audrey 0.23.1 → 1.0.1

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 (250) hide show
  1. package/CHANGELOG.md +101 -15
  2. package/LICENSE +21 -21
  3. package/README.md +232 -6
  4. package/SECURITY.md +2 -1
  5. package/benchmarks/adapter-kit.mjs +20 -0
  6. package/benchmarks/adapter-self-test.mjs +166 -0
  7. package/benchmarks/adapters/example-allow.mjs +28 -0
  8. package/benchmarks/adapters/mem0-platform.mjs +267 -0
  9. package/benchmarks/adapters/registry.json +51 -0
  10. package/benchmarks/adapters/zep-cloud.mjs +280 -0
  11. package/benchmarks/baselines.js +169 -0
  12. package/benchmarks/build-leaderboard.mjs +170 -0
  13. package/benchmarks/cases.js +537 -0
  14. package/benchmarks/create-conformance-card.mjs +139 -0
  15. package/benchmarks/create-submission-bundle.mjs +176 -0
  16. package/benchmarks/dry-run-external-adapters.mjs +165 -0
  17. package/benchmarks/guardbench.js +1125 -0
  18. package/benchmarks/output/adapter-self-test/guardbench-adapter-self-test.json +50 -0
  19. package/benchmarks/output/external/guardbench-external-dry-run.json +69 -0
  20. package/benchmarks/output/external/guardbench-external-evidence.json +56 -0
  21. package/benchmarks/output/guardbench-conformance-card.json +63 -0
  22. package/benchmarks/output/guardbench-manifest.json +414 -0
  23. package/benchmarks/output/guardbench-raw.json +1271 -0
  24. package/benchmarks/output/guardbench-summary.json +2107 -0
  25. package/benchmarks/output/leaderboard/guardbench-leaderboard.json +93 -0
  26. package/benchmarks/output/leaderboard/guardbench-leaderboard.md +7 -0
  27. package/benchmarks/output/submission-bundle/guardbench-conformance-card.json +63 -0
  28. package/benchmarks/output/submission-bundle/guardbench-manifest.json +414 -0
  29. package/benchmarks/output/submission-bundle/guardbench-raw.json +1271 -0
  30. package/benchmarks/output/submission-bundle/guardbench-summary.json +2107 -0
  31. package/benchmarks/output/submission-bundle/schemas/guardbench-adapter-registry.schema.json +69 -0
  32. package/benchmarks/output/submission-bundle/schemas/guardbench-adapter-self-test.schema.json +156 -0
  33. package/benchmarks/output/submission-bundle/schemas/guardbench-conformance-card.schema.json +184 -0
  34. package/benchmarks/output/submission-bundle/schemas/guardbench-external-dry-run.schema.json +74 -0
  35. package/benchmarks/output/submission-bundle/schemas/guardbench-external-evidence.schema.json +108 -0
  36. package/benchmarks/output/submission-bundle/schemas/guardbench-external-run.schema.json +160 -0
  37. package/benchmarks/output/submission-bundle/schemas/guardbench-leaderboard.schema.json +179 -0
  38. package/benchmarks/output/submission-bundle/schemas/guardbench-manifest.schema.json +213 -0
  39. package/benchmarks/output/submission-bundle/schemas/guardbench-publication-verification.schema.json +47 -0
  40. package/benchmarks/output/submission-bundle/schemas/guardbench-raw.schema.json +184 -0
  41. package/benchmarks/output/submission-bundle/schemas/guardbench-submission-manifest.schema.json +151 -0
  42. package/benchmarks/output/submission-bundle/schemas/guardbench-summary.schema.json +249 -0
  43. package/benchmarks/output/submission-bundle/submission-manifest.json +131 -0
  44. package/benchmarks/output/submission-bundle/validation-report.json +31 -0
  45. package/benchmarks/output/summary.json +2354 -0
  46. package/benchmarks/perf-snapshot.js +304 -0
  47. package/benchmarks/perf.bench.js +161 -0
  48. package/benchmarks/public-paths.mjs +78 -0
  49. package/benchmarks/reference-results.js +70 -0
  50. package/benchmarks/report.js +259 -0
  51. package/benchmarks/run-external-guardbench.mjs +281 -0
  52. package/benchmarks/run.js +682 -0
  53. package/benchmarks/schemas/guardbench-adapter-registry.schema.json +69 -0
  54. package/benchmarks/schemas/guardbench-adapter-self-test.schema.json +156 -0
  55. package/benchmarks/schemas/guardbench-conformance-card.schema.json +184 -0
  56. package/benchmarks/schemas/guardbench-external-dry-run.schema.json +74 -0
  57. package/benchmarks/schemas/guardbench-external-evidence.schema.json +108 -0
  58. package/benchmarks/schemas/guardbench-external-run.schema.json +160 -0
  59. package/benchmarks/schemas/guardbench-leaderboard.schema.json +179 -0
  60. package/benchmarks/schemas/guardbench-manifest.schema.json +213 -0
  61. package/benchmarks/schemas/guardbench-publication-verification.schema.json +47 -0
  62. package/benchmarks/schemas/guardbench-raw.schema.json +184 -0
  63. package/benchmarks/schemas/guardbench-submission-manifest.schema.json +151 -0
  64. package/benchmarks/schemas/guardbench-summary.schema.json +249 -0
  65. package/benchmarks/snapshots/perf-0.22.2.json +123 -0
  66. package/benchmarks/snapshots/perf-0.23.0.json +123 -0
  67. package/benchmarks/validate-adapter-module.mjs +104 -0
  68. package/benchmarks/validate-adapter-registry.mjs +134 -0
  69. package/benchmarks/validate-adapter-self-test.mjs +96 -0
  70. package/benchmarks/validate-guardbench-artifacts.mjs +343 -0
  71. package/benchmarks/verify-external-evidence.mjs +296 -0
  72. package/benchmarks/verify-publication-artifacts.mjs +286 -0
  73. package/benchmarks/verify-submission-bundle.mjs +167 -0
  74. package/dist/mcp-server/config.d.ts +1 -1
  75. package/dist/mcp-server/config.d.ts.map +1 -1
  76. package/dist/mcp-server/config.js +1 -1
  77. package/dist/mcp-server/config.js.map +1 -1
  78. package/dist/mcp-server/index.d.ts +65 -3
  79. package/dist/mcp-server/index.d.ts.map +1 -1
  80. package/dist/mcp-server/index.js +675 -157
  81. package/dist/mcp-server/index.js.map +1 -1
  82. package/dist/src/action-key.d.ts +9 -0
  83. package/dist/src/action-key.d.ts.map +1 -0
  84. package/dist/src/action-key.js +49 -0
  85. package/dist/src/action-key.js.map +1 -0
  86. package/dist/src/adaptive.js +5 -5
  87. package/dist/src/affect.js +8 -8
  88. package/dist/src/audrey.d.ts +13 -0
  89. package/dist/src/audrey.d.ts.map +1 -1
  90. package/dist/src/audrey.js +68 -3
  91. package/dist/src/audrey.js.map +1 -1
  92. package/dist/src/capsule.js +4 -4
  93. package/dist/src/causal.js +3 -3
  94. package/dist/src/consolidate.js +48 -48
  95. package/dist/src/controller.d.ts +78 -6
  96. package/dist/src/controller.d.ts.map +1 -1
  97. package/dist/src/controller.js +273 -53
  98. package/dist/src/controller.js.map +1 -1
  99. package/dist/src/db.js +172 -172
  100. package/dist/src/decay.js +8 -8
  101. package/dist/src/embedding.d.ts +2 -1
  102. package/dist/src/embedding.d.ts.map +1 -1
  103. package/dist/src/embedding.js +39 -29
  104. package/dist/src/embedding.js.map +1 -1
  105. package/dist/src/encode.js +6 -6
  106. package/dist/src/feedback.d.ts +6 -0
  107. package/dist/src/feedback.d.ts.map +1 -1
  108. package/dist/src/feedback.js +6 -0
  109. package/dist/src/feedback.js.map +1 -1
  110. package/dist/src/forget.js +12 -12
  111. package/dist/src/hybrid-recall.js +9 -9
  112. package/dist/src/impact.js +6 -6
  113. package/dist/src/import.d.ts +3 -3
  114. package/dist/src/import.js +41 -41
  115. package/dist/src/index.d.ts +5 -4
  116. package/dist/src/index.d.ts.map +1 -1
  117. package/dist/src/index.js +3 -3
  118. package/dist/src/index.js.map +1 -1
  119. package/dist/src/interference.js +14 -14
  120. package/dist/src/introspect.js +18 -18
  121. package/dist/src/preflight.d.ts.map +1 -1
  122. package/dist/src/preflight.js +41 -0
  123. package/dist/src/preflight.js.map +1 -1
  124. package/dist/src/promote.js +7 -7
  125. package/dist/src/prompts.js +118 -118
  126. package/dist/src/recall.js +30 -30
  127. package/dist/src/reflexes.d.ts +1 -0
  128. package/dist/src/reflexes.d.ts.map +1 -1
  129. package/dist/src/reflexes.js +3 -0
  130. package/dist/src/reflexes.js.map +1 -1
  131. package/dist/src/rollback.js +4 -4
  132. package/dist/src/routes.d.ts.map +1 -1
  133. package/dist/src/routes.js +71 -2
  134. package/dist/src/routes.js.map +1 -1
  135. package/dist/src/validate.js +25 -25
  136. package/docs/AUDREY_PAPER_OUTLINE.md +175 -0
  137. package/docs/MEMORY_BENCHMARKING.md +59 -0
  138. package/docs/PRODUCTION_BACKLOG.md +304 -0
  139. package/docs/paper/00-master.md +48 -0
  140. package/docs/paper/01-introduction.md +27 -0
  141. package/docs/paper/02-related-work.md +47 -0
  142. package/docs/paper/03-problem-definition.md +108 -0
  143. package/docs/paper/04-design.md +164 -0
  144. package/docs/paper/05-guardbench-spec.md +412 -0
  145. package/docs/paper/06-implementation.md +113 -0
  146. package/docs/paper/07-evaluation.md +168 -0
  147. package/docs/paper/08-discussion-limitations.md +61 -0
  148. package/docs/paper/09-conclusion.md +11 -0
  149. package/docs/paper/SUBMISSION_README.md +162 -0
  150. package/docs/paper/appendix-a-demo-transcript.md +114 -0
  151. package/docs/paper/arxiv-compile-report.schema.json +116 -0
  152. package/docs/paper/arxiv-source.schema.json +61 -0
  153. package/docs/paper/audrey-paper-v1.md +1106 -0
  154. package/docs/paper/browser-launch-plan.json +209 -0
  155. package/docs/paper/browser-launch-plan.schema.json +100 -0
  156. package/docs/paper/browser-launch-results.json +86 -0
  157. package/docs/paper/browser-launch-results.schema.json +66 -0
  158. package/docs/paper/claim-register.json +138 -0
  159. package/docs/paper/claim-register.schema.json +81 -0
  160. package/docs/paper/evidence-ledger.md +103 -0
  161. package/docs/paper/output/arxiv/README-arxiv.txt +8 -0
  162. package/docs/paper/output/arxiv/arxiv-manifest.json +41 -0
  163. package/docs/paper/output/arxiv/main.tex +949 -0
  164. package/docs/paper/output/arxiv/references.bib +222 -0
  165. package/docs/paper/output/arxiv-compile-report.json +24 -0
  166. package/docs/paper/output/submission-bundle/LICENSE +21 -0
  167. package/docs/paper/output/submission-bundle/README.md +555 -0
  168. package/docs/paper/output/submission-bundle/benchmarks/output/adapter-self-test/guardbench-adapter-self-test.json +50 -0
  169. package/docs/paper/output/submission-bundle/benchmarks/output/external/guardbench-external-dry-run.json +69 -0
  170. package/docs/paper/output/submission-bundle/benchmarks/output/external/guardbench-external-evidence.json +56 -0
  171. package/docs/paper/output/submission-bundle/benchmarks/output/guardbench-conformance-card.json +63 -0
  172. package/docs/paper/output/submission-bundle/benchmarks/output/guardbench-manifest.json +414 -0
  173. package/docs/paper/output/submission-bundle/benchmarks/output/guardbench-raw.json +1271 -0
  174. package/docs/paper/output/submission-bundle/benchmarks/output/guardbench-summary.json +2107 -0
  175. package/docs/paper/output/submission-bundle/benchmarks/output/leaderboard/guardbench-leaderboard.json +93 -0
  176. package/docs/paper/output/submission-bundle/benchmarks/output/leaderboard/guardbench-leaderboard.md +7 -0
  177. package/docs/paper/output/submission-bundle/benchmarks/output/submission-bundle/submission-manifest.json +131 -0
  178. package/docs/paper/output/submission-bundle/benchmarks/output/submission-bundle/validation-report.json +31 -0
  179. package/docs/paper/output/submission-bundle/benchmarks/output/summary.json +2354 -0
  180. package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-adapter-registry.schema.json +69 -0
  181. package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-adapter-self-test.schema.json +156 -0
  182. package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-conformance-card.schema.json +184 -0
  183. package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-external-dry-run.schema.json +74 -0
  184. package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-external-evidence.schema.json +108 -0
  185. package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-external-run.schema.json +160 -0
  186. package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-leaderboard.schema.json +179 -0
  187. package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-manifest.schema.json +213 -0
  188. package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-publication-verification.schema.json +47 -0
  189. package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-raw.schema.json +184 -0
  190. package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-submission-manifest.schema.json +151 -0
  191. package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-summary.schema.json +249 -0
  192. package/docs/paper/output/submission-bundle/docs/AUDREY_PAPER_OUTLINE.md +175 -0
  193. package/docs/paper/output/submission-bundle/docs/paper/00-master.md +48 -0
  194. package/docs/paper/output/submission-bundle/docs/paper/01-introduction.md +27 -0
  195. package/docs/paper/output/submission-bundle/docs/paper/02-related-work.md +47 -0
  196. package/docs/paper/output/submission-bundle/docs/paper/03-problem-definition.md +108 -0
  197. package/docs/paper/output/submission-bundle/docs/paper/04-design.md +164 -0
  198. package/docs/paper/output/submission-bundle/docs/paper/05-guardbench-spec.md +412 -0
  199. package/docs/paper/output/submission-bundle/docs/paper/06-implementation.md +113 -0
  200. package/docs/paper/output/submission-bundle/docs/paper/07-evaluation.md +168 -0
  201. package/docs/paper/output/submission-bundle/docs/paper/08-discussion-limitations.md +61 -0
  202. package/docs/paper/output/submission-bundle/docs/paper/09-conclusion.md +11 -0
  203. package/docs/paper/output/submission-bundle/docs/paper/SUBMISSION_README.md +162 -0
  204. package/docs/paper/output/submission-bundle/docs/paper/appendix-a-demo-transcript.md +114 -0
  205. package/docs/paper/output/submission-bundle/docs/paper/arxiv-compile-report.schema.json +116 -0
  206. package/docs/paper/output/submission-bundle/docs/paper/arxiv-source.schema.json +61 -0
  207. package/docs/paper/output/submission-bundle/docs/paper/audrey-paper-v1.md +1106 -0
  208. package/docs/paper/output/submission-bundle/docs/paper/browser-launch-plan.json +209 -0
  209. package/docs/paper/output/submission-bundle/docs/paper/browser-launch-plan.schema.json +100 -0
  210. package/docs/paper/output/submission-bundle/docs/paper/browser-launch-results.json +86 -0
  211. package/docs/paper/output/submission-bundle/docs/paper/browser-launch-results.schema.json +66 -0
  212. package/docs/paper/output/submission-bundle/docs/paper/claim-register.json +138 -0
  213. package/docs/paper/output/submission-bundle/docs/paper/claim-register.schema.json +81 -0
  214. package/docs/paper/output/submission-bundle/docs/paper/evidence-ledger.md +103 -0
  215. package/docs/paper/output/submission-bundle/docs/paper/output/arxiv/README-arxiv.txt +8 -0
  216. package/docs/paper/output/submission-bundle/docs/paper/output/arxiv/arxiv-manifest.json +41 -0
  217. package/docs/paper/output/submission-bundle/docs/paper/output/arxiv/main.tex +949 -0
  218. package/docs/paper/output/submission-bundle/docs/paper/output/arxiv/references.bib +222 -0
  219. package/docs/paper/output/submission-bundle/docs/paper/output/arxiv-compile-report.json +24 -0
  220. package/docs/paper/output/submission-bundle/docs/paper/paper-submission-bundle.schema.json +70 -0
  221. package/docs/paper/output/submission-bundle/docs/paper/publication-pack.json +81 -0
  222. package/docs/paper/output/submission-bundle/docs/paper/publication-pack.schema.json +60 -0
  223. package/docs/paper/output/submission-bundle/docs/paper/references.bib +222 -0
  224. package/docs/paper/output/submission-bundle/package.json +212 -0
  225. package/docs/paper/output/submission-bundle/paper-submission-manifest.json +379 -0
  226. package/docs/paper/paper-submission-bundle.schema.json +70 -0
  227. package/docs/paper/publication-pack.json +81 -0
  228. package/docs/paper/publication-pack.schema.json +60 -0
  229. package/docs/paper/references.bib +222 -0
  230. package/package.json +87 -4
  231. package/scripts/audit-release-completion.mjs +362 -0
  232. package/scripts/create-arxiv-source.mjs +362 -0
  233. package/scripts/create-paper-submission-bundle.mjs +210 -0
  234. package/scripts/finalize-release.mjs +526 -0
  235. package/scripts/prepare-release-cut.mjs +269 -0
  236. package/scripts/publish-release-bundle.mjs +209 -0
  237. package/scripts/publish-release-github-api.mjs +429 -0
  238. package/scripts/run-vitest.mjs +34 -0
  239. package/scripts/smoke-cli.js +92 -0
  240. package/scripts/sync-paper-artifacts.mjs +109 -0
  241. package/scripts/verify-arxiv-compile.mjs +440 -0
  242. package/scripts/verify-arxiv-source.mjs +194 -0
  243. package/scripts/verify-browser-launch-plan.mjs +237 -0
  244. package/scripts/verify-browser-launch-results.mjs +285 -0
  245. package/scripts/verify-paper-artifacts.mjs +338 -0
  246. package/scripts/verify-paper-claims.mjs +226 -0
  247. package/scripts/verify-paper-submission-bundle.mjs +207 -0
  248. package/scripts/verify-publication-pack.mjs +196 -0
  249. package/scripts/verify-python-package.py +201 -0
  250. package/scripts/verify-release-readiness.mjs +785 -0
@@ -0,0 +1,69 @@
1
+ {
2
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
3
+ "$id": "https://audrey-memory.org/schemas/guardbench-adapter-registry.schema.json",
4
+ "title": "GuardBench Adapter Registry",
5
+ "type": "object",
6
+ "additionalProperties": false,
7
+ "required": ["schemaVersion", "suite", "adapters"],
8
+ "properties": {
9
+ "schemaVersion": { "const": "1.0.0" },
10
+ "suite": { "const": "GuardBench adapter registry" },
11
+ "adapters": {
12
+ "type": "array",
13
+ "minItems": 1,
14
+ "items": { "$ref": "#/$defs/adapter" }
15
+ }
16
+ },
17
+ "$defs": {
18
+ "adapter": {
19
+ "type": "object",
20
+ "additionalProperties": false,
21
+ "required": [
22
+ "id",
23
+ "name",
24
+ "path",
25
+ "status",
26
+ "credentialMode",
27
+ "requiredEnv",
28
+ "description",
29
+ "commands"
30
+ ],
31
+ "properties": {
32
+ "id": {
33
+ "type": "string",
34
+ "pattern": "^[a-z0-9][a-z0-9-]*$"
35
+ },
36
+ "name": { "type": "string", "minLength": 1 },
37
+ "path": {
38
+ "type": "string",
39
+ "pattern": "^benchmarks/adapters/.+\\.mjs$"
40
+ },
41
+ "status": {
42
+ "enum": ["reference", "external-system", "experimental"]
43
+ },
44
+ "credentialMode": {
45
+ "enum": ["none", "runtime-env"]
46
+ },
47
+ "requiredEnv": {
48
+ "type": "array",
49
+ "items": {
50
+ "type": "string",
51
+ "pattern": "^[A-Z][A-Z0-9_]*$"
52
+ }
53
+ },
54
+ "description": { "type": "string", "minLength": 1 },
55
+ "commands": {
56
+ "type": "object",
57
+ "additionalProperties": false,
58
+ "required": ["moduleValidate", "selfTest", "selfTestValidate", "externalRun"],
59
+ "properties": {
60
+ "moduleValidate": { "type": "string", "minLength": 1 },
61
+ "selfTest": { "type": "string", "minLength": 1 },
62
+ "selfTestValidate": { "type": "string", "minLength": 1 },
63
+ "externalRun": { "type": "string", "minLength": 1 }
64
+ }
65
+ }
66
+ }
67
+ }
68
+ }
69
+ }
@@ -0,0 +1,156 @@
1
+ {
2
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
3
+ "$id": "https://audrey-memory.org/schemas/guardbench-adapter-self-test.schema.json",
4
+ "title": "GuardBench Adapter Self-Test",
5
+ "type": "object",
6
+ "additionalProperties": false,
7
+ "required": [
8
+ "schemaVersion",
9
+ "suite",
10
+ "generatedAt",
11
+ "ok",
12
+ "adapter",
13
+ "conformance",
14
+ "score",
15
+ "contract",
16
+ "failures"
17
+ ],
18
+ "properties": {
19
+ "schemaVersion": { "const": "1.0.0" },
20
+ "suite": { "const": "GuardBench adapter self-test" },
21
+ "generatedAt": { "type": "string", "minLength": 1 },
22
+ "ok": { "type": "boolean" },
23
+ "adapter": { "$ref": "#/$defs/adapter" },
24
+ "conformance": { "$ref": "#/$defs/conformance" },
25
+ "score": { "$ref": "#/$defs/score" },
26
+ "contract": { "$ref": "#/$defs/contract" },
27
+ "failures": {
28
+ "type": "array",
29
+ "items": { "type": "string" }
30
+ }
31
+ },
32
+ "$defs": {
33
+ "adapter": {
34
+ "type": "object",
35
+ "additionalProperties": false,
36
+ "required": ["name", "path", "moduleFile", "description"],
37
+ "properties": {
38
+ "name": { "type": "string", "minLength": 1 },
39
+ "path": { "type": "string", "minLength": 1 },
40
+ "moduleFile": { "type": "string", "minLength": 1 },
41
+ "description": {
42
+ "anyOf": [
43
+ { "type": "string", "minLength": 1 },
44
+ { "type": "null" }
45
+ ]
46
+ }
47
+ }
48
+ },
49
+ "conformance": {
50
+ "type": "object",
51
+ "additionalProperties": false,
52
+ "required": [
53
+ "ok",
54
+ "adapter",
55
+ "requestedAdapter",
56
+ "scenarios",
57
+ "expectedScenarios",
58
+ "fullContractPassRate",
59
+ "decisionAccuracy",
60
+ "redactionLeaks",
61
+ "failures"
62
+ ],
63
+ "properties": {
64
+ "ok": { "type": "boolean" },
65
+ "adapter": { "type": "string", "minLength": 1 },
66
+ "requestedAdapter": { "type": "string", "minLength": 1 },
67
+ "scenarios": { "type": "integer", "minimum": 0 },
68
+ "expectedScenarios": { "type": "integer", "minimum": 0 },
69
+ "fullContractPassRate": { "$ref": "#/$defs/nullableRate" },
70
+ "decisionAccuracy": { "$ref": "#/$defs/nullableRate" },
71
+ "redactionLeaks": { "$ref": "#/$defs/nullableNumber" },
72
+ "failures": {
73
+ "type": "array",
74
+ "items": { "type": "string" }
75
+ }
76
+ }
77
+ },
78
+ "score": {
79
+ "type": "object",
80
+ "additionalProperties": false,
81
+ "required": [
82
+ "scenarios",
83
+ "fullContractPassRate",
84
+ "decisionAccuracy",
85
+ "evidenceRecall",
86
+ "redactionLeaks",
87
+ "latency"
88
+ ],
89
+ "properties": {
90
+ "scenarios": { "type": "integer", "minimum": 0 },
91
+ "fullContractPassRate": { "$ref": "#/$defs/nullableRate" },
92
+ "decisionAccuracy": { "$ref": "#/$defs/nullableRate" },
93
+ "evidenceRecall": { "$ref": "#/$defs/nullableRate" },
94
+ "redactionLeaks": { "$ref": "#/$defs/nullableNumber" },
95
+ "latency": {
96
+ "anyOf": [
97
+ {
98
+ "type": "object",
99
+ "additionalProperties": false,
100
+ "required": ["p50Ms", "p95Ms", "maxMs"],
101
+ "properties": {
102
+ "p50Ms": { "type": "number", "minimum": 0 },
103
+ "p95Ms": { "type": "number", "minimum": 0 },
104
+ "maxMs": { "type": "number", "minimum": 0 }
105
+ }
106
+ },
107
+ { "type": "null" }
108
+ ]
109
+ }
110
+ }
111
+ },
112
+ "contract": {
113
+ "type": "object",
114
+ "additionalProperties": false,
115
+ "required": [
116
+ "expectedAnswersWithheld",
117
+ "lowScoreAllowed",
118
+ "requiredScenarioRows",
119
+ "requiredResultFields",
120
+ "redactionLeakTolerance"
121
+ ],
122
+ "properties": {
123
+ "expectedAnswersWithheld": { "const": true },
124
+ "lowScoreAllowed": { "const": true },
125
+ "requiredScenarioRows": { "type": "integer", "minimum": 1 },
126
+ "requiredResultFields": {
127
+ "type": "array",
128
+ "minItems": 6,
129
+ "items": {
130
+ "enum": [
131
+ "decision",
132
+ "riskScore",
133
+ "evidenceIds",
134
+ "recommendedActions",
135
+ "summary",
136
+ "recallErrors"
137
+ ]
138
+ }
139
+ },
140
+ "redactionLeakTolerance": { "const": 0 }
141
+ }
142
+ },
143
+ "nullableNumber": {
144
+ "anyOf": [
145
+ { "type": "number", "minimum": 0 },
146
+ { "type": "null" }
147
+ ]
148
+ },
149
+ "nullableRate": {
150
+ "anyOf": [
151
+ { "type": "number", "minimum": 0, "maximum": 1 },
152
+ { "type": "null" }
153
+ ]
154
+ }
155
+ }
156
+ }
@@ -0,0 +1,184 @@
1
+ {
2
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
3
+ "$id": "https://audrey-memory.org/schemas/guardbench-conformance-card.schema.json",
4
+ "title": "GuardBench Conformance Card",
5
+ "type": "object",
6
+ "additionalProperties": false,
7
+ "required": [
8
+ "schemaVersion",
9
+ "suite",
10
+ "generatedAt",
11
+ "sourceDir",
12
+ "manifestVersion",
13
+ "suiteId",
14
+ "subject",
15
+ "run",
16
+ "score",
17
+ "conformance",
18
+ "integrity",
19
+ "provenance"
20
+ ],
21
+ "properties": {
22
+ "schemaVersion": { "const": "1.0.0" },
23
+ "suite": { "const": "GuardBench conformance card" },
24
+ "generatedAt": { "type": "string", "minLength": 1 },
25
+ "sourceDir": { "type": "string", "minLength": 1 },
26
+ "manifestVersion": { "type": "string", "minLength": 1 },
27
+ "suiteId": { "type": "string", "minLength": 1 },
28
+ "subject": {
29
+ "type": "object",
30
+ "additionalProperties": false,
31
+ "required": ["name", "requestedAdapter", "external"],
32
+ "properties": {
33
+ "name": { "type": "string", "minLength": 1 },
34
+ "requestedAdapter": {
35
+ "anyOf": [
36
+ { "type": "string", "minLength": 1 },
37
+ { "type": "null" }
38
+ ]
39
+ },
40
+ "external": { "type": "boolean" }
41
+ }
42
+ },
43
+ "run": {
44
+ "type": "object",
45
+ "additionalProperties": false,
46
+ "required": ["status", "startedAt", "completedAt", "command", "validationCommand"],
47
+ "properties": {
48
+ "status": { "type": "string", "minLength": 1 },
49
+ "startedAt": {
50
+ "anyOf": [
51
+ { "type": "string", "minLength": 1 },
52
+ { "type": "null" }
53
+ ]
54
+ },
55
+ "completedAt": {
56
+ "anyOf": [
57
+ { "type": "string", "minLength": 1 },
58
+ { "type": "null" }
59
+ ]
60
+ },
61
+ "command": {
62
+ "anyOf": [
63
+ {
64
+ "type": "array",
65
+ "minItems": 1,
66
+ "items": { "type": "string", "minLength": 1 }
67
+ },
68
+ { "type": "null" }
69
+ ]
70
+ },
71
+ "validationCommand": {
72
+ "anyOf": [
73
+ {
74
+ "type": "array",
75
+ "minItems": 1,
76
+ "items": { "type": "string", "minLength": 1 }
77
+ },
78
+ { "type": "null" }
79
+ ]
80
+ }
81
+ }
82
+ },
83
+ "score": {
84
+ "type": "object",
85
+ "additionalProperties": false,
86
+ "required": [
87
+ "scenarios",
88
+ "fullContractPassed",
89
+ "fullContractPassRate",
90
+ "decisionAccuracy",
91
+ "evidenceRecall",
92
+ "redactionLeaks",
93
+ "latency"
94
+ ],
95
+ "properties": {
96
+ "scenarios": { "type": "integer", "minimum": 0 },
97
+ "fullContractPassed": { "$ref": "#/$defs/nullableNumber" },
98
+ "fullContractPassRate": { "$ref": "#/$defs/nullableRate" },
99
+ "decisionAccuracy": { "$ref": "#/$defs/nullableRate" },
100
+ "evidenceRecall": { "$ref": "#/$defs/nullableRate" },
101
+ "redactionLeaks": { "$ref": "#/$defs/nullableNumber" },
102
+ "latency": {
103
+ "anyOf": [
104
+ {
105
+ "type": "object",
106
+ "additionalProperties": false,
107
+ "required": ["p50Ms", "p95Ms", "maxMs"],
108
+ "properties": {
109
+ "p50Ms": { "type": "number", "minimum": 0 },
110
+ "p95Ms": { "type": "number", "minimum": 0 },
111
+ "maxMs": { "type": "number", "minimum": 0 }
112
+ }
113
+ },
114
+ { "type": "null" }
115
+ ]
116
+ }
117
+ }
118
+ },
119
+ "conformance": {
120
+ "type": "object",
121
+ "additionalProperties": false,
122
+ "required": ["ok", "failures", "artifactValidationOk", "artifactValidationFailures"],
123
+ "properties": {
124
+ "ok": { "type": "boolean" },
125
+ "failures": {
126
+ "type": "array",
127
+ "items": { "type": "string" }
128
+ },
129
+ "artifactValidationOk": { "type": "boolean" },
130
+ "artifactValidationFailures": {
131
+ "type": "array",
132
+ "items": { "type": "string" }
133
+ }
134
+ }
135
+ },
136
+ "integrity": {
137
+ "type": "object",
138
+ "additionalProperties": false,
139
+ "required": ["artifactHashes", "externalRunMetadataHash"],
140
+ "properties": {
141
+ "artifactHashes": { "$ref": "#/$defs/artifactHashes" },
142
+ "externalRunMetadataHash": {
143
+ "anyOf": [
144
+ { "$ref": "#/$defs/sha256" },
145
+ { "type": "null" }
146
+ ]
147
+ }
148
+ }
149
+ },
150
+ "provenance": { "type": "object" }
151
+ },
152
+ "$defs": {
153
+ "artifactHashes": {
154
+ "type": "object",
155
+ "additionalProperties": false,
156
+ "required": [
157
+ "guardbench-manifest.json",
158
+ "guardbench-summary.json",
159
+ "guardbench-raw.json"
160
+ ],
161
+ "properties": {
162
+ "guardbench-manifest.json": { "$ref": "#/$defs/sha256" },
163
+ "guardbench-summary.json": { "$ref": "#/$defs/sha256" },
164
+ "guardbench-raw.json": { "$ref": "#/$defs/sha256" }
165
+ }
166
+ },
167
+ "nullableNumber": {
168
+ "anyOf": [
169
+ { "type": "number", "minimum": 0 },
170
+ { "type": "null" }
171
+ ]
172
+ },
173
+ "nullableRate": {
174
+ "anyOf": [
175
+ { "type": "number", "minimum": 0, "maximum": 1 },
176
+ { "type": "null" }
177
+ ]
178
+ },
179
+ "sha256": {
180
+ "type": "string",
181
+ "pattern": "^[a-f0-9]{64}$"
182
+ }
183
+ }
184
+ }
@@ -0,0 +1,74 @@
1
+ {
2
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
3
+ "$id": "https://audrey-memory.org/schemas/guardbench-external-dry-run.schema.json",
4
+ "title": "GuardBench External Adapter Dry-Run Matrix",
5
+ "type": "object",
6
+ "additionalProperties": false,
7
+ "required": ["schemaVersion", "suite", "generatedAt", "ok", "registry", "outRoot", "adapters", "failures"],
8
+ "properties": {
9
+ "schemaVersion": { "const": "1.0.0" },
10
+ "suite": { "const": "GuardBench external adapter dry-run matrix" },
11
+ "generatedAt": { "type": "string", "minLength": 1 },
12
+ "ok": { "type": "boolean" },
13
+ "registry": { "type": "string", "minLength": 1 },
14
+ "outRoot": { "type": "string", "minLength": 1 },
15
+ "adapters": {
16
+ "type": "array",
17
+ "minItems": 1,
18
+ "items": { "$ref": "#/$defs/adapterDryRun" }
19
+ },
20
+ "failures": {
21
+ "type": "array",
22
+ "items": { "type": "string" }
23
+ }
24
+ },
25
+ "$defs": {
26
+ "adapterDryRun": {
27
+ "type": "object",
28
+ "additionalProperties": false,
29
+ "required": [
30
+ "id",
31
+ "name",
32
+ "credentialMode",
33
+ "requiredEnv",
34
+ "missingEnv",
35
+ "status",
36
+ "command",
37
+ "validationCommand",
38
+ "metadataPath"
39
+ ],
40
+ "properties": {
41
+ "id": {
42
+ "type": "string",
43
+ "pattern": "^[a-z0-9][a-z0-9-]*$"
44
+ },
45
+ "name": { "type": "string", "minLength": 1 },
46
+ "credentialMode": {
47
+ "enum": ["none", "runtime-env"]
48
+ },
49
+ "requiredEnv": {
50
+ "type": "array",
51
+ "items": { "type": "string", "minLength": 1 }
52
+ },
53
+ "missingEnv": {
54
+ "type": "array",
55
+ "items": { "type": "string", "minLength": 1 }
56
+ },
57
+ "status": {
58
+ "enum": ["dry-run-missing-env", "dry-run-ready"]
59
+ },
60
+ "command": {
61
+ "type": "array",
62
+ "minItems": 1,
63
+ "items": { "type": "string", "minLength": 1 }
64
+ },
65
+ "validationCommand": {
66
+ "type": "array",
67
+ "minItems": 1,
68
+ "items": { "type": "string", "minLength": 1 }
69
+ },
70
+ "metadataPath": { "type": "string", "minLength": 1 }
71
+ }
72
+ }
73
+ }
74
+ }
@@ -0,0 +1,108 @@
1
+ {
2
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
3
+ "$id": "https://audrey-memory.org/schemas/guardbench-external-evidence.schema.json",
4
+ "title": "GuardBench External Evidence Verification",
5
+ "type": "object",
6
+ "additionalProperties": false,
7
+ "required": ["schemaVersion", "suite", "generatedAt", "ok", "allowPending", "registry", "outRoot", "adapters", "failures"],
8
+ "properties": {
9
+ "schemaVersion": { "const": "1.0.0" },
10
+ "suite": { "const": "GuardBench external evidence verification" },
11
+ "generatedAt": { "type": "string", "minLength": 1 },
12
+ "ok": { "type": "boolean" },
13
+ "allowPending": { "type": "boolean" },
14
+ "registry": { "type": "string", "minLength": 1 },
15
+ "outRoot": { "type": "string", "minLength": 1 },
16
+ "adapters": {
17
+ "type": "array",
18
+ "minItems": 1,
19
+ "items": { "$ref": "#/$defs/adapterEvidence" }
20
+ },
21
+ "failures": {
22
+ "type": "array",
23
+ "items": { "type": "string" }
24
+ }
25
+ },
26
+ "$defs": {
27
+ "adapterEvidence": {
28
+ "type": "object",
29
+ "additionalProperties": false,
30
+ "required": [
31
+ "id",
32
+ "name",
33
+ "path",
34
+ "credentialMode",
35
+ "requiredEnv",
36
+ "outDir",
37
+ "metadataPath",
38
+ "status",
39
+ "evidenceKind",
40
+ "metadataStatus",
41
+ "dryRun",
42
+ "missingEnv",
43
+ "artifactValidationOk",
44
+ "adapterConformanceOk",
45
+ "secretLeakCount",
46
+ "failures"
47
+ ],
48
+ "properties": {
49
+ "id": {
50
+ "type": "string",
51
+ "pattern": "^[a-z0-9][a-z0-9-]*$"
52
+ },
53
+ "name": { "type": "string", "minLength": 1 },
54
+ "path": { "type": "string", "minLength": 1 },
55
+ "credentialMode": { "const": "runtime-env" },
56
+ "requiredEnv": {
57
+ "type": "array",
58
+ "minItems": 1,
59
+ "items": { "type": "string", "minLength": 1 }
60
+ },
61
+ "outDir": { "type": "string", "minLength": 1 },
62
+ "metadataPath": { "type": "string", "minLength": 1 },
63
+ "status": {
64
+ "enum": ["failed", "pending", "verified"]
65
+ },
66
+ "evidenceKind": {
67
+ "enum": ["blocked", "dry-run", "live", "missing"]
68
+ },
69
+ "metadataStatus": {
70
+ "anyOf": [
71
+ { "type": "null" },
72
+ { "type": "string", "minLength": 1 }
73
+ ]
74
+ },
75
+ "dryRun": {
76
+ "anyOf": [
77
+ { "type": "boolean" },
78
+ { "type": "null" }
79
+ ]
80
+ },
81
+ "missingEnv": {
82
+ "type": "array",
83
+ "items": { "type": "string", "minLength": 1 }
84
+ },
85
+ "artifactValidationOk": {
86
+ "anyOf": [
87
+ { "type": "boolean" },
88
+ { "type": "null" }
89
+ ]
90
+ },
91
+ "adapterConformanceOk": {
92
+ "anyOf": [
93
+ { "type": "boolean" },
94
+ { "type": "null" }
95
+ ]
96
+ },
97
+ "secretLeakCount": {
98
+ "type": "integer",
99
+ "minimum": 0
100
+ },
101
+ "failures": {
102
+ "type": "array",
103
+ "items": { "type": "string" }
104
+ }
105
+ }
106
+ }
107
+ }
108
+ }