audrey 1.0.1 → 1.0.2

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 (227) hide show
  1. package/CHANGELOG.md +30 -0
  2. package/README.md +5 -3
  3. package/benchmarks/adapter-self-test.mjs +6 -2
  4. package/benchmarks/adapters/example-allow.mjs +5 -2
  5. package/benchmarks/adapters/mem0-platform.mjs +19 -12
  6. package/benchmarks/adapters/zep-cloud.mjs +51 -27
  7. package/benchmarks/baselines.js +11 -6
  8. package/benchmarks/build-leaderboard.mjs +36 -23
  9. package/benchmarks/cases.js +24 -12
  10. package/benchmarks/create-conformance-card.mjs +12 -3
  11. package/benchmarks/create-submission-bundle.mjs +22 -8
  12. package/benchmarks/dry-run-external-adapters.mjs +24 -12
  13. package/benchmarks/guardbench.js +263 -123
  14. package/benchmarks/output/adapter-self-test/guardbench-adapter-self-test.json +4 -4
  15. package/benchmarks/output/external/guardbench-external-dry-run.json +1 -1
  16. package/benchmarks/output/external/guardbench-external-evidence.json +1 -1
  17. package/benchmarks/output/guardbench-conformance-card.json +11 -11
  18. package/benchmarks/output/guardbench-raw.json +107 -108
  19. package/benchmarks/output/guardbench-summary.json +170 -172
  20. package/benchmarks/output/leaderboard/guardbench-leaderboard.json +5 -5
  21. package/benchmarks/output/leaderboard/guardbench-leaderboard.md +2 -2
  22. package/benchmarks/output/submission-bundle/guardbench-conformance-card.json +11 -11
  23. package/benchmarks/output/submission-bundle/guardbench-raw.json +107 -108
  24. package/benchmarks/output/submission-bundle/guardbench-summary.json +170 -172
  25. package/benchmarks/output/submission-bundle/submission-manifest.json +11 -11
  26. package/benchmarks/output/submission-bundle/validation-report.json +1 -1
  27. package/benchmarks/output/summary.json +57 -57
  28. package/benchmarks/perf-snapshot.js +12 -9
  29. package/benchmarks/perf.bench.js +14 -6
  30. package/benchmarks/public-paths.mjs +11 -5
  31. package/benchmarks/reference-results.js +10 -5
  32. package/benchmarks/report.js +48 -27
  33. package/benchmarks/run-external-guardbench.mjs +47 -25
  34. package/benchmarks/run.js +112 -59
  35. package/benchmarks/validate-adapter-module.mjs +13 -10
  36. package/benchmarks/validate-adapter-registry.mjs +16 -5
  37. package/benchmarks/validate-guardbench-artifacts.mjs +76 -19
  38. package/benchmarks/verify-external-evidence.mjs +86 -31
  39. package/benchmarks/verify-publication-artifacts.mjs +34 -11
  40. package/benchmarks/verify-submission-bundle.mjs +9 -4
  41. package/dist/mcp-server/config.d.ts +1 -1
  42. package/dist/mcp-server/config.d.ts.map +1 -1
  43. package/dist/mcp-server/config.js +5 -3
  44. package/dist/mcp-server/config.js.map +1 -1
  45. package/dist/mcp-server/index.d.ts +4 -3
  46. package/dist/mcp-server/index.d.ts.map +1 -1
  47. package/dist/mcp-server/index.js +479 -172
  48. package/dist/mcp-server/index.js.map +1 -1
  49. package/dist/src/action-key.d.ts.map +1 -1
  50. package/dist/src/action-key.js +6 -2
  51. package/dist/src/action-key.js.map +1 -1
  52. package/dist/src/adaptive.d.ts.map +1 -1
  53. package/dist/src/adaptive.js +4 -2
  54. package/dist/src/adaptive.js.map +1 -1
  55. package/dist/src/affect.d.ts.map +1 -1
  56. package/dist/src/affect.js +8 -5
  57. package/dist/src/affect.js.map +1 -1
  58. package/dist/src/audrey.d.ts +1 -1
  59. package/dist/src/audrey.d.ts.map +1 -1
  60. package/dist/src/audrey.js +93 -49
  61. package/dist/src/audrey.js.map +1 -1
  62. package/dist/src/capsule.d.ts.map +1 -1
  63. package/dist/src/capsule.js +37 -15
  64. package/dist/src/capsule.js.map +1 -1
  65. package/dist/src/causal.d.ts +1 -1
  66. package/dist/src/causal.d.ts.map +1 -1
  67. package/dist/src/causal.js +4 -2
  68. package/dist/src/causal.js.map +1 -1
  69. package/dist/src/confidence.d.ts.map +1 -1
  70. package/dist/src/confidence.js +5 -5
  71. package/dist/src/confidence.js.map +1 -1
  72. package/dist/src/consolidate.d.ts.map +1 -1
  73. package/dist/src/consolidate.js +17 -9
  74. package/dist/src/consolidate.js.map +1 -1
  75. package/dist/src/context.js +1 -1
  76. package/dist/src/context.js.map +1 -1
  77. package/dist/src/controller.d.ts.map +1 -1
  78. package/dist/src/controller.js +24 -13
  79. package/dist/src/controller.js.map +1 -1
  80. package/dist/src/db.d.ts.map +1 -1
  81. package/dist/src/db.js +78 -27
  82. package/dist/src/db.js.map +1 -1
  83. package/dist/src/decay.d.ts +1 -1
  84. package/dist/src/decay.d.ts.map +1 -1
  85. package/dist/src/decay.js +1 -1
  86. package/dist/src/decay.js.map +1 -1
  87. package/dist/src/embedding.d.ts +12 -4
  88. package/dist/src/embedding.d.ts.map +1 -1
  89. package/dist/src/embedding.js +18 -16
  90. package/dist/src/embedding.js.map +1 -1
  91. package/dist/src/encode.d.ts.map +1 -1
  92. package/dist/src/encode.js +5 -4
  93. package/dist/src/encode.js.map +1 -1
  94. package/dist/src/events.d.ts +3 -2
  95. package/dist/src/events.d.ts.map +1 -1
  96. package/dist/src/events.js +7 -3
  97. package/dist/src/events.js.map +1 -1
  98. package/dist/src/export.d.ts.map +1 -1
  99. package/dist/src/export.js +21 -7
  100. package/dist/src/export.js.map +1 -1
  101. package/dist/src/feedback.d.ts.map +1 -1
  102. package/dist/src/feedback.js +1 -1
  103. package/dist/src/feedback.js.map +1 -1
  104. package/dist/src/forget.d.ts.map +1 -1
  105. package/dist/src/forget.js +12 -6
  106. package/dist/src/forget.js.map +1 -1
  107. package/dist/src/fts.d.ts.map +1 -1
  108. package/dist/src/fts.js +20 -8
  109. package/dist/src/fts.js.map +1 -1
  110. package/dist/src/hybrid-recall.d.ts.map +1 -1
  111. package/dist/src/hybrid-recall.js +12 -6
  112. package/dist/src/hybrid-recall.js.map +1 -1
  113. package/dist/src/impact.d.ts.map +1 -1
  114. package/dist/src/impact.js +26 -10
  115. package/dist/src/impact.js.map +1 -1
  116. package/dist/src/import.d.ts.map +1 -1
  117. package/dist/src/import.js +11 -6
  118. package/dist/src/import.js.map +1 -1
  119. package/dist/src/index.d.ts +3 -3
  120. package/dist/src/index.d.ts.map +1 -1
  121. package/dist/src/index.js +3 -3
  122. package/dist/src/index.js.map +1 -1
  123. package/dist/src/interference.d.ts.map +1 -1
  124. package/dist/src/interference.js +10 -5
  125. package/dist/src/interference.js.map +1 -1
  126. package/dist/src/introspect.d.ts.map +1 -1
  127. package/dist/src/introspect.js +12 -6
  128. package/dist/src/introspect.js.map +1 -1
  129. package/dist/src/llm.d.ts +2 -2
  130. package/dist/src/llm.d.ts.map +1 -1
  131. package/dist/src/llm.js +6 -6
  132. package/dist/src/llm.js.map +1 -1
  133. package/dist/src/migrate.d.ts.map +1 -1
  134. package/dist/src/migrate.js +10 -4
  135. package/dist/src/migrate.js.map +1 -1
  136. package/dist/src/preflight.d.ts.map +1 -1
  137. package/dist/src/preflight.js +6 -8
  138. package/dist/src/preflight.js.map +1 -1
  139. package/dist/src/profile.d.ts.map +1 -1
  140. package/dist/src/profile.js.map +1 -1
  141. package/dist/src/promote.d.ts.map +1 -1
  142. package/dist/src/promote.js +16 -7
  143. package/dist/src/promote.js.map +1 -1
  144. package/dist/src/prompts.d.ts.map +1 -1
  145. package/dist/src/prompts.js +1 -2
  146. package/dist/src/prompts.js.map +1 -1
  147. package/dist/src/recall.d.ts.map +1 -1
  148. package/dist/src/recall.js +85 -18
  149. package/dist/src/recall.js.map +1 -1
  150. package/dist/src/redact.d.ts.map +1 -1
  151. package/dist/src/redact.js +9 -4
  152. package/dist/src/redact.js.map +1 -1
  153. package/dist/src/reflexes.d.ts.map +1 -1
  154. package/dist/src/reflexes.js +1 -7
  155. package/dist/src/reflexes.js.map +1 -1
  156. package/dist/src/rollback.d.ts.map +1 -1
  157. package/dist/src/rollback.js +4 -2
  158. package/dist/src/rollback.js.map +1 -1
  159. package/dist/src/routes.d.ts.map +1 -1
  160. package/dist/src/routes.js +33 -13
  161. package/dist/src/routes.js.map +1 -1
  162. package/dist/src/rules-compiler.d.ts.map +1 -1
  163. package/dist/src/rules-compiler.js +24 -2
  164. package/dist/src/rules-compiler.js.map +1 -1
  165. package/dist/src/server.js +2 -2
  166. package/dist/src/server.js.map +1 -1
  167. package/dist/src/tool-trace.d.ts +2 -2
  168. package/dist/src/tool-trace.d.ts.map +1 -1
  169. package/dist/src/tool-trace.js +12 -4
  170. package/dist/src/tool-trace.js.map +1 -1
  171. package/dist/src/types.d.ts.map +1 -1
  172. package/dist/src/ulid.js +1 -1
  173. package/dist/src/ulid.js.map +1 -1
  174. package/dist/src/utils.d.ts.map +1 -1
  175. package/dist/src/utils.js.map +1 -1
  176. package/dist/src/validate.d.ts.map +1 -1
  177. package/dist/src/validate.js +20 -10
  178. package/dist/src/validate.js.map +1 -1
  179. package/docs/paper/07-evaluation.md +5 -5
  180. package/docs/paper/audrey-paper-v1.md +5 -5
  181. package/docs/paper/evidence-ledger.md +1 -1
  182. package/docs/paper/output/arxiv/arxiv-manifest.json +4 -4
  183. package/docs/paper/output/arxiv/main.tex +5 -5
  184. package/docs/paper/output/arxiv-compile-report.json +3 -3
  185. package/docs/paper/output/submission-bundle/README.md +5 -3
  186. package/docs/paper/output/submission-bundle/benchmarks/output/adapter-self-test/guardbench-adapter-self-test.json +4 -4
  187. package/docs/paper/output/submission-bundle/benchmarks/output/external/guardbench-external-dry-run.json +1 -1
  188. package/docs/paper/output/submission-bundle/benchmarks/output/external/guardbench-external-evidence.json +1 -1
  189. package/docs/paper/output/submission-bundle/benchmarks/output/guardbench-conformance-card.json +11 -11
  190. package/docs/paper/output/submission-bundle/benchmarks/output/guardbench-raw.json +107 -108
  191. package/docs/paper/output/submission-bundle/benchmarks/output/guardbench-summary.json +170 -172
  192. package/docs/paper/output/submission-bundle/benchmarks/output/leaderboard/guardbench-leaderboard.json +5 -5
  193. package/docs/paper/output/submission-bundle/benchmarks/output/leaderboard/guardbench-leaderboard.md +2 -2
  194. package/docs/paper/output/submission-bundle/benchmarks/output/submission-bundle/submission-manifest.json +11 -11
  195. package/docs/paper/output/submission-bundle/benchmarks/output/submission-bundle/validation-report.json +1 -1
  196. package/docs/paper/output/submission-bundle/benchmarks/output/summary.json +58 -58
  197. package/docs/paper/output/submission-bundle/docs/paper/07-evaluation.md +5 -5
  198. package/docs/paper/output/submission-bundle/docs/paper/audrey-paper-v1.md +5 -5
  199. package/docs/paper/output/submission-bundle/docs/paper/evidence-ledger.md +1 -1
  200. package/docs/paper/output/submission-bundle/docs/paper/output/arxiv/arxiv-manifest.json +4 -4
  201. package/docs/paper/output/submission-bundle/docs/paper/output/arxiv/main.tex +5 -5
  202. package/docs/paper/output/submission-bundle/docs/paper/output/arxiv-compile-report.json +3 -3
  203. package/docs/paper/output/submission-bundle/package.json +17 -4
  204. package/docs/paper/output/submission-bundle/paper-submission-manifest.json +36 -36
  205. package/examples/fintech-ops-demo.js +12 -5
  206. package/examples/healthcare-ops-demo.js +8 -4
  207. package/examples/ollama-memory-agent.js +41 -13
  208. package/examples/stripe-demo.js +12 -5
  209. package/package.json +17 -4
  210. package/scripts/audit-release-completion.mjs +179 -101
  211. package/scripts/create-arxiv-source.mjs +20 -14
  212. package/scripts/create-paper-submission-bundle.mjs +6 -2
  213. package/scripts/finalize-release.mjs +111 -36
  214. package/scripts/prepare-release-cut.mjs +14 -6
  215. package/scripts/publish-release-bundle.mjs +62 -23
  216. package/scripts/publish-release-github-api.mjs +89 -24
  217. package/scripts/smoke-cli.js +9 -9
  218. package/scripts/sync-paper-artifacts.mjs +5 -1
  219. package/scripts/verify-arxiv-compile.mjs +52 -16
  220. package/scripts/verify-arxiv-source.mjs +45 -15
  221. package/scripts/verify-browser-launch-plan.mjs +28 -11
  222. package/scripts/verify-browser-launch-results.mjs +32 -14
  223. package/scripts/verify-paper-artifacts.mjs +539 -79
  224. package/scripts/verify-paper-claims.mjs +48 -20
  225. package/scripts/verify-paper-submission-bundle.mjs +22 -11
  226. package/scripts/verify-publication-pack.mjs +23 -9
  227. package/scripts/verify-release-readiness.mjs +211 -76
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "suite": "GuardBench comparative",
3
- "generatedAt": "2026-05-15T17:52:12.761Z",
3
+ "generatedAt": "2026-05-29T03:45:36.607Z",
4
4
  "manifest": {
5
5
  "manifestVersion": "0.2.0",
6
6
  "suiteId": "guardbench-local-comparative",
@@ -416,14 +416,14 @@
416
416
  ]
417
417
  },
418
418
  "provenance": {
419
- "generatedAt": "2026-05-15T17:52:12.761Z",
420
- "gitSha": "82b0e9979680acf751b9e80f6f90f8c6ac74befb",
419
+ "generatedAt": "2026-05-29T03:45:36.607Z",
420
+ "gitSha": "ceed2f51b615175c8bb412b96b5e5a501561189f",
421
421
  "gitDirty": false,
422
- "node": "v24.15.0",
423
- "v8": "13.6.233.17-node.48",
422
+ "node": "v24.16.0",
423
+ "v8": "13.6.233.17-node.49",
424
424
  "platform": "linux",
425
425
  "arch": "x64",
426
- "osRelease": "6.17.0-1013-azure",
426
+ "osRelease": "6.17.0-1015-azure",
427
427
  "cpuModel": "AMD EPYC 9V74 80-Core Processor",
428
428
  "cpuCount": 4,
429
429
  "totalMemoryGb": 15.61,
@@ -449,14 +449,14 @@
449
449
  "redactionLeaks": 0,
450
450
  "recallDegradationDetectionRate": 1,
451
451
  "latency": {
452
- "p50Ms": 2.465,
453
- "p95Ms": 30.791,
454
- "maxMs": 30.791
452
+ "p50Ms": 2.916,
453
+ "p95Ms": 21.17,
454
+ "maxMs": 21.17
455
455
  },
456
456
  "systemSummaries": [
457
457
  {
458
458
  "system": "Audrey Guard",
459
- "generatedAt": "2026-05-15T17:52:12.760Z",
459
+ "generatedAt": "2026-05-29T03:45:36.606Z",
460
460
  "scenarios": 10,
461
461
  "passed": 10,
462
462
  "passRate": 1,
@@ -470,14 +470,14 @@
470
470
  "redactionLeaks": 0,
471
471
  "recallDegradationDetectionRate": 1,
472
472
  "latency": {
473
- "p50Ms": 2.465,
474
- "p95Ms": 30.791,
475
- "maxMs": 30.791
473
+ "p50Ms": 2.916,
474
+ "p95Ms": 21.17,
475
+ "maxMs": 21.17
476
476
  }
477
477
  },
478
478
  {
479
479
  "system": "No Memory",
480
- "generatedAt": "2026-05-15T17:52:12.760Z",
480
+ "generatedAt": "2026-05-29T03:45:36.606Z",
481
481
  "scenarios": 10,
482
482
  "passed": 1,
483
483
  "passRate": 0.1,
@@ -491,14 +491,14 @@
491
491
  "redactionLeaks": 0,
492
492
  "recallDegradationDetectionRate": 0,
493
493
  "latency": {
494
- "p50Ms": 0.006,
495
- "p95Ms": 0.028,
496
- "maxMs": 0.028
494
+ "p50Ms": 0.007,
495
+ "p95Ms": 0.035,
496
+ "maxMs": 0.035
497
497
  }
498
498
  },
499
499
  {
500
500
  "system": "Recent Window",
501
- "generatedAt": "2026-05-15T17:52:12.760Z",
501
+ "generatedAt": "2026-05-29T03:45:36.606Z",
502
502
  "scenarios": 10,
503
503
  "passed": 6,
504
504
  "passRate": 0.6,
@@ -512,14 +512,14 @@
512
512
  "redactionLeaks": 0,
513
513
  "recallDegradationDetectionRate": 0.5,
514
514
  "latency": {
515
- "p50Ms": 0.103,
516
- "p95Ms": 0.462,
517
- "maxMs": 0.462
515
+ "p50Ms": 0.106,
516
+ "p95Ms": 0.421,
517
+ "maxMs": 0.421
518
518
  }
519
519
  },
520
520
  {
521
521
  "system": "Vector Only",
522
- "generatedAt": "2026-05-15T17:52:12.761Z",
522
+ "generatedAt": "2026-05-29T03:45:36.606Z",
523
523
  "scenarios": 10,
524
524
  "passed": 4,
525
525
  "passRate": 0.4,
@@ -533,14 +533,14 @@
533
533
  "redactionLeaks": 0,
534
534
  "recallDegradationDetectionRate": 0,
535
535
  "latency": {
536
- "p50Ms": 0.35,
537
- "p95Ms": 1.051,
538
- "maxMs": 1.051
536
+ "p50Ms": 0.451,
537
+ "p95Ms": 1.551,
538
+ "maxMs": 1.551
539
539
  }
540
540
  },
541
541
  {
542
542
  "system": "FTS Only",
543
- "generatedAt": "2026-05-15T17:52:12.761Z",
543
+ "generatedAt": "2026-05-29T03:45:36.606Z",
544
544
  "scenarios": 10,
545
545
  "passed": 1,
546
546
  "passRate": 0.1,
@@ -554,16 +554,16 @@
554
554
  "redactionLeaks": 0,
555
555
  "recallDegradationDetectionRate": 0,
556
556
  "latency": {
557
- "p50Ms": 0.348,
558
- "p95Ms": 0.545,
559
- "maxMs": 0.545
557
+ "p50Ms": 0.43,
558
+ "p95Ms": 0.717,
559
+ "maxMs": 0.717
560
560
  }
561
561
  }
562
562
  ],
563
563
  "comparisons": {
564
564
  "bestBaseline": {
565
565
  "system": "Recent Window",
566
- "generatedAt": "2026-05-15T17:52:12.760Z",
566
+ "generatedAt": "2026-05-29T03:45:36.606Z",
567
567
  "scenarios": 10,
568
568
  "passed": 6,
569
569
  "passRate": 0.6,
@@ -577,9 +577,9 @@
577
577
  "redactionLeaks": 0,
578
578
  "recallDegradationDetectionRate": 0.5,
579
579
  "latency": {
580
- "p50Ms": 0.103,
581
- "p95Ms": 0.462,
582
- "maxMs": 0.462
580
+ "p50Ms": 0.106,
581
+ "p95Ms": 0.421,
582
+ "maxMs": 0.421
583
583
  }
584
584
  },
585
585
  "audreyMarginOverBestBaseline": 0.4
@@ -594,11 +594,11 @@
594
594
  "decisionCorrect": true,
595
595
  "riskScore": 0.9,
596
596
  "passed": true,
597
- "latencyMs": 6.135,
597
+ "latencyMs": 8.156,
598
598
  "evidenceCount": 2,
599
599
  "evidenceIds": [
600
- "01KRPC8HJAWZEY0085CJJHK54G",
601
- "failure:Bash:2026-05-15T17:52:01.099Z"
600
+ "01KSRXCP2KSTEGGHFKA5WRR1S8",
601
+ "failure:Bash:2026-05-29T03:45:33.268Z"
602
602
  ],
603
603
  "recommendedActions": [
604
604
  "Do not repeat the exact failed action until the prior error is understood or the command is changed.",
@@ -620,10 +620,10 @@
620
620
  "decisionCorrect": true,
621
621
  "riskScore": 0.85,
622
622
  "passed": true,
623
- "latencyMs": 1.96,
623
+ "latencyMs": 2.277,
624
624
  "evidenceCount": 1,
625
625
  "evidenceIds": [
626
- "01KRPC8HPQ1DDFJ3F929DEEJEB"
626
+ "01KSRXCP4XYS02ZWPSKVYA7A0G"
627
627
  ],
628
628
  "recommendedActions": [
629
629
  "Do not proceed until the high-severity memory warning is addressed.",
@@ -645,10 +645,10 @@
645
645
  "decisionCorrect": true,
646
646
  "riskScore": 0.55,
647
647
  "passed": true,
648
- "latencyMs": 2.654,
648
+ "latencyMs": 3.184,
649
649
  "evidenceCount": 1,
650
650
  "evidenceIds": [
651
- "failure:Bash:2026-05-15T17:52:01.365Z"
651
+ "failure:Bash:2026-05-29T03:45:33.404Z"
652
652
  ],
653
653
  "recommendedActions": [
654
654
  "Before re-running Bash, check what changed since the last failure."
@@ -669,10 +669,10 @@
669
669
  "decisionCorrect": true,
670
670
  "riskScore": 0.55,
671
671
  "passed": true,
672
- "latencyMs": 2.465,
672
+ "latencyMs": 2.916,
673
673
  "evidenceCount": 1,
674
674
  "evidenceIds": [
675
- "failure:Bash:2026-05-15T17:52:01.798Z"
675
+ "failure:Bash:2026-05-29T03:45:33.469Z"
676
676
  ],
677
677
  "recommendedActions": [
678
678
  "Before re-running Bash, check what changed since the last failure."
@@ -693,11 +693,11 @@
693
693
  "decisionCorrect": true,
694
694
  "riskScore": 0.2,
695
695
  "passed": true,
696
- "latencyMs": 2.485,
696
+ "latencyMs": 3.161,
697
697
  "evidenceCount": 2,
698
698
  "evidenceIds": [
699
- "01KRPC8JAPXFTFGGG94QP185MS",
700
- "failure:Bash:2026-05-15T17:52:01.877Z"
699
+ "01KSRXCPAXZX9BGBD93N5CDDCM",
700
+ "failure:Bash:2026-05-29T03:45:33.531Z"
701
701
  ],
702
702
  "recommendedActions": [
703
703
  "This exact action has succeeded since its last failure; proceed with normal validation.",
@@ -719,7 +719,7 @@
719
719
  "decisionCorrect": true,
720
720
  "riskScore": 0.85,
721
721
  "passed": true,
722
- "latencyMs": 2.159,
722
+ "latencyMs": 2.647,
723
723
  "evidenceCount": 1,
724
724
  "evidenceIds": [
725
725
  "recall:episodic:recall.vector_counts"
@@ -745,11 +745,11 @@
745
745
  "decisionCorrect": true,
746
746
  "riskScore": 0.85,
747
747
  "passed": true,
748
- "latencyMs": 1.561,
748
+ "latencyMs": 1.934,
749
749
  "evidenceCount": 2,
750
750
  "evidenceIds": [
751
751
  "recall:fts:recall.fts_lookup",
752
- "01KRPC8JKB36TE59QKA7Z4V2DM"
752
+ "01KSRXCPEXC1RDR4VFSV3ZV759"
753
753
  ],
754
754
  "recommendedActions": [
755
755
  "Do not proceed until the high-severity memory warning is addressed.",
@@ -772,11 +772,11 @@
772
772
  "decisionCorrect": true,
773
773
  "riskScore": 0.9,
774
774
  "passed": true,
775
- "latencyMs": 2.339,
775
+ "latencyMs": 2.599,
776
776
  "evidenceCount": 2,
777
777
  "evidenceIds": [
778
- "01KRPC8JQFVTGQBPCSTSKTRPY7",
779
- "failure:Bash:2026-05-15T17:52:02.287Z"
778
+ "01KSRXCPGV1X3H49QBRCN72084",
779
+ "failure:Bash:2026-05-29T03:45:33.723Z"
780
780
  ],
781
781
  "recommendedActions": [
782
782
  "Do not repeat the exact failed action until the prior error is understood or the command is changed.",
@@ -798,11 +798,11 @@
798
798
  "decisionCorrect": true,
799
799
  "riskScore": 0.85,
800
800
  "passed": true,
801
- "latencyMs": 1.963,
801
+ "latencyMs": 2.391,
802
802
  "evidenceCount": 2,
803
803
  "evidenceIds": [
804
- "01KRPC8K2N9C3SKKD835K921Z8",
805
- "01KRPC8K2PVBNMYZ2RBA7B2Q9X"
804
+ "01KSRXCPJTXVN9X36WASHM2QY6",
805
+ "01KSRXCPJV1JQBFZ19K6H796AG"
806
806
  ],
807
807
  "recommendedActions": [
808
808
  "Do not proceed until the high-severity memory warning is addressed.",
@@ -824,29 +824,28 @@
824
824
  "decisionCorrect": true,
825
825
  "riskScore": 0.85,
826
826
  "passed": true,
827
- "latencyMs": 30.791,
828
- "evidenceCount": 13,
827
+ "latencyMs": 21.17,
828
+ "evidenceCount": 12,
829
829
  "evidenceIds": [
830
- "01KRPC8PQ72DA5K79S9YZ7N381",
831
- "01KRPC8PQ6YCVWK55HP85M0JKB",
832
- "01KRPC8PMZ7SZFK6P2HCZQF23X",
833
- "01KRPC8PHVXXXJ1HRFGXQ9SNZD",
834
- "01KRPC8PE7CP3E77NRQKFWB01Z",
835
- "01KRPC8PC7C083T4QRW0PB54W0",
836
- "01KRPC8P76C1BBHBKMW79XHVPA",
837
- "01KRPC8NSJ25DKGHN9RM5EKGSZ",
838
- "01KRPC8NSFC7N7AHWGCBNHXP2P",
839
- "01KRPC8MWXZ9DVQJ2QAFM2EJJC",
840
- "01KRPC8MV37S2ZR305M1PCPCJA",
841
- "01KRPC8KZNCXB2CYDMJ6QVV5CJ",
842
- "01KRPC8K5SHHV6HE5MQ10DSKAT"
830
+ "01KSRXCQ4DK284E35ZKNYDXWBQ",
831
+ "01KSRXCQ3H1TVR4E552DQVV9MG",
832
+ "01KSRXCQ39QKSSNZWEFZBHMNT9",
833
+ "01KSRXCPYP6VKM8AZC7KZ4SN6W",
834
+ "01KSRXCPVWWCPWE3M38G6VM1BG",
835
+ "01KSRXCPV7YMCBAT0602VZ3DQG",
836
+ "01KSRXCPTM8GHZXKXNKH5FMRG6",
837
+ "01KSRXCPSNJTZHJK1MWE6WNNYW",
838
+ "01KSRXCPS3K2GR6MFXTMTDEKD3",
839
+ "01KSRXCPS3K2GR6MFXTMTDEKD2",
840
+ "01KSRXCPRGC2EN41NQD4MYJ1Q1",
841
+ "01KSRXCPNG135506TFPF1WMAVB"
843
842
  ],
844
843
  "recommendedActions": [
845
844
  "Do not proceed until the high-severity memory warning is addressed.",
846
845
  "Apply this must-follow rule before acting.",
847
846
  "Treat this as uncertain context and verify before relying on it."
848
847
  ],
849
- "summary": "Blocked: 13 memory signals, 1 high severity, 12 medium severity found before acting.",
848
+ "summary": "Blocked: 12 memory signals, 1 high severity, 11 medium severity found before acting.",
850
849
  "recallErrors": [],
851
850
  "leakedSecrets": [],
852
851
  "hasEvidenceForDecision": true,
@@ -869,11 +868,11 @@
869
868
  "decisionCorrect": true,
870
869
  "riskScore": 0.9,
871
870
  "passed": true,
872
- "latencyMs": 6.135,
871
+ "latencyMs": 8.156,
873
872
  "evidenceCount": 2,
874
873
  "evidenceIds": [
875
- "01KRPC8HJAWZEY0085CJJHK54G",
876
- "failure:Bash:2026-05-15T17:52:01.099Z"
874
+ "01KSRXCP2KSTEGGHFKA5WRR1S8",
875
+ "failure:Bash:2026-05-29T03:45:33.268Z"
877
876
  ],
878
877
  "recommendedActions": [
879
878
  "Do not repeat the exact failed action until the prior error is understood or the command is changed.",
@@ -895,7 +894,7 @@
895
894
  "decisionCorrect": false,
896
895
  "riskScore": 0,
897
896
  "passed": false,
898
- "latencyMs": 0.028,
897
+ "latencyMs": 0.035,
899
898
  "evidenceCount": 0,
900
899
  "evidenceIds": [],
901
900
  "recommendedActions": [],
@@ -915,10 +914,10 @@
915
914
  "decisionCorrect": false,
916
915
  "riskScore": 0.55,
917
916
  "passed": false,
918
- "latencyMs": 0.185,
917
+ "latencyMs": 0.223,
919
918
  "evidenceCount": 1,
920
919
  "evidenceIds": [
921
- "01KRPC8HKK5GH11P78E2KB66KW"
920
+ "01KSRXCP3QFKSFEB9TZNQV7P48"
922
921
  ],
923
922
  "recommendedActions": [
924
923
  "Check the recent failed event before repeating a similar action."
@@ -939,10 +938,10 @@
939
938
  "decisionCorrect": false,
940
939
  "riskScore": 0.35,
941
940
  "passed": false,
942
- "latencyMs": 0.742,
941
+ "latencyMs": 0.945,
943
942
  "evidenceCount": 1,
944
943
  "evidenceIds": [
945
- "01KRPC8HM92Q8XXJDSGGCKV8XP"
944
+ "01KSRXCP44MHYS3JJJN1C3H0SR"
946
945
  ],
947
946
  "recommendedActions": [
948
947
  "Treat retrieved memory as advisory context."
@@ -963,7 +962,7 @@
963
962
  "decisionCorrect": false,
964
963
  "riskScore": 0,
965
964
  "passed": false,
966
- "latencyMs": 0.418,
965
+ "latencyMs": 0.467,
967
966
  "evidenceCount": 0,
968
967
  "evidenceIds": [],
969
968
  "recommendedActions": [],
@@ -990,10 +989,10 @@
990
989
  "decisionCorrect": true,
991
990
  "riskScore": 0.85,
992
991
  "passed": true,
993
- "latencyMs": 1.96,
992
+ "latencyMs": 2.277,
994
993
  "evidenceCount": 1,
995
994
  "evidenceIds": [
996
- "01KRPC8HPQ1DDFJ3F929DEEJEB"
995
+ "01KSRXCP4XYS02ZWPSKVYA7A0G"
997
996
  ],
998
997
  "recommendedActions": [
999
998
  "Do not proceed until the high-severity memory warning is addressed.",
@@ -1015,7 +1014,7 @@
1015
1014
  "decisionCorrect": false,
1016
1015
  "riskScore": 0,
1017
1016
  "passed": false,
1018
- "latencyMs": 0.006,
1017
+ "latencyMs": 0.004,
1019
1018
  "evidenceCount": 0,
1020
1019
  "evidenceIds": [],
1021
1020
  "recommendedActions": [],
@@ -1035,10 +1034,10 @@
1035
1034
  "decisionCorrect": true,
1036
1035
  "riskScore": 0.85,
1037
1036
  "passed": true,
1038
- "latencyMs": 0.298,
1037
+ "latencyMs": 0.322,
1039
1038
  "evidenceCount": 1,
1040
1039
  "evidenceIds": [
1041
- "01KRPC8HS2VFCEMKZT27Y7J289"
1040
+ "01KSRXCP5QE5HDWCC6T278MZFJ"
1042
1041
  ],
1043
1042
  "recommendedActions": [
1044
1043
  "Review retrieved memory before acting."
@@ -1059,10 +1058,10 @@
1059
1058
  "decisionCorrect": true,
1060
1059
  "riskScore": 0.85,
1061
1060
  "passed": true,
1062
- "latencyMs": 0.475,
1061
+ "latencyMs": 0.508,
1063
1062
  "evidenceCount": 1,
1064
1063
  "evidenceIds": [
1065
- "01KRPC8HSJ7N9KKFGH3EZGTFWP"
1064
+ "01KSRXCP63JE2K5CC3CMMWRZVM"
1066
1065
  ],
1067
1066
  "recommendedActions": [
1068
1067
  "Review retrieved memory before acting."
@@ -1083,7 +1082,7 @@
1083
1082
  "decisionCorrect": false,
1084
1083
  "riskScore": 0,
1085
1084
  "passed": false,
1086
- "latencyMs": 0.353,
1085
+ "latencyMs": 0.428,
1087
1086
  "evidenceCount": 0,
1088
1087
  "evidenceIds": [],
1089
1088
  "recommendedActions": [],
@@ -1110,10 +1109,10 @@
1110
1109
  "decisionCorrect": true,
1111
1110
  "riskScore": 0.55,
1112
1111
  "passed": true,
1113
- "latencyMs": 2.654,
1112
+ "latencyMs": 3.184,
1114
1113
  "evidenceCount": 1,
1115
1114
  "evidenceIds": [
1116
- "failure:Bash:2026-05-15T17:52:01.365Z"
1115
+ "failure:Bash:2026-05-29T03:45:33.404Z"
1117
1116
  ],
1118
1117
  "recommendedActions": [
1119
1118
  "Before re-running Bash, check what changed since the last failure."
@@ -1134,7 +1133,7 @@
1134
1133
  "decisionCorrect": false,
1135
1134
  "riskScore": 0,
1136
1135
  "passed": false,
1137
- "latencyMs": 0.007,
1136
+ "latencyMs": 0.005,
1138
1137
  "evidenceCount": 0,
1139
1138
  "evidenceIds": [],
1140
1139
  "recommendedActions": [],
@@ -1154,10 +1153,10 @@
1154
1153
  "decisionCorrect": true,
1155
1154
  "riskScore": 0.55,
1156
1155
  "passed": true,
1157
- "latencyMs": 0.079,
1156
+ "latencyMs": 0.071,
1158
1157
  "evidenceCount": 1,
1159
1158
  "evidenceIds": [
1160
- "01KRPC8J3S4YPWSTE26SD6SVPY"
1159
+ "01KSRXCP7QG82ZMEVDA0WPBDGS"
1161
1160
  ],
1162
1161
  "recommendedActions": [
1163
1162
  "Check the recent failed event before repeating a similar action."
@@ -1178,10 +1177,10 @@
1178
1177
  "decisionCorrect": true,
1179
1178
  "riskScore": 0.35,
1180
1179
  "passed": true,
1181
- "latencyMs": 0.492,
1180
+ "latencyMs": 0.523,
1182
1181
  "evidenceCount": 1,
1183
1182
  "evidenceIds": [
1184
- "01KRPC8J4BXV5DHC2RY9YBGZZC"
1183
+ "01KSRXCP85A8VZCV4YFYWDBZY4"
1185
1184
  ],
1186
1185
  "recommendedActions": [
1187
1186
  "Treat retrieved memory as advisory context."
@@ -1202,7 +1201,7 @@
1202
1201
  "decisionCorrect": false,
1203
1202
  "riskScore": 0,
1204
1203
  "passed": false,
1205
- "latencyMs": 0.381,
1204
+ "latencyMs": 0.457,
1206
1205
  "evidenceCount": 0,
1207
1206
  "evidenceIds": [],
1208
1207
  "recommendedActions": [],
@@ -1229,10 +1228,10 @@
1229
1228
  "decisionCorrect": true,
1230
1229
  "riskScore": 0.55,
1231
1230
  "passed": true,
1232
- "latencyMs": 2.465,
1231
+ "latencyMs": 2.916,
1233
1232
  "evidenceCount": 1,
1234
1233
  "evidenceIds": [
1235
- "failure:Bash:2026-05-15T17:52:01.798Z"
1234
+ "failure:Bash:2026-05-29T03:45:33.469Z"
1236
1235
  ],
1237
1236
  "recommendedActions": [
1238
1237
  "Before re-running Bash, check what changed since the last failure."
@@ -1253,7 +1252,7 @@
1253
1252
  "decisionCorrect": false,
1254
1253
  "riskScore": 0,
1255
1254
  "passed": false,
1256
- "latencyMs": 0.006,
1255
+ "latencyMs": 0.008,
1257
1256
  "evidenceCount": 0,
1258
1257
  "evidenceIds": [],
1259
1258
  "recommendedActions": [],
@@ -1273,10 +1272,10 @@
1273
1272
  "decisionCorrect": true,
1274
1273
  "riskScore": 0.55,
1275
1274
  "passed": true,
1276
- "latencyMs": 0.052,
1275
+ "latencyMs": 0.079,
1277
1276
  "evidenceCount": 1,
1278
1277
  "evidenceIds": [
1279
- "01KRPC8J98CDFSGA2AG5E56TN0"
1278
+ "01KSRXCP9QC1E626F1EMNFSYVM"
1280
1279
  ],
1281
1280
  "recommendedActions": [
1282
1281
  "Check the recent failed event before repeating a similar action."
@@ -1297,10 +1296,10 @@
1297
1296
  "decisionCorrect": true,
1298
1297
  "riskScore": 0.35,
1299
1298
  "passed": true,
1300
- "latencyMs": 0.315,
1299
+ "latencyMs": 0.427,
1301
1300
  "evidenceCount": 1,
1302
1301
  "evidenceIds": [
1303
- "01KRPC8J9Q0CMEHTHR4TPX8SYY"
1302
+ "01KSRXCPA32S387Y36NHQRFNPT"
1304
1303
  ],
1305
1304
  "recommendedActions": [
1306
1305
  "Treat retrieved memory as advisory context."
@@ -1321,7 +1320,7 @@
1321
1320
  "decisionCorrect": false,
1322
1321
  "riskScore": 0,
1323
1322
  "passed": false,
1324
- "latencyMs": 0.348,
1323
+ "latencyMs": 0.43,
1325
1324
  "evidenceCount": 0,
1326
1325
  "evidenceIds": [],
1327
1326
  "recommendedActions": [],
@@ -1348,11 +1347,11 @@
1348
1347
  "decisionCorrect": true,
1349
1348
  "riskScore": 0.2,
1350
1349
  "passed": true,
1351
- "latencyMs": 2.485,
1350
+ "latencyMs": 3.161,
1352
1351
  "evidenceCount": 2,
1353
1352
  "evidenceIds": [
1354
- "01KRPC8JAPXFTFGGG94QP185MS",
1355
- "failure:Bash:2026-05-15T17:52:01.877Z"
1353
+ "01KSRXCPAXZX9BGBD93N5CDDCM",
1354
+ "failure:Bash:2026-05-29T03:45:33.531Z"
1356
1355
  ],
1357
1356
  "recommendedActions": [
1358
1357
  "This exact action has succeeded since its last failure; proceed with normal validation.",
@@ -1374,7 +1373,7 @@
1374
1373
  "decisionCorrect": true,
1375
1374
  "riskScore": 0,
1376
1375
  "passed": true,
1377
- "latencyMs": 0.004,
1376
+ "latencyMs": 0.007,
1378
1377
  "evidenceCount": 0,
1379
1378
  "evidenceIds": [],
1380
1379
  "recommendedActions": [],
@@ -1394,10 +1393,10 @@
1394
1393
  "decisionCorrect": false,
1395
1394
  "riskScore": 0.55,
1396
1395
  "passed": false,
1397
- "latencyMs": 0.055,
1396
+ "latencyMs": 0.077,
1398
1397
  "evidenceCount": 1,
1399
1398
  "evidenceIds": [
1400
- "01KRPC8JBSK81DRW3SP4PGA3M0"
1399
+ "01KSRXCPBRH43HFGR41QC4C6S1"
1401
1400
  ],
1402
1401
  "recommendedActions": [
1403
1402
  "Check the recent failed event before repeating a similar action."
@@ -1418,10 +1417,10 @@
1418
1417
  "decisionCorrect": false,
1419
1418
  "riskScore": 0.35,
1420
1419
  "passed": false,
1421
- "latencyMs": 0.35,
1420
+ "latencyMs": 0.451,
1422
1421
  "evidenceCount": 1,
1423
1422
  "evidenceIds": [
1424
- "01KRPC8JC7Z8S82XZQM0MC2VED"
1423
+ "01KSRXCPC5FRNJEG2MV4DA0M6B"
1425
1424
  ],
1426
1425
  "recommendedActions": [
1427
1426
  "Treat retrieved memory as advisory context."
@@ -1442,7 +1441,7 @@
1442
1441
  "decisionCorrect": true,
1443
1442
  "riskScore": 0,
1444
1443
  "passed": true,
1445
- "latencyMs": 0.322,
1444
+ "latencyMs": 0.44,
1446
1445
  "evidenceCount": 0,
1447
1446
  "evidenceIds": [],
1448
1447
  "recommendedActions": [],
@@ -1469,7 +1468,7 @@
1469
1468
  "decisionCorrect": true,
1470
1469
  "riskScore": 0.85,
1471
1470
  "passed": true,
1472
- "latencyMs": 2.159,
1471
+ "latencyMs": 2.647,
1473
1472
  "evidenceCount": 1,
1474
1473
  "evidenceIds": [
1475
1474
  "recall:episodic:recall.vector_counts"
@@ -1495,7 +1494,7 @@
1495
1494
  "decisionCorrect": false,
1496
1495
  "riskScore": 0,
1497
1496
  "passed": false,
1498
- "latencyMs": 0.005,
1497
+ "latencyMs": 0.01,
1499
1498
  "evidenceCount": 0,
1500
1499
  "evidenceIds": [],
1501
1500
  "recommendedActions": [],
@@ -1515,10 +1514,10 @@
1515
1514
  "decisionCorrect": true,
1516
1515
  "riskScore": 0.85,
1517
1516
  "passed": true,
1518
- "latencyMs": 0.128,
1517
+ "latencyMs": 0.153,
1519
1518
  "evidenceCount": 1,
1520
1519
  "evidenceIds": [
1521
- "01KRPC8JEJYKMEDCJKMDKX3Q7H"
1520
+ "01KSRXCPDRK36MH6YDNH3JKEXF"
1522
1521
  ],
1523
1522
  "recommendedActions": [
1524
1523
  "Review retrieved memory before acting."
@@ -1539,7 +1538,7 @@
1539
1538
  "decisionCorrect": false,
1540
1539
  "riskScore": 0.55,
1541
1540
  "passed": false,
1542
- "latencyMs": 0.267,
1541
+ "latencyMs": 0.304,
1543
1542
  "evidenceCount": 0,
1544
1543
  "evidenceIds": [],
1545
1544
  "recommendedActions": [
@@ -1567,7 +1566,7 @@
1567
1566
  "decisionCorrect": false,
1568
1567
  "riskScore": 0,
1569
1568
  "passed": false,
1570
- "latencyMs": 0.334,
1569
+ "latencyMs": 0.376,
1571
1570
  "evidenceCount": 0,
1572
1571
  "evidenceIds": [],
1573
1572
  "recommendedActions": [],
@@ -1594,11 +1593,11 @@
1594
1593
  "decisionCorrect": true,
1595
1594
  "riskScore": 0.85,
1596
1595
  "passed": true,
1597
- "latencyMs": 1.561,
1596
+ "latencyMs": 1.934,
1598
1597
  "evidenceCount": 2,
1599
1598
  "evidenceIds": [
1600
1599
  "recall:fts:recall.fts_lookup",
1601
- "01KRPC8JKB36TE59QKA7Z4V2DM"
1600
+ "01KSRXCPEXC1RDR4VFSV3ZV759"
1602
1601
  ],
1603
1602
  "recommendedActions": [
1604
1603
  "Do not proceed until the high-severity memory warning is addressed.",
@@ -1621,7 +1620,7 @@
1621
1620
  "decisionCorrect": false,
1622
1621
  "riskScore": 0,
1623
1622
  "passed": false,
1624
- "latencyMs": 0.005,
1623
+ "latencyMs": 0.006,
1625
1624
  "evidenceCount": 0,
1626
1625
  "evidenceIds": [],
1627
1626
  "recommendedActions": [],
@@ -1641,10 +1640,10 @@
1641
1640
  "decisionCorrect": false,
1642
1641
  "riskScore": 0.35,
1643
1642
  "passed": false,
1644
- "latencyMs": 0.103,
1643
+ "latencyMs": 0.105,
1645
1644
  "evidenceCount": 1,
1646
1645
  "evidenceIds": [
1647
- "01KRPC8JNWTHH9J03GFSYTHN4K"
1646
+ "01KSRXCPFQ579DG3V402TKWYPM"
1648
1647
  ],
1649
1648
  "recommendedActions": [
1650
1649
  "Treat retrieved memory as advisory context."
@@ -1665,10 +1664,10 @@
1665
1664
  "decisionCorrect": false,
1666
1665
  "riskScore": 0.35,
1667
1666
  "passed": false,
1668
- "latencyMs": 0.314,
1667
+ "latencyMs": 0.347,
1669
1668
  "evidenceCount": 1,
1670
1669
  "evidenceIds": [
1671
- "01KRPC8JPC7P0SJDFCJXF222DE"
1670
+ "01KSRXCPG3Q8K0YSYA2SAVRPMM"
1672
1671
  ],
1673
1672
  "recommendedActions": [
1674
1673
  "Treat retrieved memory as advisory context."
@@ -1689,7 +1688,7 @@
1689
1688
  "decisionCorrect": false,
1690
1689
  "riskScore": 0.55,
1691
1690
  "passed": false,
1692
- "latencyMs": 0.129,
1691
+ "latencyMs": 0.13,
1693
1692
  "evidenceCount": 0,
1694
1693
  "evidenceIds": [],
1695
1694
  "recommendedActions": [
@@ -1724,11 +1723,11 @@
1724
1723
  "decisionCorrect": true,
1725
1724
  "riskScore": 0.9,
1726
1725
  "passed": true,
1727
- "latencyMs": 2.339,
1726
+ "latencyMs": 2.599,
1728
1727
  "evidenceCount": 2,
1729
1728
  "evidenceIds": [
1730
- "01KRPC8JQFVTGQBPCSTSKTRPY7",
1731
- "failure:Bash:2026-05-15T17:52:02.287Z"
1729
+ "01KSRXCPGV1X3H49QBRCN72084",
1730
+ "failure:Bash:2026-05-29T03:45:33.723Z"
1732
1731
  ],
1733
1732
  "recommendedActions": [
1734
1733
  "Do not repeat the exact failed action until the prior error is understood or the command is changed.",
@@ -1750,7 +1749,7 @@
1750
1749
  "decisionCorrect": false,
1751
1750
  "riskScore": 0,
1752
1751
  "passed": false,
1753
- "latencyMs": 0.007,
1752
+ "latencyMs": 0.005,
1754
1753
  "evidenceCount": 0,
1755
1754
  "evidenceIds": [],
1756
1755
  "recommendedActions": [],
@@ -1770,10 +1769,10 @@
1770
1769
  "decisionCorrect": false,
1771
1770
  "riskScore": 0.55,
1772
1771
  "passed": false,
1773
- "latencyMs": 0.049,
1772
+ "latencyMs": 0.062,
1774
1773
  "evidenceCount": 1,
1775
1774
  "evidenceIds": [
1776
- "01KRPC8JRKBJR9Y6CTD0D1ZX47"
1775
+ "01KSRXCPHPYBHWZKFJ5XCHY1X6"
1777
1776
  ],
1778
1777
  "recommendedActions": [
1779
1778
  "Check the recent failed event before repeating a similar action."
@@ -1794,10 +1793,10 @@
1794
1793
  "decisionCorrect": false,
1795
1794
  "riskScore": 0.35,
1796
1795
  "passed": false,
1797
- "latencyMs": 0.331,
1796
+ "latencyMs": 0.396,
1798
1797
  "evidenceCount": 1,
1799
1798
  "evidenceIds": [
1800
- "01KRPC8JS661GJEJJV12PR5YD3"
1799
+ "01KSRXCPJ2NXZ1VNNKPQ5RH818"
1801
1800
  ],
1802
1801
  "recommendedActions": [
1803
1802
  "Treat retrieved memory as advisory context."
@@ -1818,7 +1817,7 @@
1818
1817
  "decisionCorrect": false,
1819
1818
  "riskScore": 0,
1820
1819
  "passed": false,
1821
- "latencyMs": 0.306,
1820
+ "latencyMs": 0.35,
1822
1821
  "evidenceCount": 0,
1823
1822
  "evidenceIds": [],
1824
1823
  "recommendedActions": [],
@@ -1845,11 +1844,11 @@
1845
1844
  "decisionCorrect": true,
1846
1845
  "riskScore": 0.85,
1847
1846
  "passed": true,
1848
- "latencyMs": 1.963,
1847
+ "latencyMs": 2.391,
1849
1848
  "evidenceCount": 2,
1850
1849
  "evidenceIds": [
1851
- "01KRPC8K2N9C3SKKD835K921Z8",
1852
- "01KRPC8K2PVBNMYZ2RBA7B2Q9X"
1850
+ "01KSRXCPJTXVN9X36WASHM2QY6",
1851
+ "01KSRXCPJV1JQBFZ19K6H796AG"
1853
1852
  ],
1854
1853
  "recommendedActions": [
1855
1854
  "Do not proceed until the high-severity memory warning is addressed.",
@@ -1891,11 +1890,11 @@
1891
1890
  "decisionCorrect": true,
1892
1891
  "riskScore": 0.85,
1893
1892
  "passed": true,
1894
- "latencyMs": 0.088,
1893
+ "latencyMs": 0.106,
1895
1894
  "evidenceCount": 2,
1896
1895
  "evidenceIds": [
1897
- "01KRPC8K3V9JQY1TQFJQDWGHGM",
1898
- "01KRPC8K3TKQJ65V280YRNH91B"
1896
+ "01KSRXCPKNY5BNX2TH3M407J48",
1897
+ "01KSRXCPKMTBHPCWYJWJ3REV9J"
1899
1898
  ],
1900
1899
  "recommendedActions": [
1901
1900
  "Review retrieved memory before acting."
@@ -1916,11 +1915,11 @@
1916
1915
  "decisionCorrect": true,
1917
1916
  "riskScore": 0.85,
1918
1917
  "passed": true,
1919
- "latencyMs": 0.345,
1918
+ "latencyMs": 0.4,
1920
1919
  "evidenceCount": 2,
1921
1920
  "evidenceIds": [
1922
- "01KRPC8K4CWWGSHGAYZ5JDF62G",
1923
- "01KRPC8K4D0KF2C6EW79KC869P"
1921
+ "01KSRXCPM08WEJAJ579D9KS053",
1922
+ "01KSRXCPM1V6CVS8380AN3F39Y"
1924
1923
  ],
1925
1924
  "recommendedActions": [
1926
1925
  "Review retrieved memory before acting."
@@ -1941,7 +1940,7 @@
1941
1940
  "decisionCorrect": false,
1942
1941
  "riskScore": 0,
1943
1942
  "passed": false,
1944
- "latencyMs": 0.329,
1943
+ "latencyMs": 0.378,
1945
1944
  "evidenceCount": 0,
1946
1945
  "evidenceIds": [],
1947
1946
  "recommendedActions": [],
@@ -1968,29 +1967,28 @@
1968
1967
  "decisionCorrect": true,
1969
1968
  "riskScore": 0.85,
1970
1969
  "passed": true,
1971
- "latencyMs": 30.791,
1972
- "evidenceCount": 13,
1970
+ "latencyMs": 21.17,
1971
+ "evidenceCount": 12,
1973
1972
  "evidenceIds": [
1974
- "01KRPC8PQ72DA5K79S9YZ7N381",
1975
- "01KRPC8PQ6YCVWK55HP85M0JKB",
1976
- "01KRPC8PMZ7SZFK6P2HCZQF23X",
1977
- "01KRPC8PHVXXXJ1HRFGXQ9SNZD",
1978
- "01KRPC8PE7CP3E77NRQKFWB01Z",
1979
- "01KRPC8PC7C083T4QRW0PB54W0",
1980
- "01KRPC8P76C1BBHBKMW79XHVPA",
1981
- "01KRPC8NSJ25DKGHN9RM5EKGSZ",
1982
- "01KRPC8NSFC7N7AHWGCBNHXP2P",
1983
- "01KRPC8MWXZ9DVQJ2QAFM2EJJC",
1984
- "01KRPC8MV37S2ZR305M1PCPCJA",
1985
- "01KRPC8KZNCXB2CYDMJ6QVV5CJ",
1986
- "01KRPC8K5SHHV6HE5MQ10DSKAT"
1973
+ "01KSRXCQ4DK284E35ZKNYDXWBQ",
1974
+ "01KSRXCQ3H1TVR4E552DQVV9MG",
1975
+ "01KSRXCQ39QKSSNZWEFZBHMNT9",
1976
+ "01KSRXCPYP6VKM8AZC7KZ4SN6W",
1977
+ "01KSRXCPVWWCPWE3M38G6VM1BG",
1978
+ "01KSRXCPV7YMCBAT0602VZ3DQG",
1979
+ "01KSRXCPTM8GHZXKXNKH5FMRG6",
1980
+ "01KSRXCPSNJTZHJK1MWE6WNNYW",
1981
+ "01KSRXCPS3K2GR6MFXTMTDEKD3",
1982
+ "01KSRXCPS3K2GR6MFXTMTDEKD2",
1983
+ "01KSRXCPRGC2EN41NQD4MYJ1Q1",
1984
+ "01KSRXCPNG135506TFPF1WMAVB"
1987
1985
  ],
1988
1986
  "recommendedActions": [
1989
1987
  "Do not proceed until the high-severity memory warning is addressed.",
1990
1988
  "Apply this must-follow rule before acting.",
1991
1989
  "Treat this as uncertain context and verify before relying on it."
1992
1990
  ],
1993
- "summary": "Blocked: 13 memory signals, 1 high severity, 12 medium severity found before acting.",
1991
+ "summary": "Blocked: 12 memory signals, 1 high severity, 11 medium severity found before acting.",
1994
1992
  "recallErrors": [],
1995
1993
  "leakedSecrets": [],
1996
1994
  "hasEvidenceForDecision": true,
@@ -2006,7 +2004,7 @@
2006
2004
  "decisionCorrect": false,
2007
2005
  "riskScore": 0,
2008
2006
  "passed": false,
2009
- "latencyMs": 0.009,
2007
+ "latencyMs": 0.011,
2010
2008
  "evidenceCount": 0,
2011
2009
  "evidenceIds": [],
2012
2010
  "recommendedActions": [],
@@ -2026,10 +2024,10 @@
2026
2024
  "decisionCorrect": true,
2027
2025
  "riskScore": 0.85,
2028
2026
  "passed": true,
2029
- "latencyMs": 0.462,
2027
+ "latencyMs": 0.421,
2030
2028
  "evidenceCount": 1,
2031
2029
  "evidenceIds": [
2032
- "01KRPC8V0CK77K0V6ZKRC1T15A"
2030
+ "01KSRXCR6FARVQ7ATWYWC5QAF9"
2033
2031
  ],
2034
2032
  "recommendedActions": [
2035
2033
  "Review retrieved memory before acting."
@@ -2050,14 +2048,14 @@
2050
2048
  "decisionCorrect": false,
2051
2049
  "riskScore": 0.35,
2052
2050
  "passed": false,
2053
- "latencyMs": 1.051,
2051
+ "latencyMs": 1.551,
2054
2052
  "evidenceCount": 5,
2055
2053
  "evidenceIds": [
2056
- "01KRPC8VCE8VNRWCGWMC1VYNA9",
2057
- "01KRPC8VCG3GQ7EPDPV9RQ23JA",
2058
- "01KRPC8VG14K20MGW0C8N1WDGH",
2059
- "01KRPC8VK28WY2BM5BB3AR9NPA",
2060
- "01KRPC8V709AFR44CVFQB5MAFW"
2054
+ "01KSRXCRC5YBFBKT1RM4SPXRZZ",
2055
+ "01KSRXCR9R09K2J5HM1BGN1PSW",
2056
+ "01KSRXCRJKF9PWQG7YRGGK1TP6",
2057
+ "01KSRXCRGTW1V1VGWWT869D36Q",
2058
+ "01KSRXCRP3KVXPGD7WMNS3KWKF"
2061
2059
  ],
2062
2060
  "recommendedActions": [
2063
2061
  "Treat retrieved memory as advisory context."
@@ -2078,7 +2076,7 @@
2078
2076
  "decisionCorrect": false,
2079
2077
  "riskScore": 0,
2080
2078
  "passed": false,
2081
- "latencyMs": 0.545,
2079
+ "latencyMs": 0.717,
2082
2080
  "evidenceCount": 0,
2083
2081
  "evidenceIds": [],
2084
2082
  "recommendedActions": [],
@@ -2093,7 +2091,7 @@
2093
2091
  }
2094
2092
  ],
2095
2093
  "artifactRedactionSweep": {
2096
- "checkedAt": "2026-05-15T17:52:12.780Z",
2094
+ "checkedAt": "2026-05-29T03:45:36.646Z",
2097
2095
  "filesChecked": [
2098
2096
  "benchmarks/output/guardbench-manifest.json",
2099
2097
  "benchmarks/output/guardbench-raw.json",