audrey 0.23.1 → 1.0.0
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.
- package/CHANGELOG.md +81 -19
- package/LICENSE +21 -21
- package/README.md +209 -5
- package/SECURITY.md +2 -1
- package/benchmarks/adapter-kit.mjs +20 -0
- package/benchmarks/adapter-self-test.mjs +166 -0
- package/benchmarks/adapters/example-allow.mjs +28 -0
- package/benchmarks/adapters/mem0-platform.mjs +267 -0
- package/benchmarks/adapters/registry.json +51 -0
- package/benchmarks/adapters/zep-cloud.mjs +280 -0
- package/benchmarks/baselines.js +169 -0
- package/benchmarks/build-leaderboard.mjs +170 -0
- package/benchmarks/cases.js +537 -0
- package/benchmarks/create-conformance-card.mjs +139 -0
- package/benchmarks/create-submission-bundle.mjs +176 -0
- package/benchmarks/dry-run-external-adapters.mjs +165 -0
- package/benchmarks/guardbench.js +1035 -0
- package/benchmarks/output/adapter-self-test/guardbench-adapter-self-test.json +50 -0
- package/benchmarks/output/external/guardbench-external-dry-run.json +69 -0
- package/benchmarks/output/external/guardbench-external-evidence.json +56 -0
- package/benchmarks/output/guardbench-conformance-card.json +63 -0
- package/benchmarks/output/guardbench-manifest.json +414 -0
- package/benchmarks/output/guardbench-raw.json +1171 -0
- package/benchmarks/output/guardbench-summary.json +1981 -0
- package/benchmarks/output/leaderboard/guardbench-leaderboard.json +93 -0
- package/benchmarks/output/leaderboard/guardbench-leaderboard.md +7 -0
- package/benchmarks/output/submission-bundle/guardbench-conformance-card.json +63 -0
- package/benchmarks/output/submission-bundle/guardbench-manifest.json +414 -0
- package/benchmarks/output/submission-bundle/guardbench-raw.json +1171 -0
- package/benchmarks/output/submission-bundle/guardbench-summary.json +1981 -0
- package/benchmarks/output/submission-bundle/schemas/guardbench-adapter-registry.schema.json +69 -0
- package/benchmarks/output/submission-bundle/schemas/guardbench-adapter-self-test.schema.json +156 -0
- package/benchmarks/output/submission-bundle/schemas/guardbench-conformance-card.schema.json +184 -0
- package/benchmarks/output/submission-bundle/schemas/guardbench-external-dry-run.schema.json +74 -0
- package/benchmarks/output/submission-bundle/schemas/guardbench-external-evidence.schema.json +108 -0
- package/benchmarks/output/submission-bundle/schemas/guardbench-external-run.schema.json +160 -0
- package/benchmarks/output/submission-bundle/schemas/guardbench-leaderboard.schema.json +179 -0
- package/benchmarks/output/submission-bundle/schemas/guardbench-manifest.schema.json +213 -0
- package/benchmarks/output/submission-bundle/schemas/guardbench-publication-verification.schema.json +47 -0
- package/benchmarks/output/submission-bundle/schemas/guardbench-raw.schema.json +164 -0
- package/benchmarks/output/submission-bundle/schemas/guardbench-submission-manifest.schema.json +151 -0
- package/benchmarks/output/submission-bundle/schemas/guardbench-summary.schema.json +228 -0
- package/benchmarks/output/submission-bundle/submission-manifest.json +131 -0
- package/benchmarks/output/submission-bundle/validation-report.json +31 -0
- package/benchmarks/output/summary.json +2354 -0
- package/benchmarks/perf-snapshot.js +304 -0
- package/benchmarks/perf.bench.js +161 -0
- package/benchmarks/public-paths.mjs +78 -0
- package/benchmarks/reference-results.js +70 -0
- package/benchmarks/report.js +259 -0
- package/benchmarks/run-external-guardbench.mjs +281 -0
- package/benchmarks/run.js +682 -0
- package/benchmarks/schemas/guardbench-adapter-registry.schema.json +69 -0
- package/benchmarks/schemas/guardbench-adapter-self-test.schema.json +156 -0
- package/benchmarks/schemas/guardbench-conformance-card.schema.json +184 -0
- package/benchmarks/schemas/guardbench-external-dry-run.schema.json +74 -0
- package/benchmarks/schemas/guardbench-external-evidence.schema.json +108 -0
- package/benchmarks/schemas/guardbench-external-run.schema.json +160 -0
- package/benchmarks/schemas/guardbench-leaderboard.schema.json +179 -0
- package/benchmarks/schemas/guardbench-manifest.schema.json +213 -0
- package/benchmarks/schemas/guardbench-publication-verification.schema.json +47 -0
- package/benchmarks/schemas/guardbench-raw.schema.json +164 -0
- package/benchmarks/schemas/guardbench-submission-manifest.schema.json +151 -0
- package/benchmarks/schemas/guardbench-summary.schema.json +228 -0
- package/benchmarks/snapshots/perf-0.22.2.json +123 -0
- package/benchmarks/snapshots/perf-0.23.0.json +123 -0
- package/benchmarks/validate-adapter-module.mjs +104 -0
- package/benchmarks/validate-adapter-registry.mjs +134 -0
- package/benchmarks/validate-adapter-self-test.mjs +96 -0
- package/benchmarks/validate-guardbench-artifacts.mjs +343 -0
- package/benchmarks/verify-external-evidence.mjs +296 -0
- package/benchmarks/verify-publication-artifacts.mjs +286 -0
- package/benchmarks/verify-submission-bundle.mjs +167 -0
- package/dist/mcp-server/config.d.ts +1 -1
- package/dist/mcp-server/config.d.ts.map +1 -1
- package/dist/mcp-server/config.js +1 -1
- package/dist/mcp-server/config.js.map +1 -1
- package/dist/mcp-server/index.d.ts +65 -3
- package/dist/mcp-server/index.d.ts.map +1 -1
- package/dist/mcp-server/index.js +675 -157
- package/dist/mcp-server/index.js.map +1 -1
- package/dist/src/action-key.d.ts +9 -0
- package/dist/src/action-key.d.ts.map +1 -0
- package/dist/src/action-key.js +49 -0
- package/dist/src/action-key.js.map +1 -0
- package/dist/src/adaptive.js +5 -5
- package/dist/src/affect.js +8 -8
- package/dist/src/audrey.d.ts +3 -0
- package/dist/src/audrey.d.ts.map +1 -1
- package/dist/src/audrey.js +55 -3
- package/dist/src/audrey.js.map +1 -1
- package/dist/src/capsule.js +4 -4
- package/dist/src/causal.js +3 -3
- package/dist/src/consolidate.js +48 -48
- package/dist/src/controller.d.ts +61 -5
- package/dist/src/controller.d.ts.map +1 -1
- package/dist/src/controller.js +230 -49
- package/dist/src/controller.js.map +1 -1
- package/dist/src/db.js +172 -172
- package/dist/src/decay.js +8 -8
- package/dist/src/embedding.d.ts +2 -1
- package/dist/src/embedding.d.ts.map +1 -1
- package/dist/src/embedding.js +39 -29
- package/dist/src/embedding.js.map +1 -1
- package/dist/src/encode.js +6 -6
- package/dist/src/feedback.d.ts +6 -0
- package/dist/src/feedback.d.ts.map +1 -1
- package/dist/src/feedback.js +6 -0
- package/dist/src/feedback.js.map +1 -1
- package/dist/src/forget.js +12 -12
- package/dist/src/hybrid-recall.js +9 -9
- package/dist/src/impact.js +6 -6
- package/dist/src/import.d.ts +3 -3
- package/dist/src/import.js +41 -41
- package/dist/src/index.d.ts +3 -3
- package/dist/src/index.d.ts.map +1 -1
- package/dist/src/index.js +2 -2
- package/dist/src/index.js.map +1 -1
- package/dist/src/interference.js +14 -14
- package/dist/src/introspect.js +18 -18
- package/dist/src/preflight.d.ts.map +1 -1
- package/dist/src/preflight.js +41 -0
- package/dist/src/preflight.js.map +1 -1
- package/dist/src/promote.js +7 -7
- package/dist/src/prompts.js +118 -118
- package/dist/src/recall.js +30 -30
- package/dist/src/reflexes.d.ts +1 -0
- package/dist/src/reflexes.d.ts.map +1 -1
- package/dist/src/reflexes.js +3 -0
- package/dist/src/reflexes.js.map +1 -1
- package/dist/src/rollback.js +4 -4
- package/dist/src/routes.d.ts.map +1 -1
- package/dist/src/routes.js +67 -1
- package/dist/src/routes.js.map +1 -1
- package/dist/src/validate.js +25 -25
- package/docs/AUDREY_PAPER_OUTLINE.md +175 -0
- package/docs/MEMORY_BENCHMARKING.md +59 -0
- package/docs/PRODUCTION_BACKLOG.md +304 -0
- package/docs/paper/00-master.md +48 -0
- package/docs/paper/01-introduction.md +27 -0
- package/docs/paper/02-related-work.md +47 -0
- package/docs/paper/03-problem-definition.md +108 -0
- package/docs/paper/04-design.md +164 -0
- package/docs/paper/05-guardbench-spec.md +412 -0
- package/docs/paper/06-implementation.md +113 -0
- package/docs/paper/07-evaluation.md +168 -0
- package/docs/paper/08-discussion-limitations.md +61 -0
- package/docs/paper/09-conclusion.md +11 -0
- package/docs/paper/SUBMISSION_README.md +162 -0
- package/docs/paper/appendix-a-demo-transcript.md +114 -0
- package/docs/paper/arxiv-compile-report.schema.json +116 -0
- package/docs/paper/arxiv-source.schema.json +61 -0
- package/docs/paper/audrey-paper-v1.md +1106 -0
- package/docs/paper/browser-launch-plan.json +209 -0
- package/docs/paper/browser-launch-plan.schema.json +100 -0
- package/docs/paper/browser-launch-results.json +86 -0
- package/docs/paper/browser-launch-results.schema.json +66 -0
- package/docs/paper/claim-register.json +138 -0
- package/docs/paper/claim-register.schema.json +81 -0
- package/docs/paper/evidence-ledger.md +103 -0
- package/docs/paper/output/arxiv/README-arxiv.txt +8 -0
- package/docs/paper/output/arxiv/arxiv-manifest.json +41 -0
- package/docs/paper/output/arxiv/main.tex +949 -0
- package/docs/paper/output/arxiv/references.bib +222 -0
- package/docs/paper/output/arxiv-compile-report.json +24 -0
- package/docs/paper/output/submission-bundle/LICENSE +21 -0
- package/docs/paper/output/submission-bundle/README.md +533 -0
- package/docs/paper/output/submission-bundle/benchmarks/output/adapter-self-test/guardbench-adapter-self-test.json +50 -0
- package/docs/paper/output/submission-bundle/benchmarks/output/external/guardbench-external-dry-run.json +69 -0
- package/docs/paper/output/submission-bundle/benchmarks/output/external/guardbench-external-evidence.json +56 -0
- package/docs/paper/output/submission-bundle/benchmarks/output/guardbench-conformance-card.json +63 -0
- package/docs/paper/output/submission-bundle/benchmarks/output/guardbench-manifest.json +414 -0
- package/docs/paper/output/submission-bundle/benchmarks/output/guardbench-raw.json +1171 -0
- package/docs/paper/output/submission-bundle/benchmarks/output/guardbench-summary.json +1981 -0
- package/docs/paper/output/submission-bundle/benchmarks/output/leaderboard/guardbench-leaderboard.json +93 -0
- package/docs/paper/output/submission-bundle/benchmarks/output/leaderboard/guardbench-leaderboard.md +7 -0
- package/docs/paper/output/submission-bundle/benchmarks/output/submission-bundle/submission-manifest.json +131 -0
- package/docs/paper/output/submission-bundle/benchmarks/output/submission-bundle/validation-report.json +31 -0
- package/docs/paper/output/submission-bundle/benchmarks/output/summary.json +2354 -0
- package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-adapter-registry.schema.json +69 -0
- package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-adapter-self-test.schema.json +156 -0
- package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-conformance-card.schema.json +184 -0
- package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-external-dry-run.schema.json +74 -0
- package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-external-evidence.schema.json +108 -0
- package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-external-run.schema.json +160 -0
- package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-leaderboard.schema.json +179 -0
- package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-manifest.schema.json +213 -0
- package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-publication-verification.schema.json +47 -0
- package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-raw.schema.json +164 -0
- package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-submission-manifest.schema.json +151 -0
- package/docs/paper/output/submission-bundle/benchmarks/schemas/guardbench-summary.schema.json +228 -0
- package/docs/paper/output/submission-bundle/docs/AUDREY_PAPER_OUTLINE.md +175 -0
- package/docs/paper/output/submission-bundle/docs/paper/00-master.md +48 -0
- package/docs/paper/output/submission-bundle/docs/paper/01-introduction.md +27 -0
- package/docs/paper/output/submission-bundle/docs/paper/02-related-work.md +47 -0
- package/docs/paper/output/submission-bundle/docs/paper/03-problem-definition.md +108 -0
- package/docs/paper/output/submission-bundle/docs/paper/04-design.md +164 -0
- package/docs/paper/output/submission-bundle/docs/paper/05-guardbench-spec.md +412 -0
- package/docs/paper/output/submission-bundle/docs/paper/06-implementation.md +113 -0
- package/docs/paper/output/submission-bundle/docs/paper/07-evaluation.md +168 -0
- package/docs/paper/output/submission-bundle/docs/paper/08-discussion-limitations.md +61 -0
- package/docs/paper/output/submission-bundle/docs/paper/09-conclusion.md +11 -0
- package/docs/paper/output/submission-bundle/docs/paper/SUBMISSION_README.md +162 -0
- package/docs/paper/output/submission-bundle/docs/paper/appendix-a-demo-transcript.md +114 -0
- package/docs/paper/output/submission-bundle/docs/paper/arxiv-compile-report.schema.json +116 -0
- package/docs/paper/output/submission-bundle/docs/paper/arxiv-source.schema.json +61 -0
- package/docs/paper/output/submission-bundle/docs/paper/audrey-paper-v1.md +1106 -0
- package/docs/paper/output/submission-bundle/docs/paper/browser-launch-plan.json +209 -0
- package/docs/paper/output/submission-bundle/docs/paper/browser-launch-plan.schema.json +100 -0
- package/docs/paper/output/submission-bundle/docs/paper/browser-launch-results.json +86 -0
- package/docs/paper/output/submission-bundle/docs/paper/browser-launch-results.schema.json +66 -0
- package/docs/paper/output/submission-bundle/docs/paper/claim-register.json +138 -0
- package/docs/paper/output/submission-bundle/docs/paper/claim-register.schema.json +81 -0
- package/docs/paper/output/submission-bundle/docs/paper/evidence-ledger.md +103 -0
- package/docs/paper/output/submission-bundle/docs/paper/output/arxiv/README-arxiv.txt +8 -0
- package/docs/paper/output/submission-bundle/docs/paper/output/arxiv/arxiv-manifest.json +41 -0
- package/docs/paper/output/submission-bundle/docs/paper/output/arxiv/main.tex +949 -0
- package/docs/paper/output/submission-bundle/docs/paper/output/arxiv/references.bib +222 -0
- package/docs/paper/output/submission-bundle/docs/paper/output/arxiv-compile-report.json +24 -0
- package/docs/paper/output/submission-bundle/docs/paper/paper-submission-bundle.schema.json +70 -0
- package/docs/paper/output/submission-bundle/docs/paper/publication-pack.json +81 -0
- package/docs/paper/output/submission-bundle/docs/paper/publication-pack.schema.json +60 -0
- package/docs/paper/output/submission-bundle/docs/paper/references.bib +222 -0
- package/docs/paper/output/submission-bundle/package.json +212 -0
- package/docs/paper/output/submission-bundle/paper-submission-manifest.json +379 -0
- package/docs/paper/paper-submission-bundle.schema.json +70 -0
- package/docs/paper/publication-pack.json +81 -0
- package/docs/paper/publication-pack.schema.json +60 -0
- package/docs/paper/references.bib +222 -0
- package/package.json +87 -4
- package/scripts/audit-release-completion.mjs +362 -0
- package/scripts/create-arxiv-source.mjs +362 -0
- package/scripts/create-paper-submission-bundle.mjs +210 -0
- package/scripts/finalize-release.mjs +526 -0
- package/scripts/prepare-release-cut.mjs +269 -0
- package/scripts/publish-release-bundle.mjs +209 -0
- package/scripts/publish-release-github-api.mjs +429 -0
- package/scripts/run-vitest.mjs +34 -0
- package/scripts/smoke-cli.js +72 -0
- package/scripts/sync-paper-artifacts.mjs +109 -0
- package/scripts/verify-arxiv-compile.mjs +440 -0
- package/scripts/verify-arxiv-source.mjs +194 -0
- package/scripts/verify-browser-launch-plan.mjs +237 -0
- package/scripts/verify-browser-launch-results.mjs +285 -0
- package/scripts/verify-paper-artifacts.mjs +338 -0
- package/scripts/verify-paper-claims.mjs +226 -0
- package/scripts/verify-paper-submission-bundle.mjs +207 -0
- package/scripts/verify-publication-pack.mjs +196 -0
- package/scripts/verify-python-package.py +201 -0
- package/scripts/verify-release-readiness.mjs +741 -0
|
@@ -0,0 +1,164 @@
|
|
|
1
|
+
{
|
|
2
|
+
"$schema": "https://json-schema.org/draft/2020-12/schema",
|
|
3
|
+
"$id": "https://audrey-memory.org/schemas/guardbench-raw.schema.json",
|
|
4
|
+
"title": "GuardBench Raw Output",
|
|
5
|
+
"type": "object",
|
|
6
|
+
"additionalProperties": false,
|
|
7
|
+
"required": [
|
|
8
|
+
"suite",
|
|
9
|
+
"generatedAt",
|
|
10
|
+
"manifestVersion",
|
|
11
|
+
"provenance",
|
|
12
|
+
"cases",
|
|
13
|
+
"artifactRedactionSweep"
|
|
14
|
+
],
|
|
15
|
+
"properties": {
|
|
16
|
+
"suite": { "const": "GuardBench comparative" },
|
|
17
|
+
"generatedAt": { "type": "string", "minLength": 1 },
|
|
18
|
+
"manifestVersion": { "type": "string", "minLength": 1 },
|
|
19
|
+
"provenance": { "$ref": "#/$defs/provenance" },
|
|
20
|
+
"cases": {
|
|
21
|
+
"type": "array",
|
|
22
|
+
"minItems": 10,
|
|
23
|
+
"items": { "$ref": "#/$defs/caseResult" }
|
|
24
|
+
},
|
|
25
|
+
"artifactRedactionSweep": { "$ref": "#/$defs/artifactRedactionSweep" }
|
|
26
|
+
},
|
|
27
|
+
"$defs": {
|
|
28
|
+
"provenance": {
|
|
29
|
+
"type": "object",
|
|
30
|
+
"additionalProperties": false,
|
|
31
|
+
"required": [
|
|
32
|
+
"generatedAt",
|
|
33
|
+
"gitSha",
|
|
34
|
+
"gitDirty",
|
|
35
|
+
"node",
|
|
36
|
+
"v8",
|
|
37
|
+
"platform",
|
|
38
|
+
"arch",
|
|
39
|
+
"osRelease",
|
|
40
|
+
"cpuModel",
|
|
41
|
+
"cpuCount",
|
|
42
|
+
"totalMemoryGb",
|
|
43
|
+
"embeddingProvider",
|
|
44
|
+
"embeddingDimensions",
|
|
45
|
+
"llmProvider"
|
|
46
|
+
],
|
|
47
|
+
"properties": {
|
|
48
|
+
"generatedAt": { "type": "string", "minLength": 1 },
|
|
49
|
+
"gitSha": {
|
|
50
|
+
"anyOf": [
|
|
51
|
+
{ "type": "null" },
|
|
52
|
+
{ "type": "string", "minLength": 7 }
|
|
53
|
+
]
|
|
54
|
+
},
|
|
55
|
+
"gitDirty": { "type": "boolean" },
|
|
56
|
+
"node": { "type": "string", "minLength": 1 },
|
|
57
|
+
"v8": { "type": "string", "minLength": 1 },
|
|
58
|
+
"platform": { "type": "string", "minLength": 1 },
|
|
59
|
+
"arch": { "type": "string", "minLength": 1 },
|
|
60
|
+
"osRelease": { "type": "string", "minLength": 1 },
|
|
61
|
+
"cpuModel": {
|
|
62
|
+
"anyOf": [
|
|
63
|
+
{ "type": "null" },
|
|
64
|
+
{ "type": "string", "minLength": 1 }
|
|
65
|
+
]
|
|
66
|
+
},
|
|
67
|
+
"cpuCount": { "type": "integer", "minimum": 1 },
|
|
68
|
+
"totalMemoryGb": { "type": "number", "minimum": 0 },
|
|
69
|
+
"embeddingProvider": { "type": "string", "minLength": 1 },
|
|
70
|
+
"embeddingDimensions": { "type": "integer", "minimum": 1 },
|
|
71
|
+
"llmProvider": { "type": "string", "minLength": 1 }
|
|
72
|
+
}
|
|
73
|
+
},
|
|
74
|
+
"caseResult": {
|
|
75
|
+
"type": "object",
|
|
76
|
+
"additionalProperties": false,
|
|
77
|
+
"required": ["id", "name", "expectedDecision", "results"],
|
|
78
|
+
"properties": {
|
|
79
|
+
"id": { "type": "string", "pattern": "^GB-[0-9]{2}$" },
|
|
80
|
+
"name": { "type": "string", "minLength": 1 },
|
|
81
|
+
"expectedDecision": { "enum": ["allow", "warn", "block"] },
|
|
82
|
+
"results": {
|
|
83
|
+
"type": "array",
|
|
84
|
+
"minItems": 5,
|
|
85
|
+
"items": { "$ref": "#/$defs/resultRow" }
|
|
86
|
+
}
|
|
87
|
+
}
|
|
88
|
+
},
|
|
89
|
+
"resultRow": {
|
|
90
|
+
"type": "object",
|
|
91
|
+
"additionalProperties": false,
|
|
92
|
+
"required": [
|
|
93
|
+
"system",
|
|
94
|
+
"id",
|
|
95
|
+
"name",
|
|
96
|
+
"expectedDecision",
|
|
97
|
+
"decision",
|
|
98
|
+
"decisionCorrect",
|
|
99
|
+
"riskScore",
|
|
100
|
+
"passed",
|
|
101
|
+
"latencyMs",
|
|
102
|
+
"evidenceCount",
|
|
103
|
+
"evidenceIds",
|
|
104
|
+
"recommendedActions",
|
|
105
|
+
"summary",
|
|
106
|
+
"recallErrors",
|
|
107
|
+
"leakedSecrets",
|
|
108
|
+
"requiredEvidenceMatched"
|
|
109
|
+
],
|
|
110
|
+
"properties": {
|
|
111
|
+
"system": { "type": "string", "minLength": 1 },
|
|
112
|
+
"external": { "type": "boolean" },
|
|
113
|
+
"id": { "type": "string", "pattern": "^GB-[0-9]{2}$" },
|
|
114
|
+
"name": { "type": "string", "minLength": 1 },
|
|
115
|
+
"expectedDecision": { "enum": ["allow", "warn", "block"] },
|
|
116
|
+
"decision": { "enum": ["allow", "warn", "block"] },
|
|
117
|
+
"decisionCorrect": { "type": "boolean" },
|
|
118
|
+
"riskScore": { "type": "number", "minimum": 0, "maximum": 1 },
|
|
119
|
+
"passed": { "type": "boolean" },
|
|
120
|
+
"latencyMs": { "type": "number", "minimum": 0 },
|
|
121
|
+
"evidenceCount": { "type": "integer", "minimum": 0 },
|
|
122
|
+
"evidenceIds": {
|
|
123
|
+
"type": "array",
|
|
124
|
+
"items": { "type": "string" }
|
|
125
|
+
},
|
|
126
|
+
"recommendedActions": {
|
|
127
|
+
"type": "array",
|
|
128
|
+
"items": { "type": "string" }
|
|
129
|
+
},
|
|
130
|
+
"summary": { "type": "string", "minLength": 1 },
|
|
131
|
+
"recallErrors": { "type": "array" },
|
|
132
|
+
"leakedSecrets": {
|
|
133
|
+
"type": "array",
|
|
134
|
+
"items": { "type": "string" }
|
|
135
|
+
},
|
|
136
|
+
"requiredEvidenceMatched": { "type": "boolean" }
|
|
137
|
+
}
|
|
138
|
+
},
|
|
139
|
+
"artifactRedactionSweep": {
|
|
140
|
+
"type": "object",
|
|
141
|
+
"additionalProperties": false,
|
|
142
|
+
"required": [
|
|
143
|
+
"checkedAt",
|
|
144
|
+
"filesChecked",
|
|
145
|
+
"seededSecretsChecked",
|
|
146
|
+
"leakCount",
|
|
147
|
+
"leaks",
|
|
148
|
+
"passed"
|
|
149
|
+
],
|
|
150
|
+
"properties": {
|
|
151
|
+
"checkedAt": { "type": "string", "minLength": 1 },
|
|
152
|
+
"filesChecked": {
|
|
153
|
+
"type": "array",
|
|
154
|
+
"minItems": 3,
|
|
155
|
+
"items": { "type": "string", "minLength": 1 }
|
|
156
|
+
},
|
|
157
|
+
"seededSecretsChecked": { "type": "integer", "minimum": 0 },
|
|
158
|
+
"leakCount": { "type": "integer", "minimum": 0 },
|
|
159
|
+
"leaks": { "type": "array" },
|
|
160
|
+
"passed": { "type": "boolean" }
|
|
161
|
+
}
|
|
162
|
+
}
|
|
163
|
+
}
|
|
164
|
+
}
|
|
@@ -0,0 +1,151 @@
|
|
|
1
|
+
{
|
|
2
|
+
"$schema": "https://json-schema.org/draft/2020-12/schema",
|
|
3
|
+
"$id": "https://audrey-memory.org/schemas/guardbench-submission-manifest.schema.json",
|
|
4
|
+
"title": "GuardBench Submission Manifest",
|
|
5
|
+
"type": "object",
|
|
6
|
+
"additionalProperties": false,
|
|
7
|
+
"required": [
|
|
8
|
+
"schemaVersion",
|
|
9
|
+
"suite",
|
|
10
|
+
"generatedAt",
|
|
11
|
+
"sourceDir",
|
|
12
|
+
"subject",
|
|
13
|
+
"score",
|
|
14
|
+
"conformance",
|
|
15
|
+
"validation",
|
|
16
|
+
"files"
|
|
17
|
+
],
|
|
18
|
+
"properties": {
|
|
19
|
+
"schemaVersion": { "const": "1.0.0" },
|
|
20
|
+
"suite": { "const": "GuardBench submission bundle" },
|
|
21
|
+
"generatedAt": { "type": "string", "minLength": 1 },
|
|
22
|
+
"sourceDir": { "type": "string", "minLength": 1 },
|
|
23
|
+
"subject": { "$ref": "#/$defs/subject" },
|
|
24
|
+
"score": { "$ref": "#/$defs/score" },
|
|
25
|
+
"conformance": { "$ref": "#/$defs/conformance" },
|
|
26
|
+
"validation": { "$ref": "#/$defs/validation" },
|
|
27
|
+
"files": {
|
|
28
|
+
"type": "array",
|
|
29
|
+
"minItems": 1,
|
|
30
|
+
"items": { "$ref": "#/$defs/fileRecord" }
|
|
31
|
+
}
|
|
32
|
+
},
|
|
33
|
+
"$defs": {
|
|
34
|
+
"subject": {
|
|
35
|
+
"type": "object",
|
|
36
|
+
"additionalProperties": false,
|
|
37
|
+
"required": ["name", "requestedAdapter", "external"],
|
|
38
|
+
"properties": {
|
|
39
|
+
"name": { "type": "string", "minLength": 1 },
|
|
40
|
+
"requestedAdapter": {
|
|
41
|
+
"anyOf": [
|
|
42
|
+
{ "type": "string", "minLength": 1 },
|
|
43
|
+
{ "type": "null" }
|
|
44
|
+
]
|
|
45
|
+
},
|
|
46
|
+
"external": { "type": "boolean" }
|
|
47
|
+
}
|
|
48
|
+
},
|
|
49
|
+
"score": {
|
|
50
|
+
"type": "object",
|
|
51
|
+
"additionalProperties": false,
|
|
52
|
+
"required": [
|
|
53
|
+
"scenarios",
|
|
54
|
+
"fullContractPassed",
|
|
55
|
+
"fullContractPassRate",
|
|
56
|
+
"decisionAccuracy",
|
|
57
|
+
"evidenceRecall",
|
|
58
|
+
"redactionLeaks",
|
|
59
|
+
"latency"
|
|
60
|
+
],
|
|
61
|
+
"properties": {
|
|
62
|
+
"scenarios": { "type": "integer", "minimum": 0 },
|
|
63
|
+
"fullContractPassed": { "$ref": "#/$defs/nullableNumber" },
|
|
64
|
+
"fullContractPassRate": { "$ref": "#/$defs/nullableRate" },
|
|
65
|
+
"decisionAccuracy": { "$ref": "#/$defs/nullableRate" },
|
|
66
|
+
"evidenceRecall": { "$ref": "#/$defs/nullableRate" },
|
|
67
|
+
"redactionLeaks": { "$ref": "#/$defs/nullableNumber" },
|
|
68
|
+
"latency": {
|
|
69
|
+
"anyOf": [
|
|
70
|
+
{
|
|
71
|
+
"type": "object",
|
|
72
|
+
"additionalProperties": false,
|
|
73
|
+
"required": ["p50Ms", "p95Ms", "maxMs"],
|
|
74
|
+
"properties": {
|
|
75
|
+
"p50Ms": { "type": "number", "minimum": 0 },
|
|
76
|
+
"p95Ms": { "type": "number", "minimum": 0 },
|
|
77
|
+
"maxMs": { "type": "number", "minimum": 0 }
|
|
78
|
+
}
|
|
79
|
+
},
|
|
80
|
+
{ "type": "null" }
|
|
81
|
+
]
|
|
82
|
+
}
|
|
83
|
+
}
|
|
84
|
+
},
|
|
85
|
+
"conformance": {
|
|
86
|
+
"type": "object",
|
|
87
|
+
"additionalProperties": false,
|
|
88
|
+
"required": ["ok", "failures", "artifactValidationOk", "artifactValidationFailures"],
|
|
89
|
+
"properties": {
|
|
90
|
+
"ok": { "type": "boolean" },
|
|
91
|
+
"failures": {
|
|
92
|
+
"type": "array",
|
|
93
|
+
"items": { "type": "string" }
|
|
94
|
+
},
|
|
95
|
+
"artifactValidationOk": { "type": "boolean" },
|
|
96
|
+
"artifactValidationFailures": {
|
|
97
|
+
"type": "array",
|
|
98
|
+
"items": { "type": "string" }
|
|
99
|
+
}
|
|
100
|
+
}
|
|
101
|
+
},
|
|
102
|
+
"validation": {
|
|
103
|
+
"type": "object",
|
|
104
|
+
"additionalProperties": false,
|
|
105
|
+
"required": ["ok", "dir", "schemasDir", "files", "optionalFiles", "failures"],
|
|
106
|
+
"properties": {
|
|
107
|
+
"ok": { "type": "boolean" },
|
|
108
|
+
"dir": { "type": "string", "minLength": 1 },
|
|
109
|
+
"schemasDir": { "type": "string", "minLength": 1 },
|
|
110
|
+
"files": {
|
|
111
|
+
"type": "array",
|
|
112
|
+
"items": { "type": "string", "minLength": 1 }
|
|
113
|
+
},
|
|
114
|
+
"optionalFiles": {
|
|
115
|
+
"type": "array",
|
|
116
|
+
"items": { "type": "string", "minLength": 1 }
|
|
117
|
+
},
|
|
118
|
+
"failures": {
|
|
119
|
+
"type": "array",
|
|
120
|
+
"items": { "type": "string" }
|
|
121
|
+
}
|
|
122
|
+
}
|
|
123
|
+
},
|
|
124
|
+
"fileRecord": {
|
|
125
|
+
"type": "object",
|
|
126
|
+
"additionalProperties": false,
|
|
127
|
+
"required": ["path", "bytes", "sha256"],
|
|
128
|
+
"properties": {
|
|
129
|
+
"path": { "type": "string", "minLength": 1 },
|
|
130
|
+
"bytes": { "type": "integer", "minimum": 0 },
|
|
131
|
+
"sha256": { "$ref": "#/$defs/sha256" }
|
|
132
|
+
}
|
|
133
|
+
},
|
|
134
|
+
"nullableNumber": {
|
|
135
|
+
"anyOf": [
|
|
136
|
+
{ "type": "number", "minimum": 0 },
|
|
137
|
+
{ "type": "null" }
|
|
138
|
+
]
|
|
139
|
+
},
|
|
140
|
+
"nullableRate": {
|
|
141
|
+
"anyOf": [
|
|
142
|
+
{ "type": "number", "minimum": 0, "maximum": 1 },
|
|
143
|
+
{ "type": "null" }
|
|
144
|
+
]
|
|
145
|
+
},
|
|
146
|
+
"sha256": {
|
|
147
|
+
"type": "string",
|
|
148
|
+
"pattern": "^[a-f0-9]{64}$"
|
|
149
|
+
}
|
|
150
|
+
}
|
|
151
|
+
}
|
|
@@ -0,0 +1,228 @@
|
|
|
1
|
+
{
|
|
2
|
+
"$schema": "https://json-schema.org/draft/2020-12/schema",
|
|
3
|
+
"$id": "https://audrey-memory.org/schemas/guardbench-summary.schema.json",
|
|
4
|
+
"title": "GuardBench Summary",
|
|
5
|
+
"type": "object",
|
|
6
|
+
"additionalProperties": false,
|
|
7
|
+
"required": [
|
|
8
|
+
"suite",
|
|
9
|
+
"generatedAt",
|
|
10
|
+
"manifest",
|
|
11
|
+
"provenance",
|
|
12
|
+
"subjects",
|
|
13
|
+
"scenarios",
|
|
14
|
+
"passed",
|
|
15
|
+
"passRate",
|
|
16
|
+
"preventionRate",
|
|
17
|
+
"falseBlockRate",
|
|
18
|
+
"decisionAccuracy",
|
|
19
|
+
"evidenceRecall",
|
|
20
|
+
"redactionLeaks",
|
|
21
|
+
"recallDegradationDetectionRate",
|
|
22
|
+
"latency",
|
|
23
|
+
"systemSummaries",
|
|
24
|
+
"comparisons",
|
|
25
|
+
"rows",
|
|
26
|
+
"cases",
|
|
27
|
+
"artifactRedactionSweep"
|
|
28
|
+
],
|
|
29
|
+
"properties": {
|
|
30
|
+
"suite": { "const": "GuardBench comparative" },
|
|
31
|
+
"generatedAt": { "type": "string", "minLength": 1 },
|
|
32
|
+
"manifest": { "type": "object" },
|
|
33
|
+
"provenance": { "type": "object" },
|
|
34
|
+
"subjects": {
|
|
35
|
+
"type": "array",
|
|
36
|
+
"minItems": 5,
|
|
37
|
+
"items": { "type": "string", "minLength": 1 }
|
|
38
|
+
},
|
|
39
|
+
"scenarios": { "type": "integer", "minimum": 10 },
|
|
40
|
+
"passed": { "type": "integer", "minimum": 0 },
|
|
41
|
+
"passRate": { "type": "number", "minimum": 0, "maximum": 1 },
|
|
42
|
+
"preventionRate": { "type": "number", "minimum": 0, "maximum": 1 },
|
|
43
|
+
"falseBlockRate": { "type": "number", "minimum": 0, "maximum": 1 },
|
|
44
|
+
"decisionAccuracy": { "type": "number", "minimum": 0, "maximum": 1 },
|
|
45
|
+
"usefulWarningPrecision": {
|
|
46
|
+
"anyOf": [
|
|
47
|
+
{ "type": "null" },
|
|
48
|
+
{ "type": "number", "minimum": 0, "maximum": 1 }
|
|
49
|
+
]
|
|
50
|
+
},
|
|
51
|
+
"evidenceRecall": { "type": "number", "minimum": 0, "maximum": 1 },
|
|
52
|
+
"redactionLeaks": { "type": "integer", "minimum": 0 },
|
|
53
|
+
"recallDegradationDetectionRate": { "type": "number", "minimum": 0, "maximum": 1 },
|
|
54
|
+
"latency": { "$ref": "#/$defs/latency" },
|
|
55
|
+
"systemSummaries": {
|
|
56
|
+
"type": "array",
|
|
57
|
+
"minItems": 5,
|
|
58
|
+
"items": { "$ref": "#/$defs/systemSummary" }
|
|
59
|
+
},
|
|
60
|
+
"comparisons": {
|
|
61
|
+
"type": "object",
|
|
62
|
+
"additionalProperties": false,
|
|
63
|
+
"required": ["bestBaseline", "audreyMarginOverBestBaseline"],
|
|
64
|
+
"properties": {
|
|
65
|
+
"bestBaseline": { "$ref": "#/$defs/systemSummary" },
|
|
66
|
+
"audreyMarginOverBestBaseline": {
|
|
67
|
+
"anyOf": [
|
|
68
|
+
{ "type": "null" },
|
|
69
|
+
{ "type": "number", "minimum": 0, "maximum": 1 }
|
|
70
|
+
]
|
|
71
|
+
}
|
|
72
|
+
}
|
|
73
|
+
},
|
|
74
|
+
"rows": {
|
|
75
|
+
"type": "array",
|
|
76
|
+
"minItems": 10,
|
|
77
|
+
"items": { "$ref": "#/$defs/resultRow" }
|
|
78
|
+
},
|
|
79
|
+
"cases": {
|
|
80
|
+
"type": "array",
|
|
81
|
+
"minItems": 10,
|
|
82
|
+
"items": { "$ref": "#/$defs/caseResult" }
|
|
83
|
+
},
|
|
84
|
+
"artifactRedactionSweep": { "$ref": "#/$defs/artifactRedactionSweep" }
|
|
85
|
+
},
|
|
86
|
+
"$defs": {
|
|
87
|
+
"latency": {
|
|
88
|
+
"type": "object",
|
|
89
|
+
"additionalProperties": false,
|
|
90
|
+
"required": ["p50Ms", "p95Ms", "maxMs"],
|
|
91
|
+
"properties": {
|
|
92
|
+
"p50Ms": { "type": "number", "minimum": 0 },
|
|
93
|
+
"p95Ms": { "type": "number", "minimum": 0 },
|
|
94
|
+
"maxMs": { "type": "number", "minimum": 0 }
|
|
95
|
+
}
|
|
96
|
+
},
|
|
97
|
+
"systemSummary": {
|
|
98
|
+
"type": "object",
|
|
99
|
+
"additionalProperties": false,
|
|
100
|
+
"required": [
|
|
101
|
+
"system",
|
|
102
|
+
"generatedAt",
|
|
103
|
+
"scenarios",
|
|
104
|
+
"passed",
|
|
105
|
+
"passRate",
|
|
106
|
+
"decisionCorrect",
|
|
107
|
+
"decisionAccuracy",
|
|
108
|
+
"preventionRate",
|
|
109
|
+
"falseBlockRate",
|
|
110
|
+
"usefulWarningPrecision",
|
|
111
|
+
"evidenceRecall",
|
|
112
|
+
"redactionLeaks",
|
|
113
|
+
"recallDegradationDetectionRate",
|
|
114
|
+
"latency"
|
|
115
|
+
],
|
|
116
|
+
"properties": {
|
|
117
|
+
"system": { "type": "string", "minLength": 1 },
|
|
118
|
+
"generatedAt": { "type": "string", "minLength": 1 },
|
|
119
|
+
"scenarios": { "type": "integer", "minimum": 0 },
|
|
120
|
+
"passed": { "type": "integer", "minimum": 0 },
|
|
121
|
+
"passRate": { "type": "number", "minimum": 0, "maximum": 1 },
|
|
122
|
+
"decisionCorrect": { "type": "integer", "minimum": 0 },
|
|
123
|
+
"decisionAccuracy": { "type": "number", "minimum": 0, "maximum": 1 },
|
|
124
|
+
"preventionRate": { "type": "number", "minimum": 0, "maximum": 1 },
|
|
125
|
+
"falseBlockRate": { "type": "number", "minimum": 0, "maximum": 1 },
|
|
126
|
+
"usefulWarningPrecision": {
|
|
127
|
+
"anyOf": [
|
|
128
|
+
{ "type": "null" },
|
|
129
|
+
{ "type": "number", "minimum": 0, "maximum": 1 }
|
|
130
|
+
]
|
|
131
|
+
},
|
|
132
|
+
"evidenceRecall": { "type": "number", "minimum": 0, "maximum": 1 },
|
|
133
|
+
"redactionLeaks": { "type": "integer", "minimum": 0 },
|
|
134
|
+
"recallDegradationDetectionRate": { "type": "number", "minimum": 0, "maximum": 1 },
|
|
135
|
+
"latency": { "$ref": "#/$defs/latency" }
|
|
136
|
+
}
|
|
137
|
+
},
|
|
138
|
+
"resultRow": {
|
|
139
|
+
"type": "object",
|
|
140
|
+
"additionalProperties": false,
|
|
141
|
+
"required": [
|
|
142
|
+
"system",
|
|
143
|
+
"id",
|
|
144
|
+
"name",
|
|
145
|
+
"expectedDecision",
|
|
146
|
+
"decision",
|
|
147
|
+
"decisionCorrect",
|
|
148
|
+
"riskScore",
|
|
149
|
+
"passed",
|
|
150
|
+
"latencyMs",
|
|
151
|
+
"evidenceCount",
|
|
152
|
+
"evidenceIds",
|
|
153
|
+
"recommendedActions",
|
|
154
|
+
"summary",
|
|
155
|
+
"recallErrors",
|
|
156
|
+
"leakedSecrets",
|
|
157
|
+
"requiredEvidenceMatched"
|
|
158
|
+
],
|
|
159
|
+
"properties": {
|
|
160
|
+
"system": { "type": "string", "minLength": 1 },
|
|
161
|
+
"external": { "type": "boolean" },
|
|
162
|
+
"id": { "type": "string", "pattern": "^GB-[0-9]{2}$" },
|
|
163
|
+
"name": { "type": "string", "minLength": 1 },
|
|
164
|
+
"expectedDecision": { "enum": ["allow", "warn", "block"] },
|
|
165
|
+
"decision": { "enum": ["allow", "warn", "block"] },
|
|
166
|
+
"decisionCorrect": { "type": "boolean" },
|
|
167
|
+
"riskScore": { "type": "number", "minimum": 0, "maximum": 1 },
|
|
168
|
+
"passed": { "type": "boolean" },
|
|
169
|
+
"latencyMs": { "type": "number", "minimum": 0 },
|
|
170
|
+
"evidenceCount": { "type": "integer", "minimum": 0 },
|
|
171
|
+
"evidenceIds": {
|
|
172
|
+
"type": "array",
|
|
173
|
+
"items": { "type": "string" }
|
|
174
|
+
},
|
|
175
|
+
"recommendedActions": {
|
|
176
|
+
"type": "array",
|
|
177
|
+
"items": { "type": "string" }
|
|
178
|
+
},
|
|
179
|
+
"summary": { "type": "string", "minLength": 1 },
|
|
180
|
+
"recallErrors": { "type": "array" },
|
|
181
|
+
"leakedSecrets": {
|
|
182
|
+
"type": "array",
|
|
183
|
+
"items": { "type": "string" }
|
|
184
|
+
},
|
|
185
|
+
"requiredEvidenceMatched": { "type": "boolean" }
|
|
186
|
+
}
|
|
187
|
+
},
|
|
188
|
+
"caseResult": {
|
|
189
|
+
"type": "object",
|
|
190
|
+
"additionalProperties": false,
|
|
191
|
+
"required": ["id", "name", "expectedDecision", "results"],
|
|
192
|
+
"properties": {
|
|
193
|
+
"id": { "type": "string", "pattern": "^GB-[0-9]{2}$" },
|
|
194
|
+
"name": { "type": "string", "minLength": 1 },
|
|
195
|
+
"expectedDecision": { "enum": ["allow", "warn", "block"] },
|
|
196
|
+
"results": {
|
|
197
|
+
"type": "array",
|
|
198
|
+
"minItems": 5,
|
|
199
|
+
"items": { "$ref": "#/$defs/resultRow" }
|
|
200
|
+
}
|
|
201
|
+
}
|
|
202
|
+
},
|
|
203
|
+
"artifactRedactionSweep": {
|
|
204
|
+
"type": "object",
|
|
205
|
+
"additionalProperties": false,
|
|
206
|
+
"required": [
|
|
207
|
+
"checkedAt",
|
|
208
|
+
"filesChecked",
|
|
209
|
+
"seededSecretsChecked",
|
|
210
|
+
"leakCount",
|
|
211
|
+
"leaks",
|
|
212
|
+
"passed"
|
|
213
|
+
],
|
|
214
|
+
"properties": {
|
|
215
|
+
"checkedAt": { "type": "string", "minLength": 1 },
|
|
216
|
+
"filesChecked": {
|
|
217
|
+
"type": "array",
|
|
218
|
+
"minItems": 3,
|
|
219
|
+
"items": { "type": "string", "minLength": 1 }
|
|
220
|
+
},
|
|
221
|
+
"seededSecretsChecked": { "type": "integer", "minimum": 0 },
|
|
222
|
+
"leakCount": { "type": "integer", "minimum": 0 },
|
|
223
|
+
"leaks": { "type": "array" },
|
|
224
|
+
"passed": { "type": "boolean" }
|
|
225
|
+
}
|
|
226
|
+
}
|
|
227
|
+
}
|
|
228
|
+
}
|