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.
- package/CHANGELOG.md +101 -15
- package/LICENSE +21 -21
- package/README.md +232 -6
- 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 +1125 -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 +1271 -0
- package/benchmarks/output/guardbench-summary.json +2107 -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 +1271 -0
- package/benchmarks/output/submission-bundle/guardbench-summary.json +2107 -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 +184 -0
- package/benchmarks/output/submission-bundle/schemas/guardbench-submission-manifest.schema.json +151 -0
- package/benchmarks/output/submission-bundle/schemas/guardbench-summary.schema.json +249 -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 +184 -0
- package/benchmarks/schemas/guardbench-submission-manifest.schema.json +151 -0
- package/benchmarks/schemas/guardbench-summary.schema.json +249 -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 +13 -0
- package/dist/src/audrey.d.ts.map +1 -1
- package/dist/src/audrey.js +68 -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 +78 -6
- package/dist/src/controller.d.ts.map +1 -1
- package/dist/src/controller.js +273 -53
- 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 +5 -4
- package/dist/src/index.d.ts.map +1 -1
- package/dist/src/index.js +3 -3
- 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 +71 -2
- 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 +555 -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 +1271 -0
- package/docs/paper/output/submission-bundle/benchmarks/output/guardbench-summary.json +2107 -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 +184 -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 +249 -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 +92 -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 +785 -0
|
@@ -0,0 +1,176 @@
|
|
|
1
|
+
import { cpSync, existsSync, mkdirSync, readFileSync, rmSync, writeFileSync } from 'node:fs';
|
|
2
|
+
import { createHash } from 'node:crypto';
|
|
3
|
+
import { basename, isAbsolute, join, relative, resolve } from 'node:path';
|
|
4
|
+
import { writeGuardBenchConformanceCard } from './create-conformance-card.mjs';
|
|
5
|
+
import { validateGuardBenchArtifacts } from './validate-guardbench-artifacts.mjs';
|
|
6
|
+
import { publicPath } from './public-paths.mjs';
|
|
7
|
+
|
|
8
|
+
const REQUIRED_ARTIFACTS = [
|
|
9
|
+
'guardbench-manifest.json',
|
|
10
|
+
'guardbench-summary.json',
|
|
11
|
+
'guardbench-raw.json',
|
|
12
|
+
'guardbench-conformance-card.json',
|
|
13
|
+
];
|
|
14
|
+
const OPTIONAL_ARTIFACTS = ['external-run-metadata.json'];
|
|
15
|
+
const SCHEMA_FILES = [
|
|
16
|
+
'guardbench-adapter-registry.schema.json',
|
|
17
|
+
'guardbench-adapter-self-test.schema.json',
|
|
18
|
+
'guardbench-external-dry-run.schema.json',
|
|
19
|
+
'guardbench-external-evidence.schema.json',
|
|
20
|
+
'guardbench-publication-verification.schema.json',
|
|
21
|
+
'guardbench-manifest.schema.json',
|
|
22
|
+
'guardbench-summary.schema.json',
|
|
23
|
+
'guardbench-raw.schema.json',
|
|
24
|
+
'guardbench-external-run.schema.json',
|
|
25
|
+
'guardbench-conformance-card.schema.json',
|
|
26
|
+
'guardbench-leaderboard.schema.json',
|
|
27
|
+
'guardbench-submission-manifest.schema.json',
|
|
28
|
+
];
|
|
29
|
+
|
|
30
|
+
function readJson(path) {
|
|
31
|
+
return JSON.parse(readFileSync(path, 'utf-8'));
|
|
32
|
+
}
|
|
33
|
+
|
|
34
|
+
function sha256File(path) {
|
|
35
|
+
return createHash('sha256').update(readFileSync(path)).digest('hex');
|
|
36
|
+
}
|
|
37
|
+
|
|
38
|
+
function copyFileInto(sourceDir, outDir, file) {
|
|
39
|
+
const source = join(sourceDir, file);
|
|
40
|
+
if (!existsSync(source)) return null;
|
|
41
|
+
const target = join(outDir, file);
|
|
42
|
+
cpSync(source, target);
|
|
43
|
+
return target;
|
|
44
|
+
}
|
|
45
|
+
|
|
46
|
+
export function bundleRelativeFilePath(path, root) {
|
|
47
|
+
const relativePath = relative(root, path);
|
|
48
|
+
if (!relativePath || relativePath.startsWith('..') || isAbsolute(relativePath)) {
|
|
49
|
+
throw new Error(`Cannot add file outside GuardBench submission bundle: ${path}`);
|
|
50
|
+
}
|
|
51
|
+
return relativePath.replaceAll('\\', '/');
|
|
52
|
+
}
|
|
53
|
+
|
|
54
|
+
function fileRecord(path, root) {
|
|
55
|
+
return {
|
|
56
|
+
path: bundleRelativeFilePath(path, root),
|
|
57
|
+
bytes: readFileSync(path).byteLength,
|
|
58
|
+
sha256: sha256File(path),
|
|
59
|
+
};
|
|
60
|
+
}
|
|
61
|
+
|
|
62
|
+
export function writeGuardBenchSubmissionBundle(options = {}) {
|
|
63
|
+
const sourceDir = resolve(options.dir ?? 'benchmarks/output');
|
|
64
|
+
const outDir = resolve(options.outDir ?? join(sourceDir, 'submission-bundle'));
|
|
65
|
+
const schemasDir = resolve(options.schemasDir ?? 'benchmarks/schemas');
|
|
66
|
+
|
|
67
|
+
writeGuardBenchConformanceCard({ dir: sourceDir });
|
|
68
|
+
const sourceValidation = validateGuardBenchArtifacts({ dir: sourceDir, schemasDir });
|
|
69
|
+
if (!sourceValidation.ok) {
|
|
70
|
+
throw new Error(`Cannot create GuardBench submission bundle from invalid artifacts: ${sourceValidation.failures.join('; ')}`);
|
|
71
|
+
}
|
|
72
|
+
|
|
73
|
+
rmSync(outDir, { recursive: true, force: true });
|
|
74
|
+
mkdirSync(join(outDir, 'schemas'), { recursive: true });
|
|
75
|
+
|
|
76
|
+
const copied = [];
|
|
77
|
+
for (const file of [...REQUIRED_ARTIFACTS, ...OPTIONAL_ARTIFACTS]) {
|
|
78
|
+
const target = copyFileInto(sourceDir, outDir, file);
|
|
79
|
+
if (target) copied.push(target);
|
|
80
|
+
}
|
|
81
|
+
for (const file of SCHEMA_FILES) {
|
|
82
|
+
const target = join(outDir, 'schemas', file);
|
|
83
|
+
cpSync(join(schemasDir, file), target);
|
|
84
|
+
copied.push(target);
|
|
85
|
+
}
|
|
86
|
+
|
|
87
|
+
const bundleValidation = validateGuardBenchArtifacts({
|
|
88
|
+
dir: outDir,
|
|
89
|
+
schemasDir: join(outDir, 'schemas'),
|
|
90
|
+
});
|
|
91
|
+
const validationReportPath = join(outDir, 'validation-report.json');
|
|
92
|
+
writeFileSync(validationReportPath, `${JSON.stringify({
|
|
93
|
+
generatedAt: new Date().toISOString(),
|
|
94
|
+
sourceValidation,
|
|
95
|
+
bundleValidation,
|
|
96
|
+
}, null, 2)}\n`, 'utf-8');
|
|
97
|
+
copied.push(validationReportPath);
|
|
98
|
+
|
|
99
|
+
const card = readJson(join(outDir, 'guardbench-conformance-card.json'));
|
|
100
|
+
const manifestPath = join(outDir, 'submission-manifest.json');
|
|
101
|
+
const manifest = {
|
|
102
|
+
schemaVersion: '1.0.0',
|
|
103
|
+
suite: 'GuardBench submission bundle',
|
|
104
|
+
generatedAt: new Date().toISOString(),
|
|
105
|
+
sourceDir: publicPath(sourceDir),
|
|
106
|
+
subject: card.subject,
|
|
107
|
+
score: card.score,
|
|
108
|
+
conformance: card.conformance,
|
|
109
|
+
validation: bundleValidation,
|
|
110
|
+
files: copied.map(path => fileRecord(path, outDir)).sort((a, b) => a.path.localeCompare(b.path)),
|
|
111
|
+
};
|
|
112
|
+
writeFileSync(manifestPath, `${JSON.stringify(manifest, null, 2)}\n`, 'utf-8');
|
|
113
|
+
|
|
114
|
+
const finalFiles = copied.map(path => fileRecord(path, outDir)).sort((a, b) => a.path.localeCompare(b.path));
|
|
115
|
+
manifest.files = finalFiles;
|
|
116
|
+
writeFileSync(manifestPath, `${JSON.stringify(manifest, null, 2)}\n`, 'utf-8');
|
|
117
|
+
|
|
118
|
+
return {
|
|
119
|
+
outDir,
|
|
120
|
+
manifestPath,
|
|
121
|
+
validation: bundleValidation,
|
|
122
|
+
subject: card.subject,
|
|
123
|
+
files: finalFiles.map(record => record.path),
|
|
124
|
+
};
|
|
125
|
+
}
|
|
126
|
+
|
|
127
|
+
function parseArgs(argv = process.argv.slice(2)) {
|
|
128
|
+
const args = {
|
|
129
|
+
dir: 'benchmarks/output',
|
|
130
|
+
outDir: null,
|
|
131
|
+
schemasDir: 'benchmarks/schemas',
|
|
132
|
+
json: false,
|
|
133
|
+
};
|
|
134
|
+
for (let i = 0; i < argv.length; i++) {
|
|
135
|
+
const token = argv[i];
|
|
136
|
+
if ((token === '--dir' || token === '--source-dir') && argv[i + 1]) args.dir = argv[++i];
|
|
137
|
+
else if (token === '--out-dir' && argv[i + 1]) args.outDir = argv[++i];
|
|
138
|
+
else if (token === '--schemas-dir' && argv[i + 1]) args.schemasDir = argv[++i];
|
|
139
|
+
else if (token === '--json') args.json = true;
|
|
140
|
+
else if (token === '--help' || token === '-h') args.help = true;
|
|
141
|
+
else throw new Error(`Unknown argument: ${token}`);
|
|
142
|
+
}
|
|
143
|
+
return args;
|
|
144
|
+
}
|
|
145
|
+
|
|
146
|
+
function usage() {
|
|
147
|
+
return [
|
|
148
|
+
'Usage: node benchmarks/create-submission-bundle.mjs [--dir benchmarks/output] [--out-dir <dir>] [--json]',
|
|
149
|
+
'',
|
|
150
|
+
'Creates a portable GuardBench submission bundle containing artifacts, schemas,',
|
|
151
|
+
'a conformance card, validation report, and submission manifest.',
|
|
152
|
+
].join('\n');
|
|
153
|
+
}
|
|
154
|
+
|
|
155
|
+
async function main() {
|
|
156
|
+
const args = parseArgs();
|
|
157
|
+
if (args.help) {
|
|
158
|
+
console.log(usage());
|
|
159
|
+
return;
|
|
160
|
+
}
|
|
161
|
+
const result = writeGuardBenchSubmissionBundle(args);
|
|
162
|
+
if (args.json) console.log(JSON.stringify(result, null, 2));
|
|
163
|
+
else {
|
|
164
|
+
console.log(`GuardBench submission bundle: ${result.outDir}`);
|
|
165
|
+
console.log(`Subject: ${result.subject.name}`);
|
|
166
|
+
console.log(`Validation: ${result.validation.ok ? 'passed' : 'failed'}`);
|
|
167
|
+
}
|
|
168
|
+
if (!result.validation.ok) process.exit(1);
|
|
169
|
+
}
|
|
170
|
+
|
|
171
|
+
if (process.argv[1] && resolve(process.argv[1]).endsWith(basename(import.meta.url))) {
|
|
172
|
+
main().catch(error => {
|
|
173
|
+
console.error(error.stack ?? error.message);
|
|
174
|
+
process.exit(1);
|
|
175
|
+
});
|
|
176
|
+
}
|
|
@@ -0,0 +1,165 @@
|
|
|
1
|
+
import { mkdirSync, readFileSync, writeFileSync } from 'node:fs';
|
|
2
|
+
import { dirname, join, resolve } from 'node:path';
|
|
3
|
+
import { fileURLToPath } from 'node:url';
|
|
4
|
+
import { buildExternalGuardBenchRun, writeExternalRunMetadata } from './run-external-guardbench.mjs';
|
|
5
|
+
import { validateAdapterRegistry } from './validate-adapter-registry.mjs';
|
|
6
|
+
import { validateSchema } from './validate-guardbench-artifacts.mjs';
|
|
7
|
+
import { publicCommand, publicPath } from './public-paths.mjs';
|
|
8
|
+
|
|
9
|
+
const DEFAULT_REGISTRY = 'benchmarks/adapters/registry.json';
|
|
10
|
+
const DEFAULT_OUT_ROOT = 'benchmarks/output/external';
|
|
11
|
+
const DEFAULT_OUT = 'benchmarks/output/external/guardbench-external-dry-run.json';
|
|
12
|
+
const DEFAULT_SCHEMA = 'benchmarks/schemas/guardbench-external-dry-run.schema.json';
|
|
13
|
+
|
|
14
|
+
export function parseExternalDryRunArgs(argv = process.argv.slice(2)) {
|
|
15
|
+
const args = {
|
|
16
|
+
registry: DEFAULT_REGISTRY,
|
|
17
|
+
outRoot: DEFAULT_OUT_ROOT,
|
|
18
|
+
out: DEFAULT_OUT,
|
|
19
|
+
includeCredentialFree: false,
|
|
20
|
+
json: false,
|
|
21
|
+
noWrite: false,
|
|
22
|
+
};
|
|
23
|
+
|
|
24
|
+
for (let i = 0; i < argv.length; i++) {
|
|
25
|
+
const token = argv[i];
|
|
26
|
+
if (token === '--registry' && argv[i + 1]) args.registry = argv[++i];
|
|
27
|
+
else if (token === '--out-root' && argv[i + 1]) args.outRoot = argv[++i];
|
|
28
|
+
else if (token === '--out' && argv[i + 1]) args.out = argv[++i];
|
|
29
|
+
else if (token === '--include-credential-free') args.includeCredentialFree = true;
|
|
30
|
+
else if (token === '--json') args.json = true;
|
|
31
|
+
else if (token === '--no-write') args.noWrite = true;
|
|
32
|
+
else if (token === '--help' || token === '-h') args.help = true;
|
|
33
|
+
else throw new Error(`Unknown argument: ${token}`);
|
|
34
|
+
}
|
|
35
|
+
|
|
36
|
+
return args;
|
|
37
|
+
}
|
|
38
|
+
|
|
39
|
+
function usage() {
|
|
40
|
+
return `Usage: node benchmarks/dry-run-external-adapters.mjs [options]
|
|
41
|
+
|
|
42
|
+
Options:
|
|
43
|
+
--registry <path> Adapter registry JSON. Default: ${DEFAULT_REGISTRY}.
|
|
44
|
+
--out-root <path> Root directory for dry-run metadata. Default: ${DEFAULT_OUT_ROOT}.
|
|
45
|
+
--out <path> Matrix JSON report path. Default: ${DEFAULT_OUT}.
|
|
46
|
+
--include-credential-free Include credential-free registry adapters.
|
|
47
|
+
--json Print the machine-readable dry-run matrix.
|
|
48
|
+
--no-write Do not write the matrix JSON report.
|
|
49
|
+
`;
|
|
50
|
+
}
|
|
51
|
+
|
|
52
|
+
function readJson(path) {
|
|
53
|
+
return JSON.parse(readFileSync(path, 'utf-8'));
|
|
54
|
+
}
|
|
55
|
+
|
|
56
|
+
export function validateExternalAdapterDryRunMatrix(matrix, options = {}) {
|
|
57
|
+
const schema = options.schemaObject ?? readJson(options.schema ?? DEFAULT_SCHEMA);
|
|
58
|
+
return validateSchema(matrix, schema, 'guardbench-external-dry-run');
|
|
59
|
+
}
|
|
60
|
+
|
|
61
|
+
export function writeExternalAdapterDryRunMatrix(matrix, out = DEFAULT_OUT) {
|
|
62
|
+
const outPath = resolve(out);
|
|
63
|
+
mkdirSync(dirname(outPath), { recursive: true });
|
|
64
|
+
writeFileSync(outPath, `${JSON.stringify(matrix, null, 2)}\n`, 'utf-8');
|
|
65
|
+
return outPath;
|
|
66
|
+
}
|
|
67
|
+
|
|
68
|
+
export async function buildExternalAdapterDryRunMatrix(options = {}) {
|
|
69
|
+
const registryPath = options.registry ?? DEFAULT_REGISTRY;
|
|
70
|
+
const outRoot = resolve(options.outRoot ?? DEFAULT_OUT_ROOT);
|
|
71
|
+
const registryValidation = await validateAdapterRegistry({ registry: registryPath });
|
|
72
|
+
const registry = readJson(registryPath);
|
|
73
|
+
const adapters = registry.adapters.filter(adapter =>
|
|
74
|
+
adapter.status === 'external-system'
|
|
75
|
+
&& (options.includeCredentialFree || adapter.credentialMode === 'runtime-env'));
|
|
76
|
+
const rows = [];
|
|
77
|
+
const failures = [];
|
|
78
|
+
|
|
79
|
+
if (!registryValidation.ok) {
|
|
80
|
+
failures.push(...registryValidation.failures.map(failure => `registry: ${failure}`));
|
|
81
|
+
}
|
|
82
|
+
|
|
83
|
+
for (const adapter of adapters) {
|
|
84
|
+
const run = buildExternalGuardBenchRun({
|
|
85
|
+
adapter: adapter.id,
|
|
86
|
+
outDir: join(outRoot, adapter.id),
|
|
87
|
+
check: true,
|
|
88
|
+
json: true,
|
|
89
|
+
}, options.env ?? process.env);
|
|
90
|
+
const metadata = {
|
|
91
|
+
suite: 'GuardBench external adapter run',
|
|
92
|
+
startedAt: new Date().toISOString(),
|
|
93
|
+
adapter: run.adapter,
|
|
94
|
+
adapterPath: run.adapterPath,
|
|
95
|
+
outDir: run.outDir,
|
|
96
|
+
requiredEnv: run.requiredEnv,
|
|
97
|
+
missingEnv: run.missingEnv,
|
|
98
|
+
command: run.command,
|
|
99
|
+
validationCommand: run.validationCommand,
|
|
100
|
+
dryRun: true,
|
|
101
|
+
status: run.missingEnv.length ? 'dry-run-missing-env' : 'dry-run-ready',
|
|
102
|
+
};
|
|
103
|
+
const metadataPath = writeExternalRunMetadata(run.outDir, metadata);
|
|
104
|
+
rows.push({
|
|
105
|
+
id: adapter.id,
|
|
106
|
+
name: adapter.name,
|
|
107
|
+
credentialMode: adapter.credentialMode,
|
|
108
|
+
requiredEnv: run.requiredEnv,
|
|
109
|
+
missingEnv: run.missingEnv,
|
|
110
|
+
status: metadata.status,
|
|
111
|
+
command: publicCommand(run.command),
|
|
112
|
+
validationCommand: publicCommand(run.validationCommand),
|
|
113
|
+
metadataPath: publicPath(metadataPath),
|
|
114
|
+
});
|
|
115
|
+
}
|
|
116
|
+
|
|
117
|
+
const matrix = {
|
|
118
|
+
schemaVersion: '1.0.0',
|
|
119
|
+
suite: 'GuardBench external adapter dry-run matrix',
|
|
120
|
+
generatedAt: new Date().toISOString(),
|
|
121
|
+
ok: failures.length === 0,
|
|
122
|
+
registry: publicPath(resolve(registryPath)),
|
|
123
|
+
outRoot: publicPath(outRoot),
|
|
124
|
+
adapters: rows,
|
|
125
|
+
failures,
|
|
126
|
+
};
|
|
127
|
+
const schemaFailures = validateExternalAdapterDryRunMatrix(matrix, options);
|
|
128
|
+
if (schemaFailures.length > 0) {
|
|
129
|
+
throw new Error(`GuardBench external adapter dry-run schema validation failed: ${schemaFailures.join('; ')}`);
|
|
130
|
+
}
|
|
131
|
+
return matrix;
|
|
132
|
+
}
|
|
133
|
+
|
|
134
|
+
async function main() {
|
|
135
|
+
const args = parseExternalDryRunArgs();
|
|
136
|
+
if (args.help) {
|
|
137
|
+
console.log(usage());
|
|
138
|
+
return;
|
|
139
|
+
}
|
|
140
|
+
|
|
141
|
+
const matrix = await buildExternalAdapterDryRunMatrix(args);
|
|
142
|
+
const outPath = args.noWrite ? null : writeExternalAdapterDryRunMatrix(matrix, args.out);
|
|
143
|
+
if (args.json) {
|
|
144
|
+
console.log(JSON.stringify(matrix, null, 2));
|
|
145
|
+
} else if (matrix.ok) {
|
|
146
|
+
console.log(`GuardBench external adapter dry-run matrix passed: ${matrix.adapters.length} adapter(s)`);
|
|
147
|
+
for (const row of matrix.adapters) {
|
|
148
|
+
const missing = row.missingEnv.length ? `missing ${row.missingEnv.join(', ')}` : 'ready';
|
|
149
|
+
console.log(`- ${row.id}: ${missing}; metadata ${row.metadataPath}`);
|
|
150
|
+
}
|
|
151
|
+
if (outPath) console.log(`Matrix report: ${outPath}`);
|
|
152
|
+
} else {
|
|
153
|
+
console.error('GuardBench external adapter dry-run matrix failed:');
|
|
154
|
+
for (const failure of matrix.failures) console.error(`- ${failure}`);
|
|
155
|
+
}
|
|
156
|
+
|
|
157
|
+
if (!matrix.ok) process.exit(1);
|
|
158
|
+
}
|
|
159
|
+
|
|
160
|
+
if (process.argv[1] && resolve(process.argv[1]) === fileURLToPath(import.meta.url)) {
|
|
161
|
+
main().catch(error => {
|
|
162
|
+
console.error(error.stack ?? error.message);
|
|
163
|
+
process.exit(1);
|
|
164
|
+
});
|
|
165
|
+
}
|