@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
@@ -1,1151 +0,0 @@
1
- {
2
- "schemaVersion": 1,
3
- "kind": "development-comparison",
4
- "classification": "known development task; 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": "28723ff3350b7572d5bd68f2d40b6feef1e0e6b4",
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": "b4f75da6dd5b96ed73d94673859cf964cfa3c3e629718da21b036ee9a77d6b17",
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": "6eb48cf207cc56b18809cd762872253c95faeb6f47bc09c7fc2f1debd9c70b76",
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": "a9b555d4c6b5f369f86a3d79ec1a3aab45886282cc7d0ef211aa084ee17e3897",
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": "e450b09d4c5c04c87389ff59dfe099563935e25347d832c100fe8d9e2ccb2ebd",
93
- "complete": true,
94
- "candidates": {
95
- "zhivex": {
96
- "planned": 1,
97
- "recorded": 1,
98
- "resolved": 0,
99
- "candidateCaptured": 1,
100
- "candidateResolved": 0,
101
- "candidateGraded": 1,
102
- "resolutionRate": 0.0,
103
- "missing": 0,
104
- "gradingFailures": 0,
105
- "usageComplete": false,
106
- "totalCostUsd": null,
107
- "costPerResolvedUsd": null,
108
- "resolvedP50Ms": null
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": 55211.05325000826
124
- }
125
- },
126
- "pairedTaskResolutionDifference": -1.0,
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": false,
136
- "gradingStatus": "completed",
137
- "usageComplete": false,
138
- "costUsd": null,
139
- "candidateCaptureFailure": null,
140
- "modelTimings": [
141
- {
142
- "durationMs": 1952.0378749999995,
143
- "firstTokenMs": 1934.8579999999997,
144
- "completed": true
145
- },
146
- {
147
- "durationMs": 1481.4574579999999,
148
- "firstTokenMs": 1442.4896250000002,
149
- "completed": true
150
- },
151
- {
152
- "durationMs": 1178.4417500000009,
153
- "firstTokenMs": 1168.2751249999992,
154
- "completed": true
155
- },
156
- {
157
- "durationMs": 1458.7090000000007,
158
- "firstTokenMs": 1399.9479580000007,
159
- "completed": true
160
- },
161
- {
162
- "durationMs": 3632.605875000001,
163
- "firstTokenMs": 3628.0615420000013,
164
- "completed": true
165
- },
166
- {
167
- "durationMs": 13409.22125,
168
- "firstTokenMs": null,
169
- "completed": false
170
- }
171
- ],
172
- "stageTimings": [],
173
- "approvalTimings": [],
174
- "toolTimings": [
175
- {
176
- "name": "list_files",
177
- "calls": 1,
178
- "errors": 0,
179
- "totalMs": 600.8902909999997,
180
- "maxMs": 600.8902909999997
181
- },
182
- {
183
- "name": "read_task",
184
- "calls": 1,
185
- "errors": 0,
186
- "totalMs": 0.15608399999928224,
187
- "maxMs": 0.15608399999928224
188
- },
189
- {
190
- "name": "search_files",
191
- "calls": 2,
192
- "errors": 0,
193
- "totalMs": 691.9165410000005,
194
- "maxMs": 440.27229100000113
195
- },
196
- {
197
- "name": "read_files",
198
- "calls": 1,
199
- "errors": 0,
200
- "totalMs": 1.4332090000007156,
201
- "maxMs": 1.4332090000007156
202
- },
203
- {
204
- "name": "repair_plan",
205
- "calls": 1,
206
- "errors": 0,
207
- "totalMs": 0.12099999999918509,
208
- "maxMs": 0.12099999999918509
209
- }
210
- ],
211
- "schemaVersion": 1,
212
- "status": "failed",
213
- "failure": "EXECUTION_FAILED",
214
- "phase": "agent",
215
- "durationMs": 32880.266167,
216
- "approvals": 0,
217
- "turns": [
218
- {
219
- "index": 1,
220
- "status": "completed",
221
- "finishReason": "tool-calls",
222
- "requestedTools": [
223
- "list_files"
224
- ],
225
- "inputTokens": 3015,
226
- "outputTokens": 42,
227
- "cachedInputTokens": 3012,
228
- "requestCharacters": 1439,
229
- "tools": [
230
- {
231
- "name": "list_files",
232
- "isError": false,
233
- "characters": 9769
234
- }
235
- ]
236
- },
237
- {
238
- "index": 2,
239
- "status": "completed",
240
- "finishReason": "tool-calls",
241
- "requestedTools": [
242
- "read_task"
243
- ],
244
- "inputTokens": 5755,
245
- "outputTokens": 73,
246
- "cachedInputTokens": 2476,
247
- "requestCharacters": 20581,
248
- "tools": [
249
- {
250
- "name": "read_task",
251
- "isError": false,
252
- "characters": 1147
253
- }
254
- ]
255
- },
256
- {
257
- "index": 3,
258
- "status": "completed",
259
- "finishReason": "tool-calls",
260
- "requestedTools": [
261
- "search_files"
262
- ],
263
- "inputTokens": 6162,
264
- "outputTokens": 37,
265
- "cachedInputTokens": 2476,
266
- "requestCharacters": 25510,
267
- "tools": [
268
- {
269
- "name": "search_files",
270
- "isError": false,
271
- "characters": 1117
272
- }
273
- ]
274
- },
275
- {
276
- "index": 4,
277
- "status": "completed",
278
- "finishReason": "tool-calls",
279
- "requestedTools": [
280
- "read_files",
281
- "search_files"
282
- ],
283
- "inputTokens": 6601,
284
- "outputTokens": 118,
285
- "cachedInputTokens": 2476,
286
- "requestCharacters": 27524,
287
- "tools": [
288
- {
289
- "name": "read_files",
290
- "isError": false,
291
- "characters": 10073
292
- },
293
- {
294
- "name": "search_files",
295
- "isError": false,
296
- "characters": 1679
297
- }
298
- ]
299
- },
300
- {
301
- "index": 5,
302
- "status": "completed",
303
- "finishReason": "tool-calls",
304
- "requestedTools": [
305
- "repair_plan"
306
- ],
307
- "inputTokens": 6938,
308
- "outputTokens": 242,
309
- "cachedInputTokens": 2476,
310
- "requestCharacters": 1439,
311
- "tools": [
312
- {
313
- "name": "repair_plan",
314
- "isError": false,
315
- "characters": 969
316
- }
317
- ]
318
- }
319
- ],
320
- "inputTokens": 28471,
321
- "outputTokens": 512,
322
- "cachedInputTokens": 12916,
323
- "modelCalls": 6,
324
- "compactions": [
325
- {
326
- "beforeStep": 5,
327
- "estimatedTokensBefore": 10611,
328
- "estimatedTokensAfter": 5834
329
- }
330
- ],
331
- "terminalDiagnostic": {
332
- "code": "EXECUTION_FAILED",
333
- "category": "execution",
334
- "failureType": "parsing",
335
- "retryable": false,
336
- "chain": [
337
- {
338
- "validationCode": null,
339
- "providerReason": "incomplete_arguments",
340
- "validationIssues": [],
341
- "status": null,
342
- "errorClass": "Object",
343
- "providerHints": [],
344
- "fingerprint": "83aa171c6634719d8782be5fb7aa7816bb4f67d6611ebe92242c6ecb68472a43",
345
- "reason": "tool call"
346
- }
347
- ]
348
- },
349
- "thrownDiagnostic": {
350
- "code": "EXECUTION_FAILED",
351
- "category": "execution",
352
- "failureType": "parsing",
353
- "retryable": false,
354
- "chain": [
355
- {
356
- "validationCode": null,
357
- "providerReason": null,
358
- "validationIssues": [],
359
- "status": null,
360
- "errorClass": "HarnessExecutionError",
361
- "providerHints": [],
362
- "fingerprint": "83aa171c6634719d8782be5fb7aa7816bb4f67d6611ebe92242c6ecb68472a43",
363
- "reason": "tool call"
364
- },
365
- {
366
- "validationCode": null,
367
- "providerReason": "incomplete_arguments",
368
- "validationIssues": [],
369
- "status": null,
370
- "errorClass": "ProviderToolCallError",
371
- "providerHints": [],
372
- "fingerprint": "83aa171c6634719d8782be5fb7aa7816bb4f67d6611ebe92242c6ecb68472a43",
373
- "reason": "tool call"
374
- },
375
- {
376
- "validationCode": null,
377
- "providerReason": "incomplete_arguments",
378
- "validationIssues": [],
379
- "status": null,
380
- "errorClass": "ProviderToolCallError",
381
- "providerHints": [],
382
- "fingerprint": "83aa171c6634719d8782be5fb7aa7816bb4f67d6611ebe92242c6ecb68472a43",
383
- "reason": "tool call"
384
- }
385
- ]
386
- },
387
- "guardrails": [],
388
- "repairProgress": {
389
- "repeatedResults": 0,
390
- "suppressedResults": 0,
391
- "blockedBroadCalls": 0,
392
- "enteredClosure": false,
393
- "phase": "explore",
394
- "blockedPhaseCalls": 0,
395
- "planRecorded": true
396
- },
397
- "contextMetrics": [
398
- {
399
- "systemCharacters": 4724,
400
- "userCharacters": 2403,
401
- "assistantCharacters": 0,
402
- "toolResultCharacters": 0,
403
- "otherMessageCharacters": 0,
404
- "toolDefinitionCharacters": 9375,
405
- "unmeasuredToolDefinitions": 0
406
- },
407
- {
408
- "systemCharacters": 4724,
409
- "userCharacters": 2403,
410
- "assistantCharacters": 3521,
411
- "toolResultCharacters": 9927,
412
- "otherMessageCharacters": 0,
413
- "toolDefinitionCharacters": 9375,
414
- "unmeasuredToolDefinitions": 0
415
- },
416
- {
417
- "systemCharacters": 4724,
418
- "userCharacters": 2403,
419
- "assistantCharacters": 7144,
420
- "toolResultCharacters": 11231,
421
- "otherMessageCharacters": 0,
422
- "toolDefinitionCharacters": 9375,
423
- "unmeasuredToolDefinitions": 0
424
- },
425
- {
426
- "systemCharacters": 4724,
427
- "userCharacters": 2403,
428
- "assistantCharacters": 7879,
429
- "toolResultCharacters": 12508,
430
- "otherMessageCharacters": 0,
431
- "toolDefinitionCharacters": 9375,
432
- "unmeasuredToolDefinitions": 0
433
- },
434
- {
435
- "systemCharacters": 4724,
436
- "userCharacters": 1437,
437
- "assistantCharacters": 6534,
438
- "toolResultCharacters": 12070,
439
- "otherMessageCharacters": 0,
440
- "toolDefinitionCharacters": 9375,
441
- "unmeasuredToolDefinitions": 0
442
- },
443
- {
444
- "systemCharacters": 4724,
445
- "userCharacters": 1845,
446
- "assistantCharacters": 12159,
447
- "toolResultCharacters": 13198,
448
- "otherMessageCharacters": 0,
449
- "toolDefinitionCharacters": 9375,
450
- "unmeasuredToolDefinitions": 0
451
- }
452
- ],
453
- "omittedContextMeasurements": 0,
454
- "toolCalls": 6,
455
- "patchSha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
456
- "protectedFilesChanged": [],
457
- "gradingMethod": "empty-patch-no-solution",
458
- "officialResolved": false,
459
- "candidateCaptured": true,
460
- "candidatePatchSha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
461
- "candidateGradingStatus": "completed",
462
- "candidateResolved": false,
463
- "repairStages": {
464
- "candidatePresent": false,
465
- "importedPatchPresent": false,
466
- "verificationAttempted": false
467
- },
468
- "gradingDurationMs": 0.1590839819982648,
469
- "totalDurationMs": 45670.68600002676
470
- },
471
- {
472
- "candidate": "mini-swe-agent",
473
- "instanceId": "django__django-15851",
474
- "repetition": 0,
475
- "resolved": true,
476
- "gradingStatus": "completed",
477
- "usageComplete": true,
478
- "costUsd": null,
479
- "schemaVersion": 1,
480
- "status": "Submitted",
481
- "failure": null,
482
- "phase": "agent",
483
- "durationMs": 27957.655458012596,
484
- "approvals": 0,
485
- "toolCalls": 11,
486
- "inputTokens": 27084,
487
- "outputTokens": 1705,
488
- "cachedInputTokens": 23814,
489
- "modelCalls": 9,
490
- "patchSha256": "873ee0e2d09108303f5ebe1fa4d0aec08ee9e4ce52b93941c70b4c8c91c21db4",
491
- "protectedFilesChanged": [],
492
- "officialResolved": true,
493
- "candidateCaptured": true,
494
- "candidatePatchSha256": "873ee0e2d09108303f5ebe1fa4d0aec08ee9e4ce52b93941c70b4c8c91c21db4",
495
- "candidateGradingStatus": "completed",
496
- "candidateResolved": true,
497
- "repairStages": {
498
- "candidatePresent": true,
499
- "importedPatchPresent": null,
500
- "verificationAttempted": null
501
- },
502
- "gradingDurationMs": 24993.36854100693,
503
- "totalDurationMs": 55211.05325000826
504
- }
505
- ]
506
- },
507
- "qwen": {
508
- "model": "qwen3.8-flash",
509
- "limits": {
510
- "steps": 24,
511
- "outputPerTurn": 2048,
512
- "inputTokens": 100000,
513
- "outputTokens": 16000,
514
- "timeoutSeconds": 300,
515
- "memoryMb": 2048,
516
- "cpus": 2
517
- },
518
- "implementation": {
519
- "gitCommit": "28723ff3350b7572d5bd68f2d40b6feef1e0e6b4",
520
- "sources": {
521
- "bun.lock": "e370459197ce52c3961770f3c5ccaab7ff35ddbe641ef72e8bd61f87a8b8e3fe",
522
- "evaluations/external/requirements.lock": "42e90a08f10f821ebb958af79256f707a9fb5cf365277c4f9e2715d612028179",
523
- "package.json": "a4a2507e3ad876d72cad49a16e149192d33521fb4175bb40a3eb077ccfdf43a1",
524
- "scripts/benchmark-swebench.ts": "ed17063e1847017ba2e72f14732b6ff8f5b1af7cd5253b91302af75f2a9e8476",
525
- "scripts/swebench/candidate.py": "f5e3410b91b950acfbe50fc7511891816a8fb1c223082397e2c6300cf1e866cf",
526
- "scripts/swebench/candidate.ts": "bf102becfcce0c9a3f0df0a8f1ce0e5aaea7db2d252f6d9454a8370c24e53982",
527
- "scripts/swebench/common.py": "ad790227dac3bec1f592c458c993efff4f9d3de8f0df39b33e3238b58b3f0b3b",
528
- "scripts/swebench/context-metrics.ts": "7b84da6789413aa634d997c5963ee6796584e06c3eafe5f68f424fb7e27815e9",
529
- "scripts/swebench/mini_driver.py": "bd0a77221878370ec44760122a51f07fb56d969612afa9a7f171248b2958baa6",
530
- "scripts/swebench/model-budget.ts": "e46dd406ee5d58a051a15c7d7133513666292303d8c6534f861a6db0e711f47c",
531
- "scripts/swebench/preflight.py": "aed092a75dd52f504fba6d14814b69bd627ff17d0e311a1f1279f8b613b69675",
532
- "scripts/swebench/prepare.py": "bb7edd3fba5dbbe26597c97d744c60b94dbb7bcdfd7740a4a058c8583bd27e4c",
533
- "scripts/swebench/report.py": "d0b8d45f32c48e0ba2cb977f50838065a497a0ecaf3b4243ed94139393ce4d6b",
534
- "scripts/swebench/run.py": "94dccd55dfe6f70d8f8f2da70331700ec85eddd958547e0c3afe3dc416c9802c",
535
- "scripts/swebench/telemetry.ts": "743c9baf273e9240d704d098b6bd5eea9e76ef3e49ae636aaafa9bfd7a85b45b",
536
- "scripts/swebench/test_candidate.py": "e77d16379b748baa2daf184a45ea62587d938b6a7d48dcf9f6061f8af21fafd9",
537
- "scripts/swebench/test_contracts.py": "902e966d0fbf8501dbb15e011849f160331c033cd2a5480f9920d0f365a08d03",
538
- "scripts/swebench/zhivex-driver.ts": "b17e937484a28cf7fc7728595b51900efa5015dfa37cf6c1d4f9322051fe1ece",
539
- "scripts/time-to-safe-fix-efficiency.ts": "dd20fd8839280332ae4ca80d6846e98e07ff15d79e200b206fa2ba6be2724d2f",
540
- "src/bounded-reads.ts": "9807bb460894be3279c5425af19863d1c6e419a1582ba6910ec7e7fc35703ed6",
541
- "src/capabilities.ts": "56fdb60a2f6e1d33b2dc26d38b6d2709807e491ba1268d47b97876c2c0591d65",
542
- "src/change-envelope.ts": "9047ec978c9e5117965b8093ec961d855f97bbc2c3accdc79b74cfe6a0e4f01e",
543
- "src/cli-options.ts": "f48d9be132fed38f2ddbcfecbd8222c664a0df2f20edc28ce8d1b9b1a1900863",
544
- "src/cli-profiles.ts": "4fe451f1645e95f4957597a8724fa989b3c661ae1693252bfd6e74a51d20fc74",
545
- "src/cli-stream.ts": "5ab8a4b1b8cf0a3dfa7b5b1f9e4e0facf8f6e7f6c341bdf700a41dc8927c825e",
546
- "src/cli.ts": "5a4acb4695a968a58fbeac05c4bb89848a9653d6189590f0ff90e430ffa98d14",
547
- "src/compaction.ts": "9e06a0abeeef155377f40c160b21c56256bca95ec18c5ee897681a5fc73db169",
548
- "src/config.ts": "b1d3991e84add58603fb85b2a604b9b3eb56ae879d0fa30b0cb40545fa44edc1",
549
- "src/console-context.ts": "0ad7e4559504b9ed09ab500d615dea09af1075d357f02ec3b77f003d55d951db",
550
- "src/console-input.ts": "dc8ea0cdbba1364db288d5156c5ff9819362c46b37ef35a98b0475b6c445b1f8",
551
- "src/context-engineering.ts": "1995e564f6b91974e75bf66dd421ddcbea29b1f8eee59b76eb4e34c9d6d8f47e",
552
- "src/context-metrics.ts": "3eb30f9db808f918ed8d73d4f8ac19a13acb09ed5c1ee70fdae05201714b9198",
553
- "src/edit-contracts.ts": "47cc3132e27f8749d54a32bbcc33bb174cf4448f0d0171b2578cdb559cf2176b",
554
- "src/errors.ts": "e02f4eb80562a83762f908bda56c950ec6a2b78725c0c8a49f41fee0057ab903",
555
- "src/execution-environment.ts": "9ce6c49afab26527d5296f0c11f106063590f7333300d58dd49d5a1ec1b4b6e9",
556
- "src/file-security.ts": "b8c9f4988c4339c73a60ab4fa8d394b3eecc19117a0c61a8d19ab6483ca253df",
557
- "src/harness.ts": "b4f75da6dd5b96ed73d94673859cf964cfa3c3e629718da21b036ee9a77d6b17",
558
- "src/index.ts": "4bbfcd11376d48c4f1c1ec17d6b5c9f51baa9331fa999432be178bae94530a5e",
559
- "src/json-contracts.ts": "0780145636b47829a1aa6087c574d5654eb31b7534ea7d9c38c188d648c90272",
560
- "src/mcp.ts": "7722de031921c620ec03b9e42a6596ef8415aa8e4bb5f93ea84d2427c466e554",
561
- "src/model-budget.ts": "e25eca12f998998acb2e03f808ed0d84ba69b97711fc577716f230244c62aa80",
562
- "src/operations.ts": "bcccaf4d1d0ec117c2ed35df92185564d33399e5fdc5c12225ac61f14fd0b4b1",
563
- "src/orchestration.ts": "b56c8c7d526a99b745cc17769e0a1b7b500b25920e77f34aa0dd01c9758de319",
564
- "src/package-manager.ts": "cec4666bbab2f42e0a269709631f79f9f87c2c4dbb3cafeffacfaff761069e2c",
565
- "src/process-runtime.ts": "b7286731a0b5c96dcad0b8e3a4bf1155d0539d73303cd7ebe77ffabf858f1729",
566
- "src/providers.ts": "70ccfc7f315cd7238e83d69b719ce4908f3b620c564cbb4f00f93e102e025602",
567
- "src/repair-controller.ts": "6eb48cf207cc56b18809cd762872253c95faeb6f47bc09c7fc2f1debd9c70b76",
568
- "src/repair-progress.ts": "3550509a2a6c1e4a2e091c1012d5fb63ad8512a06f09b41e6b3a915a9a2b31a3",
569
- "src/repair-verifier.ts": "b70fb11feee8cdc7933947ccac0b78f8bb5c0eaccc3b29ad527197d77f40ffc9",
570
- "src/replacement-edits.ts": "8a0efdd3a7a9fc3e5a5475916cbd17ef882265881bb61076db58673e80b72957",
571
- "src/routing.ts": "8e7c9f5968b4bd638cbbc86b19847ca183f3418b9a298783b96a88cffa568e22",
572
- "src/run-interruption.ts": "df99129cf2dd6b938eca629cb897c4350a7d57fe6bf0ce437f5f3f30271617df",
573
- "src/runtime-checkpoints.ts": "a9b555d4c6b5f369f86a3d79ec1a3aab45886282cc7d0ef211aa084ee17e3897",
574
- "src/runtime-diagnostics.ts": "e1e5a3a558fa2aaeb477335dc336551f022c1234fe5e742eefc6599c8803e364",
575
- "src/runtime-policy.ts": "8356acf0102f0e73839b81ed628a5c0d9a89d9c21b0ca416efedde8f1b3414e2",
576
- "src/sessions.ts": "5167c8cf36bd4b8845ee278883394700321b65397f49c1dda0cf0d0de09dd06e",
577
- "src/sqlite-database.ts": "18765d1ca0c2f2ad87bfd82d311312dfe8a42006556f27bc21da5892ed3b044e",
578
- "src/state-backup.ts": "b6f8f1aedeb5dce31b65c23c7ffc50030870655b58a5aaaef9989bafb9da5c68",
579
- "src/state-directory.ts": "3711550991951e6b738a08deaa0e06da84d01864c03a9602600f592963fa3211",
580
- "src/task-memory.ts": "5b62c863be5a23c33608633cd7aac0464f5536d01cba930aca58b11870170731",
581
- "src/terminal-markdown.ts": "8cffb6ebaf25108cb2bee04956d9b2973635b010f5d7ce6be21b3ceffe3e8bfd",
582
- "src/terminal-ui.ts": "4da1878a76ba608a2517bf5d5c992a9a88702aba3fd800a9980129ba93e157bc",
583
- "src/time-to-safe-fix.ts": "4add1d49c0ea209cc496c66fa2b2012dc7bcab87e678a58c8591b3c2cdc8df8a",
584
- "src/tool-registry.ts": "7e90fff11940e525883c9ec0d2adbea2e6616f8bd9bc4ec77e1b7277e4217f2c",
585
- "src/version.ts": "b2c2f80a40adc08e8258b81e82434273f0bdfa787528b2221f14efe7f06aa747",
586
- "src/workspace-mutation-lock.ts": "ca201242db16173bd3a4d86c5d8bb0c2b6f3c43bfed3e4fa4ff5f06a75b19e80",
587
- "src/workspace.ts": "809f8bbfcf72fed10d860c48585089ca65bb3ef7c672ee6df3c918907a968302"
588
- }
589
- },
590
- "report": {
591
- "schemaVersion": 1,
592
- "kind": "swebench-comparison",
593
- "manifestSha256": "98eec035da763f41c7f0b3bd796bdfb4a5db4b7af04e59fa7e9afe8e435c8ccd",
594
- "complete": true,
595
- "candidates": {
596
- "zhivex": {
597
- "planned": 1,
598
- "recorded": 1,
599
- "resolved": 0,
600
- "candidateCaptured": 1,
601
- "candidateResolved": 0,
602
- "candidateGraded": 1,
603
- "resolutionRate": 0.0,
604
- "missing": 0,
605
- "gradingFailures": 0,
606
- "usageComplete": true,
607
- "totalCostUsd": null,
608
- "costPerResolvedUsd": null,
609
- "resolvedP50Ms": null
610
- },
611
- "mini-swe-agent": {
612
- "planned": 1,
613
- "recorded": 1,
614
- "resolved": 1,
615
- "candidateCaptured": 1,
616
- "candidateResolved": 1,
617
- "candidateGraded": 1,
618
- "resolutionRate": 1.0,
619
- "missing": 0,
620
- "gradingFailures": 0,
621
- "usageComplete": true,
622
- "totalCostUsd": null,
623
- "costPerResolvedUsd": null,
624
- "resolvedP50Ms": 48502.77654099045
625
- }
626
- },
627
- "pairedTaskResolutionDifference": -1.0,
628
- "pairedTaskBootstrap95": null,
629
- "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."
630
- },
631
- "samples": [
632
- {
633
- "candidate": "zhivex",
634
- "instanceId": "django__django-15851",
635
- "repetition": 0,
636
- "resolved": false,
637
- "gradingStatus": "completed",
638
- "usageComplete": true,
639
- "costUsd": null,
640
- "candidateCaptureFailure": null,
641
- "modelTimings": [
642
- {
643
- "durationMs": 3584.256166,
644
- "firstTokenMs": 2470.53525,
645
- "completed": true
646
- },
647
- {
648
- "durationMs": 2939.302667,
649
- "firstTokenMs": 2939.0016670000005,
650
- "completed": true
651
- },
652
- {
653
- "durationMs": 1977.4265410000007,
654
- "firstTokenMs": 1440.8752910000003,
655
- "completed": true
656
- },
657
- {
658
- "durationMs": 3907.1031669999993,
659
- "firstTokenMs": 3906.919833,
660
- "completed": true
661
- },
662
- {
663
- "durationMs": 2242.0869999999995,
664
- "firstTokenMs": 1290.0013340000005,
665
- "completed": true
666
- },
667
- {
668
- "durationMs": 2864.407124999998,
669
- "firstTokenMs": 1045.6169999999984,
670
- "completed": true
671
- },
672
- {
673
- "durationMs": 2212.9751250000045,
674
- "firstTokenMs": 1399.6929579999996,
675
- "completed": true
676
- },
677
- {
678
- "durationMs": 4297.833332999995,
679
- "firstTokenMs": null,
680
- "completed": true
681
- }
682
- ],
683
- "stageTimings": [
684
- {
685
- "event": "command-requested",
686
- "elapsedMs": 18268.558042
687
- },
688
- {
689
- "event": "edit-requested",
690
- "elapsedMs": 34937.461792
691
- },
692
- {
693
- "event": "verification-requested",
694
- "elapsedMs": 36549.677917
695
- },
696
- {
697
- "event": "governed-edit-completed",
698
- "elapsedMs": 36554.98225
699
- }
700
- ],
701
- "approvalTimings": [
702
- {
703
- "durationMs": 0.11495799999829615,
704
- "resolved": true
705
- },
706
- {
707
- "durationMs": 0.03387500000098953,
708
- "resolved": true
709
- },
710
- {
711
- "durationMs": 0.04270900000119582,
712
- "resolved": true
713
- },
714
- {
715
- "durationMs": 0.03220800000417512,
716
- "resolved": true
717
- }
718
- ],
719
- "toolTimings": [
720
- {
721
- "name": "search_files",
722
- "calls": 2,
723
- "errors": 0,
724
- "totalMs": 363.4452080000001,
725
- "maxMs": 349.44487500000014
726
- },
727
- {
728
- "name": "list_files",
729
- "calls": 1,
730
- "errors": 0,
731
- "totalMs": 11.903999999999542,
732
- "maxMs": 11.903999999999542
733
- },
734
- {
735
- "name": "read_files",
736
- "calls": 2,
737
- "errors": 0,
738
- "totalMs": 2.5855829999982234,
739
- "maxMs": 1.6948749999992287
740
- },
741
- {
742
- "name": "repair_plan",
743
- "calls": 1,
744
- "errors": 0,
745
- "totalMs": 0.08933299999989686,
746
- "maxMs": 0.08933299999989686
747
- },
748
- {
749
- "name": "run_environment_command",
750
- "calls": 1,
751
- "errors": 0,
752
- "totalMs": 11884.953834,
753
- "maxMs": 11884.953834
754
- },
755
- {
756
- "name": "apply_reviewed_replacement",
757
- "calls": 1,
758
- "errors": 0,
759
- "totalMs": 6.329749999997148,
760
- "maxMs": 6.329749999997148
761
- },
762
- {
763
- "name": "verify_and_apply_environment_patch",
764
- "calls": 1,
765
- "errors": 1,
766
- "totalMs": 4357.139000000003,
767
- "maxMs": 4357.139000000003
768
- },
769
- {
770
- "name": "run_environment_shell",
771
- "calls": 1,
772
- "errors": 0,
773
- "totalMs": 11737.942125000001,
774
- "maxMs": 11737.942125000001
775
- }
776
- ],
777
- "schemaVersion": 1,
778
- "status": "failed",
779
- "failure": "AGENT_FAILED",
780
- "phase": "agent",
781
- "durationMs": 67282.181709,
782
- "approvals": 4,
783
- "turns": [
784
- {
785
- "index": 1,
786
- "status": "completed",
787
- "finishReason": "tool-calls",
788
- "requestedTools": [
789
- "search_files",
790
- "list_files"
791
- ],
792
- "inputTokens": 4558,
793
- "outputTokens": 142,
794
- "cachedInputTokens": 0,
795
- "requestCharacters": 7131,
796
- "tools": [
797
- {
798
- "name": "search_files",
799
- "isError": false,
800
- "characters": 158
801
- },
802
- {
803
- "name": "list_files",
804
- "isError": false,
805
- "characters": 452
806
- }
807
- ]
808
- },
809
- {
810
- "index": 2,
811
- "status": "completed",
812
- "finishReason": "tool-calls",
813
- "requestedTools": [
814
- "read_files",
815
- "search_files"
816
- ],
817
- "inputTokens": 4878,
818
- "outputTokens": 115,
819
- "cachedInputTokens": 4096,
820
- "requestCharacters": 8590,
821
- "tools": [
822
- {
823
- "name": "read_files",
824
- "isError": false,
825
- "characters": 3979
826
- },
827
- {
828
- "name": "search_files",
829
- "isError": false,
830
- "characters": 806
831
- }
832
- ]
833
- },
834
- {
835
- "index": 3,
836
- "status": "completed",
837
- "finishReason": "tool-calls",
838
- "requestedTools": [
839
- "read_files"
840
- ],
841
- "inputTokens": 6632,
842
- "outputTokens": 65,
843
- "cachedInputTokens": 4096,
844
- "requestCharacters": 14113,
845
- "tools": [
846
- {
847
- "name": "read_files",
848
- "isError": false,
849
- "characters": 7151
850
- }
851
- ]
852
- },
853
- {
854
- "index": 4,
855
- "status": "completed",
856
- "finishReason": "tool-calls",
857
- "requestedTools": [
858
- "repair_plan"
859
- ],
860
- "inputTokens": 8874,
861
- "outputTokens": 255,
862
- "cachedInputTokens": 6144,
863
- "requestCharacters": 21728,
864
- "tools": [
865
- {
866
- "name": "repair_plan",
867
- "isError": false,
868
- "characters": 910
869
- }
870
- ]
871
- },
872
- {
873
- "index": 5,
874
- "status": "waiting_approval",
875
- "finishReason": "tool-calls",
876
- "requestedTools": [
877
- "run_environment_command"
878
- ],
879
- "inputTokens": 9415,
880
- "outputTokens": 111,
881
- "cachedInputTokens": 8192,
882
- "requestCharacters": 24097,
883
- "tools": []
884
- },
885
- {
886
- "index": 6,
887
- "status": "waiting_approval",
888
- "finishReason": "tool-calls",
889
- "requestedTools": [
890
- "apply_reviewed_replacement"
891
- ],
892
- "inputTokens": 9877,
893
- "outputTokens": 213,
894
- "cachedInputTokens": 8320,
895
- "requestCharacters": 25664,
896
- "tools": [
897
- {
898
- "name": "run_environment_command",
899
- "isError": false,
900
- "characters": 600,
901
- "phaseLatencies": {
902
- "commandAndAttestationMs": 2562.427416,
903
- "workspaceExportMs": 6934.03625,
904
- "totalMs": 11884.348958,
905
- "sessionCreationMs": 2387.131208
906
- },
907
- "exitCode": 0,
908
- "timedOut": false
909
- }
910
- ]
911
- },
912
- {
913
- "index": 7,
914
- "status": "completed",
915
- "finishReason": "tool-calls",
916
- "requestedTools": [
917
- "verify_and_apply_environment_patch"
918
- ],
919
- "inputTokens": 0,
920
- "outputTokens": 0,
921
- "cachedInputTokens": 0,
922
- "requestCharacters": 27114,
923
- "tools": [
924
- {
925
- "name": "apply_reviewed_replacement",
926
- "isError": false,
927
- "characters": 483
928
- },
929
- {
930
- "name": "verify_and_apply_environment_patch",
931
- "isError": true,
932
- "characters": 276,
933
- "exitCode": 1,
934
- "timedOut": false,
935
- "verificationOutcome": "failed",
936
- "diagnostic": {
937
- "code": "EXECUTION_FAILED",
938
- "category": "execution",
939
- "failureType": "execution",
940
- "retryable": false,
941
- "chain": [
942
- {
943
- "validationCode": null,
944
- "providerReason": null,
945
- "validationIssues": [],
946
- "status": null,
947
- "errorClass": "Object",
948
- "providerHints": [],
949
- "fingerprint": "d22955ecb1810cd7e960e732636eec570e04ab3b19eab276ebf23e2096a3c941",
950
- "reason": null
951
- }
952
- ]
953
- }
954
- }
955
- ]
956
- },
957
- {
958
- "index": 8,
959
- "status": "waiting_approval",
960
- "finishReason": "tool-calls",
961
- "requestedTools": [
962
- "run_environment_shell"
963
- ],
964
- "inputTokens": 5988,
965
- "outputTokens": 83,
966
- "cachedInputTokens": 4096,
967
- "requestCharacters": 12101,
968
- "tools": []
969
- },
970
- {
971
- "index": 9,
972
- "status": "completed",
973
- "finishReason": "stop",
974
- "requestedTools": [],
975
- "inputTokens": 3980,
976
- "outputTokens": 329,
977
- "cachedInputTokens": 0,
978
- "requestCharacters": 13266,
979
- "tools": []
980
- }
981
- ],
982
- "inputTokens": 54202,
983
- "outputTokens": 1313,
984
- "cachedInputTokens": 34944,
985
- "modelCalls": 8,
986
- "compactions": [
987
- {
988
- "beforeStep": 8,
989
- "estimatedTokensBefore": 6481,
990
- "estimatedTokensAfter": 2400
991
- }
992
- ],
993
- "terminalDiagnostic": {
994
- "code": "EXECUTION_FAILED",
995
- "category": "execution",
996
- "failureType": "execution",
997
- "retryable": false,
998
- "chain": [
999
- {
1000
- "validationCode": null,
1001
- "providerReason": null,
1002
- "validationIssues": [],
1003
- "status": null,
1004
- "errorClass": "Object",
1005
- "providerHints": [],
1006
- "fingerprint": "df4ad18f75f381d4abf1c3a9a9cb1ce99246204330b132a8d70938de039a0be4",
1007
- "reason": null
1008
- }
1009
- ]
1010
- },
1011
- "thrownDiagnostic": null,
1012
- "guardrails": [],
1013
- "repairProgress": {
1014
- "repeatedResults": 0,
1015
- "suppressedResults": 0,
1016
- "blockedBroadCalls": 0,
1017
- "enteredClosure": false,
1018
- "phase": "verify",
1019
- "blockedPhaseCalls": 0,
1020
- "planRecorded": true
1021
- },
1022
- "contextMetrics": [
1023
- {
1024
- "systemCharacters": 4724,
1025
- "userCharacters": 2403,
1026
- "assistantCharacters": 0,
1027
- "toolResultCharacters": 0,
1028
- "otherMessageCharacters": 0,
1029
- "toolDefinitionCharacters": 9375,
1030
- "unmeasuredToolDefinitions": 0
1031
- },
1032
- {
1033
- "systemCharacters": 4724,
1034
- "userCharacters": 2403,
1035
- "assistantCharacters": 528,
1036
- "toolResultCharacters": 928,
1037
- "otherMessageCharacters": 0,
1038
- "toolDefinitionCharacters": 9375,
1039
- "unmeasuredToolDefinitions": 0
1040
- },
1041
- {
1042
- "systemCharacters": 4724,
1043
- "userCharacters": 2403,
1044
- "assistantCharacters": 945,
1045
- "toolResultCharacters": 6031,
1046
- "otherMessageCharacters": 0,
1047
- "toolDefinitionCharacters": 9375,
1048
- "unmeasuredToolDefinitions": 0
1049
- },
1050
- {
1051
- "systemCharacters": 4724,
1052
- "userCharacters": 2403,
1053
- "assistantCharacters": 1249,
1054
- "toolResultCharacters": 13340,
1055
- "otherMessageCharacters": 0,
1056
- "toolDefinitionCharacters": 9375,
1057
- "unmeasuredToolDefinitions": 0
1058
- },
1059
- {
1060
- "systemCharacters": 4724,
1061
- "userCharacters": 2698,
1062
- "assistantCharacters": 2252,
1063
- "toolResultCharacters": 14409,
1064
- "otherMessageCharacters": 0,
1065
- "toolDefinitionCharacters": 9375,
1066
- "unmeasuredToolDefinitions": 0
1067
- },
1068
- {
1069
- "systemCharacters": 4724,
1070
- "userCharacters": 3005,
1071
- "assistantCharacters": 2739,
1072
- "toolResultCharacters": 15180,
1073
- "otherMessageCharacters": 0,
1074
- "toolDefinitionCharacters": 9375,
1075
- "unmeasuredToolDefinitions": 0
1076
- },
1077
- {
1078
- "systemCharacters": 4724,
1079
- "userCharacters": 2503,
1080
- "assistantCharacters": 4288,
1081
- "toolResultCharacters": 580,
1082
- "otherMessageCharacters": 0,
1083
- "toolDefinitionCharacters": 9375,
1084
- "unmeasuredToolDefinitions": 0
1085
- },
1086
- {
1087
- "systemCharacters": 4724,
1088
- "userCharacters": 2588,
1089
- "assistantCharacters": 4690,
1090
- "toolResultCharacters": 1256,
1091
- "otherMessageCharacters": 0,
1092
- "toolDefinitionCharacters": 1635,
1093
- "unmeasuredToolDefinitions": 0
1094
- }
1095
- ],
1096
- "omittedContextMeasurements": 0,
1097
- "toolCalls": 10,
1098
- "patchSha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
1099
- "protectedFilesChanged": [],
1100
- "gradingMethod": "empty-patch-no-solution",
1101
- "officialResolved": false,
1102
- "candidateCaptured": true,
1103
- "candidatePatchSha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
1104
- "candidateGradingStatus": "completed",
1105
- "candidateResolved": false,
1106
- "repairStages": {
1107
- "candidatePresent": false,
1108
- "importedPatchPresent": false,
1109
- "verificationAttempted": true
1110
- },
1111
- "gradingDurationMs": 0.1834590220823884,
1112
- "totalDurationMs": 78735.57133303257
1113
- },
1114
- {
1115
- "candidate": "mini-swe-agent",
1116
- "instanceId": "django__django-15851",
1117
- "repetition": 0,
1118
- "resolved": true,
1119
- "gradingStatus": "completed",
1120
- "usageComplete": true,
1121
- "costUsd": null,
1122
- "schemaVersion": 1,
1123
- "status": "Submitted",
1124
- "failure": null,
1125
- "phase": "agent",
1126
- "durationMs": 21109.537707990967,
1127
- "approvals": 0,
1128
- "toolCalls": 8,
1129
- "inputTokens": 22034,
1130
- "outputTokens": 616,
1131
- "cachedInputTokens": 17408,
1132
- "modelCalls": 8,
1133
- "patchSha256": "873ee0e2d09108303f5ebe1fa4d0aec08ee9e4ce52b93941c70b4c8c91c21db4",
1134
- "protectedFilesChanged": [],
1135
- "officialResolved": true,
1136
- "candidateCaptured": true,
1137
- "candidatePatchSha256": "873ee0e2d09108303f5ebe1fa4d0aec08ee9e4ce52b93941c70b4c8c91c21db4",
1138
- "candidateGradingStatus": "completed",
1139
- "candidateResolved": true,
1140
- "repairStages": {
1141
- "candidatePresent": true,
1142
- "importedPatchPresent": null,
1143
- "verificationAttempted": null
1144
- },
1145
- "gradingDurationMs": 25295.337165996898,
1146
- "totalDurationMs": 48502.77654099045
1147
- }
1148
- ]
1149
- }
1150
- }
1151
- }