65bb3ef7c672ee6df3c918907a968302"
87
- }
88
- },
89
- "report": {
90
- "schemaVersion": 1,
91
- "kind": "swebench-comparison",
92
- "manifestSha256": "198ac9eb7e214ba667f121fc7155a0a3e985bdccd1688b473895edecd15a49a8",
93
- "complete": true,
94
- "candidates": {
95
- "zhivex": {
96
- "planned": 1,
97
- "recorded": 1,
98
- "resolved": 1,
99
- "candidateCaptured": 1,
100
- "candidateResolved": 0,
101
- "candidateGraded": 1,
102
- "resolutionRate": 1.0,
103
- "missing": 0,
104
- "gradingFailures": 0,
105
- "usageComplete": true,
106
- "totalCostUsd": null,
107
- "costPerResolvedUsd": null,
108
- "resolvedP50Ms": 76190.25887496537
109
- },
110
- "mini-swe-agent": {
111
- "planned": 1,
112
- "recorded": 1,
113
- "resolved": 1,
114
- "candidateCaptured": 1,
115
- "candidateResolved": 1,
116
- "candidateGraded": 1,
117
- "resolutionRate": 1.0,
118
- "missing": 0,
119
- "gradingFailures": 0,
120
- "usageComplete": true,
121
- "totalCostUsd": null,
122
- "costPerResolvedUsd": null,
123
- "resolvedP50Ms": 45844.67733296333
124
- }
125
- },
126
- "pairedTaskResolutionDifference": 0.0,
127
- "pairedTaskBootstrap95": null,
128
- "evidenceBoundary": "Local matched-model comparison. Missing and failed runs remain in the denominator. Small-sample intervals are exploratory. No safety superiority is inferred from SWE-bench correctness."
129
- },
130
- "samples": [
131
- {
132
- "candidate": "zhivex",
133
- "instanceId": "django__django-15851",
134
- "repetition": 0,
135
- "resolved": true,
136
- "gradingStatus": "completed",
137
- "usageComplete": true,
138
- "costUsd": null,
139
- "candidateCaptureFailure": null,
140
- "modelTimings": [
141
- {
142
- "durationMs": 1849.6875419999997,
143
- "firstTokenMs": 1847.9273750000002,
144
- "completed": true
145
- },
146
- {
147
- "durationMs": 2198.1834580000004,
148
- "firstTokenMs": 2197.9787080000006,
149
- "completed": true
150
- },
151
- {
152
- "durationMs": 4292.538541000001,
153
- "firstTokenMs": 4292.125666000001,
154
- "completed": true
155
- },
156
- {
157
- "durationMs": 2761.5964159999985,
158
- "firstTokenMs": 2695.658540999999,
159
- "completed": true
160
- },
161
- {
162
- "durationMs": 2379.7958750000016,
163
- "firstTokenMs": 2360.6923750000005,
164
- "completed": true
165
- }
166
- ],
167
- "stageTimings": [
168
- {
169
- "event": "edit-requested",
170
- "elapsedMs": 17290.478875
171
- },
172
- {
173
- "event": "verification-requested",
174
- "elapsedMs": 19068.900917
175
- },
176
- {
177
- "event": "governed-edit-completed",
178
- "elapsedMs": 19076.435707999997
179
- },
180
- {
181
- "event": "verified-import-completed",
182
- "elapsedMs": 35410.575333
183
- }
184
- ],
185
- "approvalTimings": [
186
- {
187
- "durationMs": 0.16533300000082818,
188
- "resolved": true
189
- },
190
- {
191
- "durationMs": 0.04420800000298186,
192
- "resolved": true
193
- }
194
- ],
195
- "toolTimings": [
196
- {
197
- "name": "list_files",
198
- "calls": 1,
199
- "errors": 0,
200
- "totalMs": 349.51416700000027,
201
- "maxMs": 349.51416700000027
202
- },
203
- {
204
- "name": "search_files",
205
- "calls": 5,
206
- "errors": 0,
207
- "totalMs": 571.2362929999981,
208
- "maxMs": 395.7137499999999
209
- },
210
- {
211
- "name": "read_files",
212
- "calls": 1,
213
- "errors": 0,
214
- "totalMs": 1.33916599999975,
215
- "maxMs": 1.33916599999975
216
- },
217
- {
218
- "name": "repair_plan",
219
- "calls": 1,
220
- "errors": 0,
221
- "totalMs": 0.09883400000035181,
222
- "maxMs": 0.09883400000035181
223
- },
224
- {
225
- "name": "apply_reviewed_replacement",
226
- "calls": 1,
227
- "errors": 0,
228
- "totalMs": 9.801250000000437,
229
- "maxMs": 9.801250000000437
230
- },
231
- {
232
- "name": "verify_and_apply_environment_patch",
233
- "calls": 1,
234
- "errors": 0,
235
- "totalMs": 16106.744416000001,
236
- "maxMs": 16106.744416000001
237
- }
238
- ],
239
- "schemaVersion": 1,
240
- "status": "completed",
241
- "failure": null,
242
- "phase": "agent",
243
- "durationMs": 40222.662625,
244
- "approvals": 2,
245
- "turns": [
246
- {
247
- "index": 1,
248
- "status": "completed",
249
- "finishReason": "tool-calls",
250
- "requestedTools": [
251
- "list_files",
252
- "search_files"
253
- ],
254
- "inputTokens": 3037,
255
- "outputTokens": 90,
256
- "cachedInputTokens": 0,
257
- "requestCharacters": 7258,
258
- "tools": [
259
- {
260
- "name": "list_files",
261
- "isError": false,
262
- "characters": 9769
263
- },
264
- {
265
- "name": "search_files",
266
- "isError": false,
267
- "characters": 2329
268
- }
269
- ]
270
- },
271
- {
272
- "index": 2,
273
- "status": "completed",
274
- "finishReason": "tool-calls",
275
- "requestedTools": [
276
- "search_files",
277
- "search_files",
278
- "search_files"
279
- ],
280
- "inputTokens": 6826,
281
- "outputTokens": 141,
282
- "cachedInputTokens": 2498,
283
- "requestCharacters": 23744,
284
- "tools": [
285
- {
286
- "name": "search_files",
287
- "isError": false,
288
- "characters": 351
289
- },
290
- {
291
- "name": "search_files",
292
- "isError": false,
293
- "characters": 342
294
- },
295
- {
296
- "name": "search_files",
297
- "isError": false,
298
- "characters": 2528
299
- }
300
- ]
301
- },
302
- {
303
- "index": 3,
304
- "status": "completed",
305
- "finishReason": "tool-calls",
306
- "requestedTools": [
307
- "read_files",
308
- "search_files"
309
- ],
310
- "inputTokens": 8153,
311
- "outputTokens": 129,
312
- "cachedInputTokens": 2498,
313
- "requestCharacters": 32312,
314
- "tools": [
315
- {
316
- "name": "read_files",
317
- "isError": false,
318
- "characters": 10466
319
- },
320
- {
321
- "name": "search_files",
322
- "isError": false,
323
- "characters": 2558
324
- }
325
- ]
326
- },
327
- {
328
- "index": 4,
329
- "status": "completed",
330
- "finishReason": "tool-calls",
331
- "requestedTools": [
332
- "repair_plan"
333
- ],
334
- "inputTokens": 7727,
335
- "outputTokens": 330,
336
- "cachedInputTokens": 2498,
337
- "requestCharacters": 27124,
338
- "tools": [
339
- {
340
- "name": "repair_plan",
341
- "isError": false,
342
- "characters": 1244
343
- }
344
- ]
345
- },
346
- {
347
- "index": 5,
348
- "status": "waiting_approval",
349
- "finishReason": "tool-calls",
350
- "requestedTools": [
351
- "apply_reviewed_replacement"
352
- ],
353
- "inputTokens": 8422,
354
- "outputTokens": 168,
355
- "cachedInputTokens": 2379,
356
- "requestCharacters": 35217,
357
- "tools": []
358
- },
359
- {
360
- "index": 6,
361
- "status": "completed",
362
- "finishReason": "tool-calls",
363
- "requestedTools": [
364
- "verify_and_apply_environment_patch"
365
- ],
366
- "inputTokens": 0,
367
- "outputTokens": 0,
368
- "cachedInputTokens": 0,
369
- "requestCharacters": 40354,
370
- "tools": [
371
- {
372
- "name": "apply_reviewed_replacement",
373
- "isError": false,
374
- "characters": 483
375
- },
376
- {
377
- "name": "verify_and_apply_environment_patch",
378
- "isError": false,
379
- "characters": 1343,
380
- "phaseLatencies": {
381
- "commandAndAttestationMs": 2442.160958,
382
- "workspaceExportMs": 6614.687625,
383
- "totalMs": 11389.767625,
384
- "sessionCreationMs": 2332.262667
385
- },
386
- "exitCode": 0,
387
- "timedOut": false,
388
- "verificationOutcome": "passed"
389
- }
390
- ]
391
- }
392
- ],
393
- "inputTokens": 34165,
394
- "outputTokens": 858,
395
- "cachedInputTokens": 9873,
396
- "modelCalls": 5,
397
- "compactions": [
398
- {
399
- "beforeStep": 4,
400
- "estimatedTokensBefore": 12160,
401
- "estimatedTokensAfter": 6422
402
- }
403
- ],
404
- "terminalDiagnostic": null,
405
- "thrownDiagnostic": null,
406
- "guardrails": [],
407
- "repairProgress": {
408
- "repeatedResults": 0,
409
- "suppressedResults": 0,
410
- "blockedBroadCalls": 0,
411
- "enteredClosure": false,
412
- "phase": "verify",
413
- "blockedPhaseCalls": 0,
414
- "planRecorded": true
415
- },
416
- "contextMetrics": [
417
- {
418
- "systemCharacters": 4851,
419
- "userCharacters": 2403,
420
- "assistantCharacters": 0,
421
- "toolResultCharacters": 0,
422
- "otherMessageCharacters": 0,
423
- "toolDefinitionCharacters": 9375,
424
- "unmeasuredToolDefinitions": 0
425
- },
426
- {
427
- "systemCharacters": 4851,
428
- "userCharacters": 2403,
429
- "assistantCharacters": 4067,
430
- "toolResultCharacters": 12416,
431
- "otherMessageCharacters": 0,
432
- "toolDefinitionCharacters": 9375,
433
- "unmeasuredToolDefinitions": 0
434
- },
435
- {
436
- "systemCharacters": 4851,
437
- "userCharacters": 2403,
438
- "assistantCharacters": 8930,
439
- "toolResultCharacters": 16117,
440
- "otherMessageCharacters": 0,
441
- "toolDefinitionCharacters": 9375,
442
- "unmeasuredToolDefinitions": 0
443
- },
444
- {
445
- "systemCharacters": 4851,
446
- "userCharacters": 1437,
447
- "assistantCharacters": 7487,
448
- "toolResultCharacters": 13342,
449
- "otherMessageCharacters": 0,
450
- "toolDefinitionCharacters": 9375,
451
- "unmeasuredToolDefinitions": 0
452
- },
453
- {
454
- "systemCharacters": 4851,
455
- "userCharacters": 1816,
456
- "assistantCharacters": 13796,
457
- "toolResultCharacters": 14745,
458
- "otherMessageCharacters": 0,
459
- "toolDefinitionCharacters": 9375,
460
- "unmeasuredToolDefinitions": 0
461
- }
462
- ],
463
- "omittedContextMeasurements": 0,
464
- "toolCalls": 10,
465
- "patchSha256": "15989da845ac1f766848c5c2e153296d7eef926a752ec658548702ac0e028bc7",
466
- "protectedFilesChanged": [],
467
- "officialResolved": true,
468
- "candidateCaptured": true,
469
- "candidatePatchSha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
470
- "candidateGradingStatus": "completed",
471
- "candidateResolved": false,
472
- "repairStages": {
473
- "candidatePresent": false,
474
- "importedPatchPresent": true,
475
- "verificationAttempted": true
476
- },
477
- "gradingDurationMs": 24984.549375018105,
478
- "totalDurationMs": 76190.25887496537
479
- },
480
- {
481
- "candidate": "mini-swe-agent",
482
- "instanceId": "django__django-15851",
483
- "repetition": 0,
484
- "resolved": true,
485
- "gradingStatus": "completed",
486
- "usageComplete": true,
487
- "costUsd": null,
488
- "schemaVersion": 1,
489
- "status": "Submitted",
490
- "failure": null,
491
- "phase": "agent",
492
- "durationMs": 20709.29804199841,
493
- "approvals": 0,
494
- "toolCalls": 9,
495
- "inputTokens": 27952,
496
- "outputTokens": 1153,
497
- "cachedInputTokens": 23129,
498
- "modelCalls": 7,
499
- "patchSha256": "873ee0e2d09108303f5ebe1fa4d0aec08ee9e4ce52b93941c70b4c8c91c21db4",
500
- "protectedFilesChanged": [],
501
- "officialResolved": true,
502
- "candidateCaptured": true,
503
- "candidatePatchSha256": "873ee0e2d09108303f5ebe1fa4d0aec08ee9e4ce52b93941c70b4c8c91c21db4",
504
- "candidateGradingStatus": "completed",
505
- "candidateResolved": true,
506
- "repairStages": {
507
- "candidatePresent": true,
508
- "importedPatchPresent": null,
509
- "verificationAttempted": null
510
- },
511
- "gradingDurationMs": 22967.088915989734,
512
- "totalDurationMs": 45844.67733296333
513
- }
514
- ]
515
- },
516
- "qwen": {
517
- "model": "qwen3.8-flash",
518
- "limits": {
519
- "steps": 24,
520
- "outputPerTurn": 2048,
521
- "inputTokens": 100000,
522
- "outputTokens": 16000,
523
- "timeoutSeconds": 300,
524
- "memoryMb": 2048,
525
- "cpus": 2
526
- },
527
- "implementation": {
528
- "gitCommit": "41bd1cb4ab8481a7413f096a0817f612fe9806fe",
529
- "sources": {
530
- "bun.lock": "e370459197ce52c3961770f3c5ccaab7ff35ddbe641ef72e8bd61f87a8b8e3fe",
531
- "evaluations/external/requirements.lock": "42e90a08f10f821ebb958af79256f707a9fb5cf365277c4f9e2715d612028179",
532
- "package.json": "a4a2507e3ad876d72cad49a16e149192d33521fb4175bb40a3eb077ccfdf43a1",
533
- "scripts/benchmark-swebench.ts": "ed17063e1847017ba2e72f14732b6ff8f5b1af7cd5253b91302af75f2a9e8476",
534
- "scripts/swebench/candidate.py": "f5e3410b91b950acfbe50fc7511891816a8fb1c223082397e2c6300cf1e866cf",
535
- "scripts/swebench/candidate.ts": "bf102becfcce0c9a3f0df0a8f1ce0e5aaea7db2d252f6d9454a8370c24e53982",
536
- "scripts/swebench/common.py": "ad790227dac3bec1f592c458c993efff4f9d3de8f0df39b33e3238b58b3f0b3b",
537
- "scripts/swebench/context-metrics.ts": "7b84da6789413aa634d997c5963ee6796584e06c3eafe5f68f424fb7e27815e9",
538
- "scripts/swebench/mini_driver.py": "bd0a77221878370ec44760122a51f07fb56d969612afa9a7f171248b2958baa6",
539
- "scripts/swebench/model-budget.ts": "e46dd406ee5d58a051a15c7d7133513666292303d8c6534f861a6db0e711f47c",
540
- "scripts/swebench/preflight.py": "aed092a75dd52f504fba6d14814b69bd627ff17d0e311a1f1279f8b613b69675",
541
- "scripts/swebench/prepare.py": "bb7edd3fba5dbbe26597c97d744c60b94dbb7bcdfd7740a4a058c8583bd27e4c",
542
- "scripts/swebench/report.py": "d0b8d45f32c48e0ba2cb977f50838065a497a0ecaf3b4243ed94139393ce4d6b",
543
- "scripts/swebench/run.py": "94dccd55dfe6f70d8f8f2da70331700ec85eddd958547e0c3afe3dc416c9802c",
544
- "scripts/swebench/telemetry.ts": "743c9baf273e9240d704d098b6bd5eea9e76ef3e49ae636aaafa9bfd7a85b45b",
545
- "scripts/swebench/test_candidate.py": "e77d16379b748baa2daf184a45ea62587d938b6a7d48dcf9f6061f8af21fafd9",
546
- "scripts/swebench/test_contracts.py": "902e966d0fbf8501dbb15e011849f160331c033cd2a5480f9920d0f365a08d03",
547
- "scripts/swebench/zhivex-driver.ts": "b17e937484a28cf7fc7728595b51900efa5015dfa37cf6c1d4f9322051fe1ece",
548
- "scripts/time-to-safe-fix-efficiency.ts": "dd20fd8839280332ae4ca80d6846e98e07ff15d79e200b206fa2ba6be2724d2f",
549
- "src/bounded-reads.ts": "9807bb460894be3279c5425af19863d1c6e419a1582ba6910ec7e7fc35703ed6",
550
- "src/capabilities.ts": "56fdb60a2f6e1d33b2dc26d38b6d2709807e491ba1268d47b97876c2c0591d65",
551
- "src/change-envelope.ts": "9047ec978c9e5117965b8093ec961d855f97bbc2c3accdc79b74cfe6a0e4f01e",
552
- "src/cli-options.ts": "f48d9be132fed38f2ddbcfecbd8222c664a0df2f20edc28ce8d1b9b1a1900863",
553
- "src/cli-profiles.ts": "4fe451f1645e95f4957597a8724fa989b3c661ae1693252bfd6e74a51d20fc74",
554
- "src/cli-stream.ts": "5ab8a4b1b8cf0a3dfa7b5b1f9e4e0facf8f6e7f6c341bdf700a41dc8927c825e",
555
- "src/cli.ts": "5a4acb4695a968a58fbeac05c4bb89848a9653d6189590f0ff90e430ffa98d14",
556
- "src/compaction.ts": "9e06a0abeeef155377f40c160b21c56256bca95ec18c5ee897681a5fc73db169",
557
- "src/config.ts": "b1d3991e84add58603fb85b2a604b9b3eb56ae879d0fa30b0cb40545fa44edc1",
558
- "src/console-context.ts": "0ad7e4559504b9ed09ab500d615dea09af1075d357f02ec3b77f003d55d951db",
559
- "src/console-input.ts": "dc8ea0cdbba1364db288d5156c5ff9819362c46b37ef35a98b0475b6c445b1f8",
560
- "src/context-engineering.ts": "1995e564f6b91974e75bf66dd421ddcbea29b1f8eee59b76eb4e34c9d6d8f47e",
561
- "src/context-metrics.ts": "3eb30f9db808f918ed8d73d4f8ac19a13acb09ed5c1ee70fdae05201714b9198",
562
- "src/edit-contracts.ts": "47cc3132e27f8749d54a32bbcc33bb174cf4448f0d0171b2578cdb559cf2176b",
563
- "src/errors.ts": "e02f4eb80562a83762f908bda56c950ec6a2b78725c0c8a49f41fee0057ab903",
564
- "src/execution-environment.ts": "9ce6c49afab26527d5296f0c11f106063590f7333300d58dd49d5a1ec1b4b6e9",
565
- "src/file-security.ts": "b8c9f4988c4339c73a60ab4fa8d394b3eecc19117a0c61a8d19ab6483ca253df",
566
- "src/harness.ts": "866115b64d90c0b2ffa3e07da6a015210afae0babe46052912d0cc288e23bc4b",
567
- "src/index.ts": "4bbfcd11376d48c4f1c1ec17d6b5c9f51baa9331fa999432be178bae94530a5e",
568
- "src/json-contracts.ts": "0780145636b47829a1aa6087c574d5654eb31b7534ea7d9c38c188d648c90272",
569
- "src/mcp.ts": "7722de031921c620ec03b9e42a6596ef8415aa8e4bb5f93ea84d2427c466e554",
570
- "src/model-budget.ts": "e25eca12f998998acb2e03f808ed0d84ba69b97711fc577716f230244c62aa80",
571
- "src/operations.ts": "bcccaf4d1d0ec117c2ed35df92185564d33399e5fdc5c12225ac61f14fd0b4b1",
572
- "src/orchestration.ts": "b56c8c7d526a99b745cc17769e0a1b7b500b25920e77f34aa0dd01c9758de319",
573
- "src/package-manager.ts": "cec4666bbab2f42e0a269709631f79f9f87c2c4dbb3cafeffacfaff761069e2c",
574
- "src/process-runtime.ts": "b7286731a0b5c96dcad0b8e3a4bf1155d0539d73303cd7ebe77ffabf858f1729",
575
- "src/providers.ts": "70ccfc7f315cd7238e83d69b719ce4908f3b620c564cbb4f00f93e102e025602",
576
- "src/repair-controller.ts": "280aa4d33918ab5b38b170059603eb1368a28e910f2ec2a39afa854af8082cc8",
577
- "src/repair-progress.ts": "3550509a2a6c1e4a2e091c1012d5fb63ad8512a06f09b41e6b3a915a9a2b31a3",
578
- "src/repair-verifier.ts": "b70fb11feee8cdc7933947ccac0b78f8bb5c0eaccc3b29ad527197d77f40ffc9",
579
- "src/replacement-edits.ts": "8a0efdd3a7a9fc3e5a5475916cbd17ef882265881bb61076db58673e80b72957",
580
- "src/routing.ts": "8e7c9f5968b4bd638cbbc86b19847ca183f3418b9a298783b96a88cffa568e22",
581
- "src/run-interruption.ts": "df99129cf2dd6b938eca629cb897c4350a7d57fe6bf0ce437f5f3f30271617df",
582
- "src/runtime-checkpoints.ts": "e3b75a0b86fb5d21992b1a126b35446110ba5fcaf4317ed5005b1a44555118dc",
583
- "src/runtime-diagnostics.ts": "e1e5a3a558fa2aaeb477335dc336551f022c1234fe5e742eefc6599c8803e364",
584
- "src/runtime-policy.ts": "8356acf0102f0e73839b81ed628a5c0d9a89d9c21b0ca416efedde8f1b3414e2",
585
- "src/sessions.ts": "5167c8cf36bd4b8845ee278883394700321b65397f49c1dda0cf0d0de09dd06e",
586
- "src/sqlite-database.ts": "18765d1ca0c2f2ad87bfd82d311312dfe8a42006556f27bc21da5892ed3b044e",
587
- "src/state-backup.ts": "b6f8f1aedeb5dce31b65c23c7ffc50030870655b58a5aaaef9989bafb9da5c68",
588
- "src/state-directory.ts": "3711550991951e6b738a08deaa0e06da84d01864c03a9602600f592963fa3211",
589
- "src/task-memory.ts": "5b62c863be5a23c33608633cd7aac0464f5536d01cba930aca58b11870170731",
590
- "src/terminal-markdown.ts": "8cffb6ebaf25108cb2bee04956d9b2973635b010f5d7ce6be21b3ceffe3e8bfd",
591
- "src/terminal-ui.ts": "4da1878a76ba608a2517bf5d5c992a9a88702aba3fd800a9980129ba93e157bc",
592
- "src/time-to-safe-fix.ts": "4add1d49c0ea209cc496c66fa2b2012dc7bcab87e678a58c8591b3c2cdc8df8a",
593
- "src/tool-registry.ts": "7e90fff11940e525883c9ec0d2adbea2e6616f8bd9bc4ec77e1b7277e4217f2c",
594
- "src/version.ts": "b2c2f80a40adc08e8258b81e82434273f0bdfa787528b2221f14efe7f06aa747",
595
- "src/workspace-mutation-lock.ts": "ca201242db16173bd3a4d86c5d8bb0c2b6f3c43bfed3e4fa4ff5f06a75b19e80",
596
- "src/workspace.ts": "809f8bbfcf72fed10d860c48585089ca65bb3ef7c672ee6df3c918907a968302"
597
- }
598
- },
599
- "report": {
600
- "schemaVersion": 1,
601
- "kind": "swebench-comparison",
602
- "manifestSha256": "331ea93fbfb74a631cacd7d7c126a8814595f09f2549c93907c7c9be79e4fecd",
603
- "complete": true,
604
- "candidates": {
605
- "zhivex": {
606
- "planned": 1,
607
- "recorded": 1,
608
- "resolved": 0,
609
- "candidateCaptured": 1,
610
- "candidateResolved": 0,
611
- "candidateGraded": 1,
612
- "resolutionRate": 0.0,
613
- "missing": 0,
614
- "gradingFailures": 0,
615
- "usageComplete": true,
616
- "totalCostUsd": null,
617
- "costPerResolvedUsd": null,
618
- "resolvedP50Ms": null
619
- },
620
- "mini-swe-agent": {
621
- "planned": 1,
622
- "recorded": 1,
623
- "resolved": 1,
624
- "candidateCaptured": 1,
625
- "candidateResolved": 1,
626
- "candidateGraded": 1,
627
- "resolutionRate": 1.0,
628
- "missing": 0,
629
- "gradingFailures": 0,
630
- "usageComplete": true,
631
- "totalCostUsd": null,
632
- "costPerResolvedUsd": null,
633
- "resolvedP50Ms": 50293.124082963914
634
- }
635
- },
636
- "pairedTaskResolutionDifference": -1.0,
637
- "pairedTaskBootstrap95": null,
638
- "evidenceBoundary": "Local matched-model comparison. Missing and failed runs remain in the denominator. Small-sample intervals are exploratory. No safety superiority is inferred from SWE-bench correctness."
639
- },
640
- "samples": [
641
- {
642
- "candidate": "zhivex",
643
- "instanceId": "django__django-15851",
644
- "repetition": 0,
645
- "resolved": false,
646
- "gradingStatus": "completed",
647
- "usageComplete": true,
648
- "costUsd": null,
649
- "candidateCaptureFailure": null,
650
- "modelTimings": [
651
- {
652
- "durationMs": 48835.026291,
653
- "firstTokenMs": 47222.434833,
654
- "completed": true
655
- },
656
- {
657
- "durationMs": 1737.1307080000042,
658
- "firstTokenMs": 1736.2095830000035,
659
- "completed": true
660
- },
661
- {
662
- "durationMs": 5377.519709,
663
- "firstTokenMs": 1118.0824999999968,
664
- "completed": true
665
- },
666
- {
667
- "durationMs": 1724.487957999998,
668
- "firstTokenMs": 1724.2807919999977,
669
- "completed": true
670
- },
671
- {
672
- "durationMs": 1809.7376250000016,
673
- "firstTokenMs": 1795.096292000002,
674
- "completed": true
675
- },
676
- {
677
- "durationMs": 1926.1222500000003,
678
- "firstTokenMs": 1926.0149999999994,
679
- "completed": true
680
- },
681
- {
682
- "durationMs": 1574.1296659999934,
683
- "firstTokenMs": 1574.0151659999974,
684
- "completed": true
685
- },
686
- {
687
- "durationMs": 1819.671000000002,
688
- "firstTokenMs": 1819.585500000001,
689
- "completed": true
690
- },
691
- {
692
- "durationMs": 4230.956542,
693
- "firstTokenMs": 4230.676166999998,
694
- "completed": true
695
- },
696
- {
697
- "durationMs": 2831.794167,
698
- "firstTokenMs": 2829.474125000008,
699
- "completed": true
700
- },
701
- {
702
- "durationMs": 5494.217833000002,
703
- "firstTokenMs": null,
704
- "completed": true
705
- }
706
- ],
707
- "stageTimings": [
708
- {
709
- "event": "edit-requested",
710
- "elapsedMs": 72615.767708
711
- },
712
- {
713
- "event": "verification-requested",
714
- "elapsedMs": 74446.437833
715
- },
716
- {
717
- "event": "governed-edit-completed",
718
- "elapsedMs": 74452.12225
719
- },
720
- {
721
- "event": "command-requested",
722
- "elapsedMs": 82857.248083
723
- }
724
- ],
725
- "approvalTimings": [
726
- {
727
- "durationMs": 0.12895800000114832,
728
- "resolved": true
729
- },
730
- {
731
- "durationMs": 0.11733400000957772,
732
- "resolved": true
733
- },
734
- {
735
- "durationMs": 0.02383400000690017,
736
- "resolved": true
737
- }
738
- ],
739
- "toolTimings": [
740
- {
741
- "name": "search_files",
742
- "calls": 4,
743
- "errors": 1,
744
- "totalMs": 472.0544590000063,
745
- "maxMs": 390.1018750000003
746
- },
747
- {
748
- "name": "read_files",
749
- "calls": 4,
750
- "errors": 0,
751
- "totalMs": 4.181332999985898,
752
- "maxMs": 1.452041000004101
753
- },
754
- {
755
- "name": "repair_plan",
756
- "calls": 1,
757
- "errors": 0,
758
- "totalMs": 0.08670899999560788,
759
- "maxMs": 0.08670899999560788
760
- },
761
- {
762
- "name": "apply_reviewed_replacement",
763
- "calls": 1,
764
- "errors": 0,
765
- "totalMs": 6.189165999996476,
766
- "maxMs": 6.189165999996476
767
- },
768
- {
769
- "name": "verify_and_apply_environment_patch",
770
- "calls": 1,
771
- "errors": 1,
772
- "totalMs": 4402.687459000008,
773
- "maxMs": 4402.687459000008
774
- },
775
- {
776
- "name": "run_environment_command",
777
- "calls": 1,
778
- "errors": 0,
779
- "totalMs": 11553.121500000008,
780
- "maxMs": 11553.121500000008
781
- }
782
- ],
783
- "schemaVersion": 1,
784
- "status": "failed",
785
- "failure": "AGENT_FAILED",
786
- "phase": "agent",
787
- "durationMs": 106168.325708,
788
- "approvals": 3,
789
- "turns": [
790
- {
791
- "index": 1,
792
- "status": "completed",
793
- "finishReason": "tool-calls",
794
- "requestedTools": [
795
- "search_files",
796
- "search_files"
797
- ],
798
- "inputTokens": 4580,
799
- "outputTokens": 152,
800
- "cachedInputTokens": 0,
801
- "requestCharacters": 7258,
802
- "tools": [
803
- {
804
- "name": "search_files",
805
- "isError": false,
806
- "characters": 714
807
- },
808
- {
809
- "name": "search_files",
810
- "isError": false,
811
- "characters": 158
812
- }
813
- ]
814
- },
815
- {
816
- "index": 2,
817
- "status": "completed",
818
- "finishReason": "tool-calls",
819
- "requestedTools": [
820
- "read_files"
821
- ],
822
- "inputTokens": 5123,
823
- "outputTokens": 53,
824
- "cachedInputTokens": 4096,
825
- "requestCharacters": 9004,
826
- "tools": [
827
- {
828
- "name": "read_files",
829
- "isError": false,
830
- "characters": 2673
831
- }
832
- ]
833
- },
834
- {
835
- "index": 3,
836
- "status": "completed",
837
- "finishReason": "tool-calls",
838
- "requestedTools": [
839
- "repair_plan"
840
- ],
841
- "inputTokens": 6058,
842
- "outputTokens": 281,
843
- "cachedInputTokens": 4096,
844
- "requestCharacters": 12056,
845
- "tools": [
846
- {
847
- "name": "repair_plan",
848
- "isError": false,
849
- "characters": 897
850
- }
851
- ]
852
- },
853
- {
854
- "index": 4,
855
- "status": "completed",
856
- "finishReason": "tool-calls",
857
- "requestedTools": [
858
- "search_files"
859
- ],
860
- "inputTokens": 6636,
861
- "outputTokens": 66,
862
- "cachedInputTokens": 5120,
863
- "requestCharacters": 14573,
864
- "tools": [
865
- {
866
- "name": "search_files",
867
- "isError": true,
868
- "characters": 4,
869
- "diagnostic": {
870
- "code": "EXECUTION_FAILED",
871
- "category": "execution",
872
- "failureType": "execution",
873
- "retryable": false,
874
- "chain": [
875
- {
876
- "validationCode": null,
877
- "providerReason": null,
878
- "validationIssues": [],
879
- "status": null,
880
- "errorClass": "Object",
881
- "providerHints": [],
882
- "fingerprint": "6f7194f25cfa44d78d03006c1b5da35b4d95b2e39a676c7d881dd9c039203944",
883
- "reason": "enoent"
884
- }
885
- ]
886
- }
887
- }
888
- ]
889
- },
890
- {
891
- "index": 5,
892
- "status": "completed",
893
- "finishReason": "tool-calls",
894
- "requestedTools": [
895
- "search_files"
896
- ],
897
- "inputTokens": 6816,
898
- "outputTokens": 62,
899
- "cachedInputTokens": 6144,
900
- "requestCharacters": 15180,
901
- "tools": [
902
- {
903
- "name": "search_files",
904
- "isError": false,
905
- "characters": 4511
906
- }
907
- ]
908
- },
909
- {
910
- "index": 6,
911
- "status": "completed",
912
- "finishReason": "tool-calls",
913
- "requestedTools": [
914
- "read_files"
915
- ],
916
- "inputTokens": 9044,
917
- "outputTokens": 50,
918
- "cachedInputTokens": 6144,
919
- "requestCharacters": 20069,
920
- "tools": [
921
- {
922
- "name": "read_files",
923
- "isError": false,
924
- "characters": 4790
925
- }
926
- ]
927
- },
928
- {
929
- "index": 7,
930
- "status": "completed",
931
- "finishReason": "tool-calls",
932
- "requestedTools": [
933
- "read_files"
934
- ],
935
- "inputTokens": 10546,
936
- "outputTokens": 52,
937
- "cachedInputTokens": 8320,
938
- "requestCharacters": 25232,
939
- "tools": [
940
- {
941
- "name": "read_files",
942
- "isError": false,
943
- "characters": 3399
944
- }
945
- ]
946
- },
947
- {
948
- "index": 8,
949
- "status": "completed",
950
- "finishReason": "tool-calls",
951
- "requestedTools": [
952
- "read_files"
953
- ],
954
- "inputTokens": 6849,
955
- "outputTokens": 50,
956
- "cachedInputTokens": 3328,
957
- "requestCharacters": 14496,
958
- "tools": [
959
- {
960
- "name": "read_files",
961
- "isError": false,
962
- "characters": 2480
963
- }
964
- ]
965
- },
966
- {
967
- "index": 9,
968
- "status": "waiting_approval",
969
- "finishReason": "tool-calls",
970
- "requestedTools": [
971
- "apply_reviewed_replacement"
972
- ],
973
- "inputTokens": 7733,
974
- "outputTokens": 233,
975
- "cachedInputTokens": 6144,
976
- "requestCharacters": 17355,
977
- "tools": []
978
- },
979
- {
980
- "index": 10,
981
- "status": "completed",
982
- "finishReason": "tool-calls",
983
- "requestedTools": [
984
- "verify_and_apply_environment_patch"
985
- ],
986
- "inputTokens": 0,
987
- "outputTokens": 0,
988
- "cachedInputTokens": 0,
989
- "requestCharacters": 18842,
990
- "tools": [
991
- {
992
- "name": "apply_reviewed_replacement",
993
- "isError": false,
994
- "characters": 483
995
- },
996
- {
997
- "name": "verify_and_apply_environment_patch",
998
- "isError": true,
999
- "characters": 437,
1000
- "exitCode": 1,
1001
- "timedOut": false,
1002
- "verificationOutcome": "failed",
1003
- "diagnostic": {
1004
- "code": "EXECUTION_FAILED",
1005
- "category": "execution",
1006
- "failureType": "execution",
1007
- "retryable": false,
1008
- "chain": [
1009
- {
1010
- "validationCode": null,
1011
- "providerReason": null,
1012
- "validationIssues": [],
1013
- "status": null,
1014
- "errorClass": "Object",
1015
- "providerHints": [],
1016
- "fingerprint": "d22955ecb1810cd7e960e732636eec570e04ab3b19eab276ebf23e2096a3c941",
1017
- "reason": null
1018
- }
1019
- ]
1020
- }
1021
- }
1022
- ]
1023
- },
1024
- {
1025
- "index": 11,
1026
- "status": "waiting_approval",
1027
- "finishReason": "tool-calls",
1028
- "requestedTools": [
1029
- "run_environment_command"
1030
- ],
1031
- "inputTokens": 8713,
1032
- "outputTokens": 135,
1033
- "cachedInputTokens": 7168,
1034
- "requestCharacters": 20327,
1035
- "tools": []
1036
- },
1037
- {
1038
- "index": 12,
1039
- "status": "completed",
1040
- "finishReason": "stop",
1041
- "requestedTools": [],
1042
- "inputTokens": 6978,
1043
- "outputTokens": 340,
1044
- "cachedInputTokens": 0,
1045
- "requestCharacters": 22196,
1046
- "tools": []
1047
- }
1048
- ],
1049
- "inputTokens": 79076,
1050
- "outputTokens": 1474,
1051
- "cachedInputTokens": 50560,
1052
- "modelCalls": 11,
1053
- "compactions": [
1054
- {
1055
- "beforeStep": 8,
1056
- "estimatedTokensBefore": 6810,
1057
- "estimatedTokensAfter": 3182
1058
- }
1059
- ],
1060
- "terminalDiagnostic": {
1061
- "code": "EXECUTION_FAILED",
1062
- "category": "execution",
1063
- "failureType": "execution",
1064
- "retryable": false,
1065
- "chain": [
1066
- {
1067
- "validationCode": null,
1068
- "providerReason": null,
1069
- "validationIssues": [],
1070
- "status": null,
1071
- "errorClass": "Object",
1072
- "providerHints": [],
1073
- "fingerprint": "df4ad18f75f381d4abf1c3a9a9cb1ce99246204330b132a8d70938de039a0be4",
1074
- "reason": null
1075
- }
1076
- ]
1077
- },
1078
- "thrownDiagnostic": null,
1079
- "guardrails": [],
1080
- "repairProgress": {
1081
- "repeatedResults": 0,
1082
- "suppressedResults": 0,
1083
- "blockedBroadCalls": 0,
1084
- "enteredClosure": true,
1085
- "phase": "verify",
1086
- "blockedPhaseCalls": 0,
1087
- "planRecorded": true
1088
- },
1089
- "contextMetrics": [
1090
- {
1091
- "systemCharacters": 4851,
1092
- "userCharacters": 2403,
1093
- "assistantCharacters": 0,
1094
- "toolResultCharacters": 0,
1095
- "otherMessageCharacters": 0,
1096
- "toolDefinitionCharacters": 9375,
1097
- "unmeasuredToolDefinitions": 0
1098
- },
1099
- {
1100
- "systemCharacters": 4851,
1101
- "userCharacters": 2403,
1102
- "assistantCharacters": 551,
1103
- "toolResultCharacters": 1192,
1104
- "otherMessageCharacters": 0,
1105
- "toolDefinitionCharacters": 9375,
1106
- "unmeasuredToolDefinitions": 0
1107
- },
1108
- {
1109
- "systemCharacters": 4851,
1110
- "userCharacters": 2403,
1111
- "assistantCharacters": 770,
1112
- "toolResultCharacters": 4023,
1113
- "otherMessageCharacters": 0,
1114
- "toolDefinitionCharacters": 9375,
1115
- "unmeasuredToolDefinitions": 0
1116
- },
1117
- {
1118
- "systemCharacters": 4851,
1119
- "userCharacters": 2733,
1120
- "assistantCharacters": 1899,
1121
- "toolResultCharacters": 5079,
1122
- "otherMessageCharacters": 0,
1123
- "toolDefinitionCharacters": 9375,
1124
- "unmeasuredToolDefinitions": 0
1125
- },
1126
- {
1127
- "systemCharacters": 4851,
1128
- "userCharacters": 2733,
1129
- "assistantCharacters": 2133,
1130
- "toolResultCharacters": 5450,
1131
- "otherMessageCharacters": 0,
1132
- "toolDefinitionCharacters": 9375,
1133
- "unmeasuredToolDefinitions": 0
1134
- },
1135
- {
1136
- "systemCharacters": 4851,
1137
- "userCharacters": 2733,
1138
- "assistantCharacters": 2349,
1139
- "toolResultCharacters": 10121,
1140
- "otherMessageCharacters": 0,
1141
- "toolDefinitionCharacters": 9375,
1142
- "unmeasuredToolDefinitions": 0
1143
- },
1144
- {
1145
- "systemCharacters": 4851,
1146
- "userCharacters": 2733,
1147
- "assistantCharacters": 2562,
1148
- "toolResultCharacters": 15069,
1149
- "otherMessageCharacters": 0,
1150
- "toolDefinitionCharacters": 9375,
1151
- "unmeasuredToolDefinitions": 0
1152
- },
1153
- {
1154
- "systemCharacters": 4851,
1155
- "userCharacters": 1767,
1156
- "assistantCharacters": 4315,
1157
- "toolResultCharacters": 3557,
1158
- "otherMessageCharacters": 0,
1159
- "toolDefinitionCharacters": 9375,
1160
- "unmeasuredToolDefinitions": 0
1161
- },
1162
- {
1163
- "systemCharacters": 4851,
1164
- "userCharacters": 1767,
1165
- "assistantCharacters": 4534,
1166
- "toolResultCharacters": 6195,
1167
- "otherMessageCharacters": 0,
1168
- "toolDefinitionCharacters": 9375,
1169
- "unmeasuredToolDefinitions": 0
1170
- },
1171
- {
1172
- "systemCharacters": 4851,
1173
- "userCharacters": 2237,
1174
- "assistantCharacters": 5634,
1175
- "toolResultCharacters": 7593,
1176
- "otherMessageCharacters": 0,
1177
- "toolDefinitionCharacters": 9375,
1178
- "unmeasuredToolDefinitions": 0
1179
- },
1180
- {
1181
- "systemCharacters": 4851,
1182
- "userCharacters": 2630,
1183
- "assistantCharacters": 6150,
1184
- "toolResultCharacters": 8551,
1185
- "otherMessageCharacters": 0,
1186
- "toolDefinitionCharacters": 1635,
1187
- "unmeasuredToolDefinitions": 0
1188
- }
1189
- ],
1190
- "omittedContextMeasurements": 0,
1191
- "toolCalls": 12,
1192
- "patchSha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
1193
- "protectedFilesChanged": [],
1194
- "gradingMethod": "empty-patch-no-solution",
1195
- "officialResolved": false,
1196
- "candidateCaptured": true,
1197
- "candidatePatchSha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
1198
- "candidateGradingStatus": "completed",
1199
- "candidateResolved": false,
1200
- "repairStages": {
1201
- "candidatePresent": false,
1202
- "importedPatchPresent": false,
1203
- "verificationAttempted": true
1204
- },
1205
- "gradingDurationMs": 0.20687503274530172,
1206
- "totalDurationMs": 117300.89512496488
1207
- },
1208
- {
1209
- "candidate": "mini-swe-agent",
1210
- "instanceId": "django__django-15851",
1211
- "repetition": 0,
1212
- "resolved": true,
1213
- "gradingStatus": "completed",
1214
- "usageComplete": true,
1215
- "costUsd": null,
1216
- "schemaVersion": 1,
1217
- "status": "Submitted",
1218
- "failure": null,
1219
- "phase": "agent",
1220
- "durationMs": 23792.422500031535,
1221
- "approvals": 0,
1222
- "toolCalls": 9,
1223
- "inputTokens": 23391,
1224
- "outputTokens": 710,
1225
- "cachedInputTokens": 15360,
1226
- "modelCalls": 9,
1227
- "patchSha256": "873ee0e2d09108303f5ebe1fa4d0aec08ee9e4ce52b93941c70b4c8c91c21db4",
1228
- "protectedFilesChanged": [],
1229
- "officialResolved": true,
1230
- "candidateCaptured": true,
1231
- "candidatePatchSha256": "873ee0e2d09108303f5ebe1fa4d0aec08ee9e4ce52b93941c70b4c8c91c21db4",
1232
- "candidateGradingStatus": "completed",
1233
- "candidateResolved": true,
1234
- "repairStages": {
1235
- "candidatePresent": true,
1236
- "importedPatchPresent": null,
1237
- "verificationAttempted": null
1238
- },
1239
- "gradingDurationMs": 24378.981000045314,
1240
- "totalDurationMs": 50293.124082963914
1241
- }
1242
- ]
1243
- }
1244
- }
1245
- }