create-windy 0.8.3 → 0.9.1
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/dist/cli.js +838 -560
- package/package.json +1 -1
- package/template/.windy-template.json +2 -2
- package/template/apps/agent-server/Dockerfile +92 -22
- package/template/apps/agent-server/README.md +47 -85
- package/template/apps/agent-server/documents/office.py +216 -0
- package/template/apps/agent-server/documents/query_report.py +136 -0
- package/template/apps/agent-server/documents/smoke.py +177 -0
- package/template/apps/agent-server/documents/test_boundaries.py +185 -0
- package/template/apps/agent-server/host-contracts.ts +56 -0
- package/template/apps/agent-server/install-native.ts +167 -0
- package/template/apps/agent-server/main.ts +179 -0
- package/template/apps/agent-server/native/EXTENSIONS.md +46 -0
- package/template/apps/agent-server/native/admission.ts +73 -0
- package/template/apps/agent-server/native/authorization.test.ts +176 -0
- package/template/apps/agent-server/native/authorization.ts +118 -0
- package/template/apps/agent-server/native/boundary-io.ts +93 -0
- package/template/apps/agent-server/native/boundary-product.ts +107 -0
- package/template/apps/agent-server/native/boundary-project.test.ts +247 -0
- package/template/apps/agent-server/native/boundary-project.ts +188 -0
- package/template/apps/agent-server/native/boundary.ts +188 -0
- package/template/apps/agent-server/native/compaction-policy.test.ts +175 -0
- package/template/apps/agent-server/native/config.ts +95 -0
- package/template/apps/agent-server/native/conversation-instructions.ts +9 -0
- package/template/apps/agent-server/native/extensions.test.ts +190 -0
- package/template/apps/agent-server/native/extensions.ts +31 -0
- package/template/apps/agent-server/native/host-profile.test.ts +91 -0
- package/template/apps/agent-server/native/host-profile.ts +33 -0
- package/template/apps/agent-server/native/host-test-fixture.ts +13 -0
- package/template/apps/agent-server/native/http.ts +129 -0
- package/template/apps/agent-server/native/lifecycle.test.ts +285 -0
- package/template/apps/agent-server/native/lifecycle.ts +136 -0
- package/template/apps/agent-server/native/manager.ts +292 -0
- package/template/apps/agent-server/native/native.test.ts +207 -0
- package/template/apps/agent-server/native/office-preview.test.ts +50 -0
- package/template/apps/agent-server/native/office-preview.ts +25 -0
- package/template/apps/agent-server/native/parallel-harness.ts +242 -0
- package/template/apps/agent-server/native/parallel-mcp-fixture.ts +97 -0
- package/template/apps/agent-server/native/parallel.test.ts +256 -0
- package/template/apps/agent-server/native/profile-composition.test.ts +147 -0
- package/template/apps/agent-server/native/profile.ts +307 -0
- package/template/apps/agent-server/native/request.ts +61 -0
- package/template/apps/agent-server/native/session-conflict.test.ts +235 -0
- package/template/apps/agent-server/native/session-conflict.ts +21 -0
- package/template/apps/agent-server/native/shared-skills.test.ts +128 -0
- package/template/apps/agent-server/native/shared-skills.ts +240 -0
- package/template/apps/agent-server/native/socket-lifecycle.test.ts +48 -0
- package/template/apps/agent-server/native/socket.ts +130 -0
- package/template/apps/agent-server/native/storage-lifecycle.test.ts +217 -0
- package/template/apps/agent-server/native/storage.ts +188 -0
- package/template/apps/agent-server/native/stream-authorization.test.ts +89 -0
- package/template/apps/agent-server/native/worker.ts +77 -0
- package/template/apps/agent-server/native-download.test.ts +119 -0
- package/template/apps/agent-server/native-download.ts +145 -0
- package/template/apps/agent-server/package.json +63 -0
- package/template/apps/agent-server/plugins/access.mjs +126 -0
- package/template/apps/agent-server/plugins/branding/README.md +9 -0
- package/template/apps/agent-server/plugins/branding/brand.css +85 -0
- package/template/apps/agent-server/plugins/branding/client.js +338 -0
- package/template/apps/agent-server/plugins/branding/index.mjs +35 -0
- package/template/apps/agent-server/plugins/branding/loading.d.mts +2 -0
- package/template/apps/agent-server/plugins/branding/loading.mjs +66 -0
- package/template/apps/agent-server/plugins/branding/loading.test.ts +89 -0
- package/template/apps/agent-server/plugins/branding/package.json +21 -0
- package/template/apps/agent-server/plugins/branding.test.ts +294 -0
- package/template/apps/agent-server/plugins/compaction.mjs +83 -0
- package/template/apps/agent-server/plugins/compaction.test.ts +91 -0
- package/template/apps/agent-server/plugins/confine-network.py +57 -0
- package/template/apps/agent-server/plugins/default-skills.test.ts +62 -0
- package/template/apps/agent-server/plugins/downloads/README.md +28 -0
- package/template/apps/agent-server/plugins/downloads/client.js +126 -0
- package/template/apps/agent-server/plugins/downloads/downloads.test.ts +222 -0
- package/template/apps/agent-server/plugins/downloads/index.d.mts +19 -0
- package/template/apps/agent-server/plugins/downloads/index.mjs +180 -0
- package/template/apps/agent-server/plugins/downloads/package.json +20 -0
- package/template/apps/agent-server/plugins/drop-overlay/README.md +24 -0
- package/template/apps/agent-server/plugins/drop-overlay/drop-overlay.test.ts +199 -0
- package/template/apps/agent-server/plugins/drop-overlay/index.d.mts +8 -0
- package/template/apps/agent-server/plugins/drop-overlay/index.mjs +12 -0
- package/template/apps/agent-server/plugins/drop-overlay/style.css +102 -0
- package/template/apps/agent-server/plugins/locale/client.js +61 -0
- package/template/apps/agent-server/plugins/locale/index.d.mts +9 -0
- package/template/apps/agent-server/plugins/locale/index.mjs +16 -0
- package/template/apps/agent-server/plugins/locale/locale.test.ts +217 -0
- package/template/apps/agent-server/plugins/locale/package.json +19 -0
- package/template/apps/agent-server/plugins/parallel.mjs +146 -0
- package/template/apps/agent-server/plugins/preset-registry.d.mts +2 -0
- package/template/apps/agent-server/plugins/preset-registry.mjs +41 -0
- package/template/apps/agent-server/plugins/preset-registry.test.ts +46 -0
- package/template/apps/agent-server/plugins/questions.d.mts +25 -0
- package/template/apps/agent-server/plugins/questions.mjs +142 -0
- package/template/apps/agent-server/plugins/questions.test.ts +105 -0
- package/template/apps/agent-server/plugins/settings/README.md +17 -0
- package/template/apps/agent-server/plugins/settings/client.js +102 -0
- package/template/apps/agent-server/plugins/settings/index.mjs +17 -0
- package/template/apps/agent-server/plugins/settings/package.json +19 -0
- package/template/apps/agent-server/plugins/settings/settings.test.ts +129 -0
- package/template/apps/agent-server/plugins/settings-controller.d.mts +4 -0
- package/template/apps/agent-server/plugins/settings-controller.mjs +31 -0
- package/template/apps/agent-server/plugins/settings-controller.test.ts +17 -0
- package/template/apps/agent-server/plugins/settings-dispatch.test.ts +34 -0
- package/template/apps/agent-server/plugins/shared-skills.d.mts +15 -0
- package/template/apps/agent-server/plugins/shared-skills.mjs +82 -0
- package/template/apps/agent-server/plugins/shared-skills.test.ts +51 -0
- package/template/apps/agent-server/plugins/skill-library/README.md +11 -0
- package/template/apps/agent-server/plugins/skill-library/archive.test.ts +43 -0
- package/template/apps/agent-server/plugins/skill-library/bundle.mjs +234 -0
- package/template/apps/agent-server/plugins/skill-library/client-chunks.test.ts +107 -0
- package/template/apps/agent-server/plugins/skill-library/client-form-mount.test.ts +139 -0
- package/template/apps/agent-server/plugins/skill-library/client-form.test.ts +261 -0
- package/template/apps/agent-server/plugins/skill-library/client-ui.test.ts +268 -0
- package/template/apps/agent-server/plugins/skill-library/client-use.test.ts +158 -0
- package/template/apps/agent-server/plugins/skill-library/client.editor.js +93 -0
- package/template/apps/agent-server/plugins/skill-library/client.form.js +316 -0
- package/template/apps/agent-server/plugins/skill-library/client.js +348 -0
- package/template/apps/agent-server/plugins/skill-library/client.request.js +52 -0
- package/template/apps/agent-server/plugins/skill-library/client.transfer.js +48 -0
- package/template/apps/agent-server/plugins/skill-library/client.use.js +46 -0
- package/template/apps/agent-server/plugins/skill-library/default-roots.test.ts +84 -0
- package/template/apps/agent-server/plugins/skill-library/draft.d.mts +17 -0
- package/template/apps/agent-server/plugins/skill-library/draft.mjs +152 -0
- package/template/apps/agent-server/plugins/skill-library/draft.test.ts +287 -0
- package/template/apps/agent-server/plugins/skill-library/index.d.mts +49 -0
- package/template/apps/agent-server/plugins/skill-library/index.mjs +226 -0
- package/template/apps/agent-server/plugins/skill-library/package.json +24 -0
- package/template/apps/agent-server/plugins/skill-library/presentation/README.md +27 -0
- package/template/apps/agent-server/plugins/skill-library/presentation/UPSTREAM.md +54 -0
- package/template/apps/agent-server/plugins/skill-library/presentation/boot.test.ts +150 -0
- package/template/apps/agent-server/plugins/skill-library/presentation/client.js +146 -0
- package/template/apps/agent-server/plugins/skill-library/presentation/client.test.ts +195 -0
- package/template/apps/agent-server/plugins/skill-library/presentation/index.d.mts +9 -0
- package/template/apps/agent-server/plugins/skill-library/presentation/index.mjs +30 -0
- package/template/apps/agent-server/plugins/skill-library/presentation/metadata.test.ts +82 -0
- package/template/apps/agent-server/plugins/skill-library/presentation/package.json +21 -0
- package/template/apps/agent-server/plugins/skill-library/presentation/react-boot.test.ts +162 -0
- package/template/apps/agent-server/plugins/skill-library/skills.test.ts +342 -0
- package/template/apps/agent-server/plugins/skill-library/storage.mjs +293 -0
- package/template/apps/agent-server/plugins/skill-library/style.css +229 -0
- package/template/apps/agent-server/plugins/skill-library/test-fixture.ts +77 -0
- package/template/apps/agent-server/plugins/skill-library/workflow.d.mts +18 -0
- package/template/apps/agent-server/plugins/skill-library/workflow.mjs +189 -0
- package/template/apps/agent-server/plugins/skill-library/workflow.test.ts +164 -0
- package/template/apps/agent-server/plugins/skills/README.md +5 -0
- package/template/apps/agent-server/plugins/skills/windy-data-summary/SKILL.md +14 -0
- package/template/apps/agent-server/plugins/skills/windy-extension-guide/SKILL.md +14 -0
- package/template/apps/agent-server/plugins/skills/windy-office/SKILL.md +14 -0
- package/template/apps/agent-server/plugins/skills/windy-skill-capture/SKILL.md +14 -0
- package/template/apps/agent-server/plugins/skills/windy-source-compare/SKILL.md +14 -0
- package/template/apps/agent-server/plugins/subprocess.mjs +40 -0
- package/template/apps/agent-server/plugins/theme/README.md +18 -0
- package/template/apps/agent-server/plugins/theme/client.js +18 -0
- package/template/apps/agent-server/plugins/theme/index.d.mts +9 -0
- package/template/apps/agent-server/plugins/theme/index.mjs +18 -0
- package/template/apps/agent-server/plugins/theme/package.json +19 -0
- package/template/apps/agent-server/plugins/theme/source-tokens.ts +82 -0
- package/template/apps/agent-server/plugins/theme/sync-tokens.ts +12 -0
- package/template/apps/agent-server/plugins/theme/theme.test.ts +193 -0
- package/template/apps/agent-server/plugins/theme/tokens.json +166 -0
- package/template/apps/agent-server/plugins/tool-catalog/README.md +79 -0
- package/template/apps/agent-server/plugins/tool-catalog/categories.mjs +111 -0
- package/template/apps/agent-server/plugins/tool-catalog/categories.test.ts +118 -0
- package/template/apps/agent-server/plugins/tool-catalog/client-cache.test.ts +264 -0
- package/template/apps/agent-server/plugins/tool-catalog/client.js +333 -0
- package/template/apps/agent-server/plugins/tool-catalog/client.test.ts +136 -0
- package/template/apps/agent-server/plugins/tool-catalog/discovery.d.mts +42 -0
- package/template/apps/agent-server/plugins/tool-catalog/discovery.mjs +242 -0
- package/template/apps/agent-server/plugins/tool-catalog/discovery.test.ts +229 -0
- package/template/apps/agent-server/plugins/tool-catalog/index.d.mts +5 -0
- package/template/apps/agent-server/plugins/tool-catalog/index.mjs +91 -0
- package/template/apps/agent-server/plugins/tool-catalog/package.json +23 -0
- package/template/apps/agent-server/plugins/tool-catalog/presentation.d.mts +21 -0
- package/template/apps/agent-server/plugins/tool-catalog/presentation.mjs +153 -0
- package/template/apps/agent-server/plugins/tool-catalog/presentation.test.ts +166 -0
- package/template/apps/agent-server/plugins/tool-catalog/routes.test.ts +209 -0
- package/template/apps/agent-server/plugins/tool-catalog/style.css +155 -0
- package/template/apps/agent-server/plugins/tool-catalog/use-tool.test.ts +160 -0
- package/template/apps/agent-server/plugins/tool-presentation/README.md +7 -0
- package/template/apps/agent-server/plugins/tool-presentation/build-client.ts +29 -0
- package/template/apps/agent-server/plugins/tool-presentation/client.js +312 -0
- package/template/apps/agent-server/plugins/tool-presentation/client.results.js +9 -0
- package/template/apps/agent-server/plugins/tool-presentation/client.test.ts +178 -0
- package/template/apps/agent-server/plugins/tool-presentation/index.mjs +10 -0
- package/template/apps/agent-server/plugins/tool-presentation/package.json +24 -0
- package/template/apps/agent-server/plugins/tool-presentation/result-discovery.mjs +43 -0
- package/template/apps/agent-server/plugins/tool-presentation/result-model.d.mts +51 -0
- package/template/apps/agent-server/plugins/tool-presentation/result-model.mjs +96 -0
- package/template/apps/agent-server/plugins/workspace/README.md +5 -0
- package/template/apps/agent-server/plugins/workspace/index.mjs +8 -0
- package/template/apps/agent-server/plugins/workspace/package.json +10 -0
- package/template/apps/agent-server/plugins/workspace/style.css +124 -0
- package/template/apps/agent-server/project/host.ts +13 -0
- package/template/apps/agent-server/runtime/artifacts.ts +115 -0
- package/template/apps/agent-server/runtime/audit.nodecheck.ts +60 -0
- package/template/apps/agent-server/runtime/capacity.ts +61 -0
- package/template/apps/agent-server/runtime/client-completion-budget.nodecheck.ts +92 -0
- package/template/apps/agent-server/runtime/compatibility.ts +33 -0
- package/template/apps/agent-server/runtime/completions.ts +320 -0
- package/template/apps/agent-server/runtime/config.ts +246 -0
- package/template/apps/agent-server/runtime/context-budget.ts +70 -0
- package/template/apps/agent-server/runtime/context-plugin.ts +306 -0
- package/template/apps/agent-server/runtime/context.nodecheck.ts +241 -0
- package/template/apps/agent-server/runtime/contracts.ts +201 -0
- package/template/apps/agent-server/runtime/dsh-engine.ts +196 -0
- package/template/apps/agent-server/runtime/evaluation-plugin.ts +115 -0
- package/template/apps/agent-server/runtime/evaluation.nodecheck.ts +194 -0
- package/template/apps/agent-server/runtime/events.ts +71 -0
- package/template/apps/agent-server/runtime/governance.ts +194 -0
- package/template/apps/agent-server/runtime/index.ts +339 -0
- package/template/apps/agent-server/runtime/internal-output.ts +37 -0
- package/template/apps/agent-server/runtime/isolation.nodecheck.ts +287 -0
- package/template/apps/agent-server/runtime/model-payload.ts +97 -0
- package/template/apps/agent-server/runtime/model-plugin.ts +308 -0
- package/template/apps/agent-server/runtime/model-route.ts +34 -0
- package/template/apps/agent-server/runtime/model-usage.ts +51 -0
- package/template/apps/agent-server/runtime/model-wire.ts +45 -0
- package/template/apps/agent-server/runtime/native-authorization.nodecheck.ts +126 -0
- package/template/apps/agent-server/runtime/native-completion.nodecheck.ts +81 -0
- package/template/apps/agent-server/runtime/output-validation.ts +66 -0
- package/template/apps/agent-server/runtime/production-capacity.nodecheck.ts +121 -0
- package/template/apps/agent-server/runtime/questions-workflow.nodecheck.ts +249 -0
- package/template/apps/agent-server/runtime/report-breakdown.ts +79 -0
- package/template/apps/agent-server/runtime/reporting.ts +219 -0
- package/template/apps/agent-server/runtime/run-completion.ts +39 -0
- package/template/apps/agent-server/runtime/run-request-digest.ts +18 -0
- package/template/apps/agent-server/runtime/runtime-test-fixtures.ts +264 -0
- package/template/apps/agent-server/runtime/runtime.nodecheck.ts +150 -0
- package/template/apps/agent-server/runtime/store-idempotency.nodecheck.ts +74 -0
- package/template/apps/agent-server/runtime/store.ts +324 -0
- package/template/apps/agent-server/runtime/structured-output-mode.nodecheck.ts +51 -0
- package/template/apps/agent-server/runtime/tool-name.ts +6 -0
- package/template/apps/agent-server/runtime/tool-plugin.ts +201 -0
- package/template/apps/agent-server/runtime/vision-readiness.ts +200 -0
- package/template/apps/agent-server/scripts/confinement-check.mjs +103 -0
- package/template/apps/agent-server/scripts/coordinator-check.mjs +344 -0
- package/template/apps/agent-server/scripts/documents-delivery-check.mjs +309 -0
- package/template/apps/agent-server/scripts/office-preview-check.mjs +92 -0
- package/template/apps/agent-server/scripts/test-confinement.ts +58 -0
- package/template/apps/agent-server/scripts/test-coordinator.ts +36 -0
- package/template/apps/agent-server/scripts/test-runtime.ts +46 -0
- package/template/apps/agent-server/tsconfig.json +21 -0
- package/template/apps/server/Dockerfile +3 -0
- package/template/apps/server/package.json +5 -1
- package/template/apps/server/scripts/check-dsh-platform-example.ts +251 -0
- package/template/apps/server/src/agent/dsh-runtime-wire.test.ts +33 -0
- package/template/apps/server/src/agent/governance-runtime.ts +41 -0
- package/template/apps/server/src/agent/remote-runtime-response.ts +1 -1
- package/template/apps/server/src/agent/run-contracts.ts +1 -0
- package/template/apps/server/src/agent/run-definition-support.ts +9 -0
- package/template/apps/server/src/agent/runtime-tool-composition.ts +2 -0
- package/template/apps/server/src/agent/workflow-runtime-bootstrap.ts +9 -0
- package/template/apps/server/src/agent-skills/access.ts +88 -0
- package/template/apps/server/src/agent-skills/bootstrap.ts +49 -0
- package/template/apps/server/src/agent-skills/builtin-registration.ts +83 -0
- package/template/apps/server/src/agent-skills/builtins.ts +19 -0
- package/template/apps/server/src/agent-skills/bundle-zip.ts +163 -0
- package/template/apps/server/src/agent-skills/bundle.test.ts +79 -0
- package/template/apps/server/src/agent-skills/bundle.ts +133 -0
- package/template/apps/server/src/agent-skills/contracts.ts +46 -0
- package/template/apps/server/src/agent-skills/drizzle-repository.ts +125 -0
- package/template/apps/server/src/agent-skills/input.ts +29 -0
- package/template/apps/server/src/agent-skills/postgres.integration.test.ts +348 -0
- package/template/apps/server/src/agent-skills/read-routes.test.ts +116 -0
- package/template/apps/server/src/agent-skills/reader.ts +108 -0
- package/template/apps/server/src/agent-skills/repository.ts +105 -0
- package/template/apps/server/src/agent-skills/routes.ts +176 -0
- package/template/apps/server/src/agent-skills/service.test.ts +295 -0
- package/template/apps/server/src/agent-skills/service.ts +301 -0
- package/template/apps/server/src/agent-workflow/preset-registry.test.ts +137 -0
- package/template/apps/server/src/agent-workflow/preset-registry.ts +191 -0
- package/template/apps/server/src/agent-workflow/preset-routes.test.ts +94 -0
- package/template/apps/server/src/agent-workflow/preset-routes.ts +45 -0
- package/template/apps/server/src/ai-runtime-report/service.test.ts +10 -7
- package/template/apps/server/src/ai-runtime-report/service.ts +1 -1
- package/template/apps/server/src/app/agent-workflow-presets.ts +4 -0
- package/template/apps/server/src/app/dsh-skill-draft-audit.test.ts +195 -0
- package/template/apps/server/src/app/dsh-skill-draft-audit.ts +115 -0
- package/template/apps/server/src/app/dsh-workbench-audit.test.ts +71 -0
- package/template/apps/server/src/app/dsh-workbench-audit.ts +34 -0
- package/template/apps/server/src/app/dsh-workbench-handshake.test.ts +63 -0
- package/template/apps/server/src/app/dsh-workbench-handshake.ts +48 -0
- package/template/apps/server/src/app/dsh-workbench-http.test.ts +96 -0
- package/template/apps/server/src/app/dsh-workbench-lan-origin.test.ts +114 -0
- package/template/apps/server/src/app/dsh-workbench-launch.test.ts +70 -0
- package/template/apps/server/src/app/dsh-workbench-origin.test.ts +54 -0
- package/template/apps/server/src/app/dsh-workbench-proxy.ts +205 -0
- package/template/apps/server/src/app/dsh-workbench-session-conflict.test.ts +61 -0
- package/template/apps/server/src/app/dsh-workbench-session.ts +59 -0
- package/template/apps/server/src/app/dsh-workbench-socket-audit.test.ts +115 -0
- package/template/apps/server/src/app/dsh-workbench-socket.test.ts +246 -0
- package/template/apps/server/src/app/dsh-workbench-socket.ts +272 -0
- package/template/apps/server/src/app/dsh-workbench-test-support.ts +158 -0
- package/template/apps/server/src/app/dsh-workbench.test.ts +213 -0
- package/template/apps/server/src/app/dsh-workbench.ts +279 -0
- package/template/apps/server/src/app/workbench-public-origin.ts +16 -0
- package/template/apps/server/src/data-governance/user-plaintext-routes.test.ts +3 -0
- package/template/apps/server/src/http-app.ts +31 -0
- package/template/apps/server/src/mcp/adapter.ts +1 -0
- package/template/apps/server/src/module-composition.test.ts +23 -0
- package/template/apps/server/src/module-composition.ts +9 -2
- package/template/apps/server/src/observability/bootstrap.ts +12 -0
- package/template/apps/server/src/observability/diagnostic-failure-recorder.ts +94 -0
- package/template/apps/server/src/observability/diagnostic-http-contract.test.ts +42 -0
- package/template/apps/server/src/observability/diagnostic-index.test.ts +89 -0
- package/template/apps/server/src/observability/diagnostic-index.ts +88 -0
- package/template/apps/server/src/observability/diagnostic-repository.ts +67 -0
- package/template/apps/server/src/platform-app-bootstrap.ts +42 -0
- package/template/apps/server/src/platform-http-runtime.ts +6 -0
- package/template/apps/server/src/runtime-development.test.ts +25 -0
- package/template/apps/server/src/runtime-development.ts +6 -44
- package/template/apps/server/src/system/diagnostic-routes.test.ts +78 -0
- package/template/apps/server/src/system/diagnostic-routes.ts +66 -0
- package/template/apps/web/Dockerfile +9 -1
- package/template/apps/web/dev/start.ts +21 -10
- package/template/apps/web/dsh-workbench-location.ts +21 -0
- package/template/apps/web/dsh-workbench-proxy.test.ts +121 -0
- package/template/apps/web/dsh-workbench-proxy.ts +194 -0
- package/template/apps/web/package.json +3 -1
- package/template/apps/web/server.ts +27 -6
- package/template/apps/web/src/composables/useLoginFlow.ts +13 -1
- package/template/apps/web/src/pages/errors/PlatformErrorPage.vue +6 -0
- package/template/apps/web/src/pages/errors/PlatformErrorPage.webtest.ts +14 -1
- package/template/apps/web/src/pages/errors/error-page-extension.ts +1 -0
- package/template/apps/web/src/pages/system/AgentNativePage.vue +44 -0
- package/template/apps/web/src/pages/system/AgentWorkflowPage.vue +20 -7
- package/template/apps/web/src/pages/system/DiagnosticLookupPage.vue +56 -0
- package/template/apps/web/src/pages/system/DiagnosticLookupPage.webtest.ts +99 -0
- package/template/apps/web/src/pages/system/SystemAgentSkillsPage.vue +88 -0
- package/template/apps/web/src/pages/system/components/DiagnosticRecordDetails.vue +52 -0
- package/template/apps/web/src/pages/system/components/agent-skills/AgentSkillDetailDialog.vue +166 -0
- package/template/apps/web/src/pages/system/components/agent-skills/AgentSkillGovernanceActionDialog.vue +51 -0
- package/template/apps/web/src/pages/system/components/agent-skills/AgentSkillImportDialog.vue +320 -0
- package/template/apps/web/src/pages/system/components/agent-skills/AgentSkillImportDialog.webtest.ts +96 -0
- package/template/apps/web/src/pages/system/components/agent-skills/AgentSkillsTable.vue +69 -0
- package/template/apps/web/src/pages/system/components/agent-workflow/WorkflowPresetCatalog.vue +82 -0
- package/template/apps/web/src/pages/system/components/agent-workflow/WorkflowPresetCatalog.webtest.ts +57 -0
- package/template/apps/web/src/pages/system/composables/useAgentSkillsGovernance.ts +105 -0
- package/template/apps/web/src/pages/system/composables/useAgentWorkflowPresets.ts +44 -0
- package/template/apps/web/src/pages/system/composables/useDiagnosticLookup.ts +64 -0
- package/template/apps/web/src/router/error-navigation.ts +3 -1
- package/template/apps/web/src/router/error-navigation.webtest.ts +14 -2
- package/template/apps/web/src/router/index.ts +8 -49
- package/template/apps/web/src/router/platform-admin-pages.ts +79 -0
- package/template/apps/web/src/router/platform-admin-route-bindings.ts +8 -0
- package/template/apps/web/src/services/agent-skills-api.ts +55 -0
- package/template/apps/web/src/services/agent-workflow-api.ts +12 -0
- package/template/apps/web/src/services/diagnostics-api.ts +12 -0
- package/template/apps/web/src/services/http.ts +18 -0
- package/template/apps/web/src/services/http.webtest.ts +51 -0
- package/template/apps/web/vite.config.ts +39 -0
- package/template/deploy/dsh/README.md +5 -0
- package/template/deploy/dsh/documents/README.md +9 -0
- package/template/deploy/dsh/documents/requirements.in +6 -0
- package/template/deploy/dsh/documents/requirements.lock +72 -0
- package/template/docker-compose.yml +24 -10
- package/template/docs/adr/0004-embedded-dsh-runtime.md +9 -0
- package/template/docs/architecture/agent-admin-workbench-implementation.md +73 -0
- package/template/docs/architecture/agent-admin-workbench-plan.md +41 -0
- package/template/docs/architecture/agent-harness.md +2 -0
- package/template/docs/architecture/agent-runtime-compatibility.md +2 -0
- package/template/docs/architecture/embedded-dsh-integration.md +44 -0
- package/template/docs/development/dsh-plugins.md +55 -0
- package/template/docs/operations/dsh-agent-migration.md +98 -0
- package/template/docs/platform/agent-production-store.md +5 -78
- package/template/docs/platform/agent-runtime-observability.md +2 -0
- package/template/docs/platform/agent-runtime.md +2 -0
- package/template/docs/platform/agent-skills.md +25 -0
- package/template/package.json +4 -3
- package/template/packages/database/drizzle/0064_dsh_skills_diagnostics.sql +42 -0
- package/template/packages/database/drizzle/meta/0064_snapshot.json +15349 -0
- package/template/packages/database/drizzle/meta/_journal.json +7 -0
- package/template/packages/database/package.json +2 -2
- package/template/packages/database/src/schema/agent-skills.ts +65 -0
- package/template/packages/database/src/schema/diagnostics.ts +27 -0
- package/template/packages/database/src/schema/index.ts +8 -0
- package/template/packages/modules/package.json +1 -1
- package/template/packages/modules/src/system-admin-routes.ts +24 -0
- package/template/packages/modules/src/system-admin-ui-catalog.ts +17 -10
- package/template/packages/modules/src/system-agent-admin-bindings.ts +2 -0
- package/template/packages/modules/src/system-agent-admin-menus.ts +20 -0
- package/template/packages/modules/src/system-agent-audit-actions.ts +3 -0
- package/template/packages/modules/src/system-agent-dsh.ts +119 -0
- package/template/packages/modules/src/system-api-permissions.ts +9 -0
- package/template/packages/modules/src/system-audit-actions.ts +11 -0
- package/template/packages/modules/src/system-features.ts +5 -0
- package/template/packages/modules/src/system-module-audit-selection.ts +25 -0
- package/template/packages/modules/src/system-module-catalog.ts +12 -32
- package/template/packages/modules/src/system-modules.ts +14 -3
- package/template/packages/modules/src/system-permissions.ts +10 -0
- package/template/packages/server-sdk/package.json +2 -2
- package/template/packages/shared/index.ts +12 -0
- package/template/packages/shared/package.json +1 -1
- package/template/packages/shared/src/agent-skills.ts +68 -0
- package/template/packages/shared/src/agent-workflow-presets.ts +26 -0
- package/template/packages/shared/src/diagnostics.ts +13 -0
- package/template/packages/shared/src/rbac.ts +1 -0
- package/template/apps/agent-server/install-python-dependencies.sh +0 -92
- package/template/apps/agent-server/pyproject.toml +0 -51
- package/template/apps/agent-server/src/southwind_agent_server/__init__.py +0 -5
- package/template/apps/agent-server/src/southwind_agent_server/app.py +0 -344
- package/template/apps/agent-server/src/southwind_agent_server/approval_resolution.py +0 -30
- package/template/apps/agent-server/src/southwind_agent_server/artifact_input.py +0 -153
- package/template/apps/agent-server/src/southwind_agent_server/asgi.py +0 -8
- package/template/apps/agent-server/src/southwind_agent_server/attempt_context.py +0 -199
- package/template/apps/agent-server/src/southwind_agent_server/attempt_context_projection.py +0 -203
- package/template/apps/agent-server/src/southwind_agent_server/client_completion_contracts.py +0 -61
- package/template/apps/agent-server/src/southwind_agent_server/client_completion_governance.py +0 -82
- package/template/apps/agent-server/src/southwind_agent_server/client_completion_routes.py +0 -121
- package/template/apps/agent-server/src/southwind_agent_server/client_completion_runtime.py +0 -158
- package/template/apps/agent-server/src/southwind_agent_server/config.py +0 -267
- package/template/apps/agent-server/src/southwind_agent_server/context_message_digest.py +0 -24
- package/template/apps/agent-server/src/southwind_agent_server/context_version_contracts.py +0 -71
- package/template/apps/agent-server/src/southwind_agent_server/context_version_port.py +0 -307
- package/template/apps/agent-server/src/southwind_agent_server/contracts.py +0 -297
- package/template/apps/agent-server/src/southwind_agent_server/deterministic.py +0 -63
- package/template/apps/agent-server/src/southwind_agent_server/engine.py +0 -134
- package/template/apps/agent-server/src/southwind_agent_server/engine_capacity.py +0 -31
- package/template/apps/agent-server/src/southwind_agent_server/engine_identity.py +0 -2
- package/template/apps/agent-server/src/southwind_agent_server/fingerprint.py +0 -61
- package/template/apps/agent-server/src/southwind_agent_server/health_routes.py +0 -93
- package/template/apps/agent-server/src/southwind_agent_server/http_helpers.py +0 -62
- package/template/apps/agent-server/src/southwind_agent_server/lease.py +0 -15
- package/template/apps/agent-server/src/southwind_agent_server/multimodal.py +0 -76
- package/template/apps/agent-server/src/southwind_agent_server/openai_engine.py +0 -329
- package/template/apps/agent-server/src/southwind_agent_server/openai_engine_validation.py +0 -165
- package/template/apps/agent-server/src/southwind_agent_server/openai_provider.py +0 -345
- package/template/apps/agent-server/src/southwind_agent_server/postgres_report_insert.py +0 -37
- package/template/apps/agent-server/src/southwind_agent_server/postgres_report_queries.py +0 -235
- package/template/apps/agent-server/src/southwind_agent_server/postgres_report_replay.py +0 -44
- package/template/apps/agent-server/src/southwind_agent_server/postgres_report_status.py +0 -83
- package/template/apps/agent-server/src/southwind_agent_server/postgres_report_store.py +0 -336
- package/template/apps/agent-server/src/southwind_agent_server/postgres_report_summary.py +0 -304
- package/template/apps/agent-server/src/southwind_agent_server/postgres_store.py +0 -335
- package/template/apps/agent-server/src/southwind_agent_server/postgres_store_backend.py +0 -333
- package/template/apps/agent-server/src/southwind_agent_server/provider_audit.py +0 -144
- package/template/apps/agent-server/src/southwind_agent_server/provider_governance.py +0 -309
- package/template/apps/agent-server/src/southwind_agent_server/provider_output_guidance.py +0 -34
- package/template/apps/agent-server/src/southwind_agent_server/readiness_pin.py +0 -37
- package/template/apps/agent-server/src/southwind_agent_server/report_lifecycle.py +0 -218
- package/template/apps/agent-server/src/southwind_agent_server/report_metrics.py +0 -149
- package/template/apps/agent-server/src/southwind_agent_server/report_presenter.py +0 -300
- package/template/apps/agent-server/src/southwind_agent_server/report_routes.py +0 -132
- package/template/apps/agent-server/src/southwind_agent_server/report_store.py +0 -298
- package/template/apps/agent-server/src/southwind_agent_server/report_store_protocol.py +0 -75
- package/template/apps/agent-server/src/southwind_agent_server/report_types.py +0 -118
- package/template/apps/agent-server/src/southwind_agent_server/run_admission.py +0 -64
- package/template/apps/agent-server/src/southwind_agent_server/run_capacity.py +0 -59
- package/template/apps/agent-server/src/southwind_agent_server/runtime_budget.py +0 -74
- package/template/apps/agent-server/src/southwind_agent_server/runtime_compatibility.py +0 -89
- package/template/apps/agent-server/src/southwind_agent_server/runtime_factory.py +0 -105
- package/template/apps/agent-server/src/southwind_agent_server/runtime_logging.py +0 -57
- package/template/apps/agent-server/src/southwind_agent_server/service.py +0 -325
- package/template/apps/agent-server/src/southwind_agent_server/service_control.py +0 -64
- package/template/apps/agent-server/src/southwind_agent_server/service_errors.py +0 -4
- package/template/apps/agent-server/src/southwind_agent_server/service_execution.py +0 -34
- package/template/apps/agent-server/src/southwind_agent_server/service_tasks.py +0 -34
- package/template/apps/agent-server/src/southwind_agent_server/signal_events.py +0 -53
- package/template/apps/agent-server/src/southwind_agent_server/sqlite_catalog.py +0 -118
- package/template/apps/agent-server/src/southwind_agent_server/sqlite_errors.py +0 -2
- package/template/apps/agent-server/src/southwind_agent_server/sqlite_schema.py +0 -73
- package/template/apps/agent-server/src/southwind_agent_server/sqlite_serialization.py +0 -16
- package/template/apps/agent-server/src/southwind_agent_server/sqlite_store.py +0 -350
- package/template/apps/agent-server/src/southwind_agent_server/sse.py +0 -41
- package/template/apps/agent-server/src/southwind_agent_server/store.py +0 -255
- package/template/apps/agent-server/src/southwind_agent_server/store_cipher.py +0 -40
- package/template/apps/agent-server/src/southwind_agent_server/structured_logging.py +0 -145
- package/template/apps/agent-server/src/southwind_agent_server/telemetry.py +0 -93
- package/template/apps/agent-server/src/southwind_agent_server/tool_host.py +0 -152
- package/template/apps/agent-server/src/southwind_agent_server/trajectory.py +0 -137
- package/template/apps/agent-server/src/southwind_agent_server/unavailable_engine.py +0 -58
- package/template/apps/agent-server/src/southwind_agent_server/vision_probe.py +0 -87
- package/template/apps/agent-server/src/southwind_agent_server/vision_requirement.py +0 -28
- package/template/apps/agent-server/src/southwind_agent_server/vlm_readiness.py +0 -247
- package/template/apps/agent-server/tests/__init__.py +0 -1
- package/template/apps/agent-server/tests/openai_provider_test_support.py +0 -131
- package/template/apps/agent-server/tests/test_api.py +0 -305
- package/template/apps/agent-server/tests/test_artifact_input.py +0 -110
- package/template/apps/agent-server/tests/test_client_completion.py +0 -230
- package/template/apps/agent-server/tests/test_config.py +0 -106
- package/template/apps/agent-server/tests/test_fingerprint.py +0 -10
- package/template/apps/agent-server/tests/test_multimodal.py +0 -81
- package/template/apps/agent-server/tests/test_openai_provider.py +0 -320
- package/template/apps/agent-server/tests/test_postgres_report_status.py +0 -59
- package/template/apps/agent-server/tests/test_postgres_report_summary.py +0 -298
- package/template/apps/agent-server/tests/test_postgres_store.py +0 -135
- package/template/apps/agent-server/tests/test_provider_attempt_reporting.py +0 -140
- package/template/apps/agent-server/tests/test_provider_governance.py +0 -276
- package/template/apps/agent-server/tests/test_report_attempt_lifecycle.py +0 -126
- package/template/apps/agent-server/tests/test_report_resilience.py +0 -183
- package/template/apps/agent-server/tests/test_report_routes.py +0 -143
- package/template/apps/agent-server/tests/test_report_store.py +0 -347
- package/template/apps/agent-server/tests/test_report_summary_parity.py +0 -116
- package/template/apps/agent-server/tests/test_report_terminal_sync.py +0 -177
- package/template/apps/agent-server/tests/test_run_capacity.py +0 -112
- package/template/apps/agent-server/tests/test_run_operational_logging.py +0 -114
- package/template/apps/agent-server/tests/test_runtime_compatibility.py +0 -119
- package/template/apps/agent-server/tests/test_runtime_factory_readiness.py +0 -52
- package/template/apps/agent-server/tests/test_sqlite_store.py +0 -182
- package/template/apps/agent-server/tests/test_structured_logging.py +0 -142
- package/template/apps/agent-server/tests/test_telemetry.py +0 -25
- package/template/apps/agent-server/tests/test_tool_approval.py +0 -28
- package/template/apps/agent-server/tests/test_tool_structured_output.py +0 -347
- package/template/apps/agent-server/tests/test_trajectory.py +0 -91
- package/template/apps/agent-server/tests/test_vision_readiness_enforcement.py +0 -72
- package/template/apps/agent-server/tests/test_vlm_readiness.py +0 -106
- package/template/apps/agent-server/uv.lock +0 -831
|
@@ -0,0 +1,115 @@
|
|
|
1
|
+
import { randomUUID } from "node:crypto";
|
|
2
|
+
import { z } from "zod";
|
|
3
|
+
import { json, readJson, SafeError } from "./contracts.js";
|
|
4
|
+
import { canonicalDigest } from "./context-plugin.js";
|
|
5
|
+
import { terminal, type RuntimeStore } from "./store.js";
|
|
6
|
+
const sha = z.string().regex(/^sha256:[a-f0-9]{64}$/);
|
|
7
|
+
const evaluationSchema = z
|
|
8
|
+
.object({
|
|
9
|
+
runIds: z
|
|
10
|
+
.array(z.string().trim().min(1).max(160))
|
|
11
|
+
.min(1)
|
|
12
|
+
.max(100)
|
|
13
|
+
.refine((ids) => new Set(ids).size === ids.length),
|
|
14
|
+
evaluationSetKey: z.string().min(1).max(160),
|
|
15
|
+
evaluationSetVersion: z.string().min(1).max(80),
|
|
16
|
+
runnerVersion: z.string().min(1).max(80),
|
|
17
|
+
runtimePin: z
|
|
18
|
+
.object({
|
|
19
|
+
key: z.string().min(1).max(160),
|
|
20
|
+
identity: z.string().min(1).max(200),
|
|
21
|
+
releaseDigest: sha,
|
|
22
|
+
offersDigest: sha,
|
|
23
|
+
})
|
|
24
|
+
.strict(),
|
|
25
|
+
resultDigest: sha,
|
|
26
|
+
metrics: z
|
|
27
|
+
.record(z.string().min(1).max(160), z.number().finite())
|
|
28
|
+
.refine(
|
|
29
|
+
(value) =>
|
|
30
|
+
Object.keys(value).length >= 1 && Object.keys(value).length <= 100,
|
|
31
|
+
),
|
|
32
|
+
})
|
|
33
|
+
.strict();
|
|
34
|
+
/** Stores externally measured evidence against actual terminal DSH Runs. */
|
|
35
|
+
export class EvaluationPlugin {
|
|
36
|
+
constructor(private store: RuntimeStore) {
|
|
37
|
+
store.db.exec(`CREATE TABLE IF NOT EXISTS evaluations (
|
|
38
|
+
id TEXT PRIMARY KEY, release_id TEXT NOT NULL, dataset_key TEXT NOT NULL,
|
|
39
|
+
dataset_version TEXT NOT NULL, evidence_digest TEXT NOT NULL,
|
|
40
|
+
evidence TEXT NOT NULL, created_at TEXT NOT NULL,
|
|
41
|
+
UNIQUE(release_id,dataset_key,dataset_version));`);
|
|
42
|
+
store.db
|
|
43
|
+
.prepare("DELETE FROM evaluations WHERE created_at<?")
|
|
44
|
+
.run(new Date(Date.now() - 90 * 86400000).toISOString());
|
|
45
|
+
}
|
|
46
|
+
async handle(request: Request): Promise<Response> {
|
|
47
|
+
const body = evaluationSchema.parse(await readJson(request, 262144));
|
|
48
|
+
request.signal.throwIfAborted();
|
|
49
|
+
const runs = body.runIds.map((id) => this.store.get(id));
|
|
50
|
+
const releaseIds = new Set(
|
|
51
|
+
runs.map((run) => String(run.request.release.releaseId ?? "")),
|
|
52
|
+
),
|
|
53
|
+
operations = new Set(runs.map((run) => run.request.operationKey));
|
|
54
|
+
if (
|
|
55
|
+
runs.some((run) => !terminal(run.status)) ||
|
|
56
|
+
releaseIds.size !== 1 ||
|
|
57
|
+
operations.size !== 1 ||
|
|
58
|
+
new Set(runs.map((run) => canonicalDigest(run.request.release))).size !==
|
|
59
|
+
1
|
|
60
|
+
)
|
|
61
|
+
throw new SafeError(
|
|
62
|
+
"RUN_CONFLICT",
|
|
63
|
+
"评估必须绑定同一 Operation Release 的已结束 Run",
|
|
64
|
+
);
|
|
65
|
+
const releaseId = [...releaseIds][0]!;
|
|
66
|
+
if (!releaseId)
|
|
67
|
+
throw new SafeError(
|
|
68
|
+
"RUN_CONFLICT",
|
|
69
|
+
"评估绑定的 Release 不存在",
|
|
70
|
+
false,
|
|
71
|
+
404,
|
|
72
|
+
);
|
|
73
|
+
const evidence = {
|
|
74
|
+
schemaVersion: 1,
|
|
75
|
+
runIds: body.runIds,
|
|
76
|
+
operationKey: [...operations][0],
|
|
77
|
+
runnerVersion: body.runnerVersion,
|
|
78
|
+
runtimePin: body.runtimePin,
|
|
79
|
+
resultDigest: body.resultDigest,
|
|
80
|
+
metrics: body.metrics,
|
|
81
|
+
};
|
|
82
|
+
const evidenceDigest = canonicalDigest(evidence);
|
|
83
|
+
const existing = this.store.db
|
|
84
|
+
.prepare(
|
|
85
|
+
"SELECT id,evidence_digest FROM evaluations WHERE release_id=? AND dataset_key=? AND dataset_version=?",
|
|
86
|
+
)
|
|
87
|
+
.get(releaseId, body.evaluationSetKey, body.evaluationSetVersion) as
|
|
88
|
+
| { id: string; evidence_digest: string }
|
|
89
|
+
| undefined;
|
|
90
|
+
if (existing) {
|
|
91
|
+
if (existing.evidence_digest !== evidenceDigest)
|
|
92
|
+
throw new SafeError("RUN_CONFLICT", "评估版本已经绑定其他结果");
|
|
93
|
+
return json({ evaluationId: existing.id, releaseId });
|
|
94
|
+
}
|
|
95
|
+
const count = this.store.db
|
|
96
|
+
.prepare("SELECT COUNT(*) AS count FROM evaluations")
|
|
97
|
+
.get() as { count: number };
|
|
98
|
+
if (count.count >= 10000)
|
|
99
|
+
throw new SafeError("BUDGET_EXCEEDED", "评估留存容量已满", true, 429);
|
|
100
|
+
request.signal.throwIfAborted();
|
|
101
|
+
const evaluationId = `eval_${randomUUID().replaceAll("-", "")}`;
|
|
102
|
+
this.store.db
|
|
103
|
+
.prepare("INSERT INTO evaluations VALUES(?,?,?,?,?,?,?)")
|
|
104
|
+
.run(
|
|
105
|
+
evaluationId,
|
|
106
|
+
releaseId,
|
|
107
|
+
body.evaluationSetKey,
|
|
108
|
+
body.evaluationSetVersion,
|
|
109
|
+
evidenceDigest,
|
|
110
|
+
JSON.stringify(evidence),
|
|
111
|
+
new Date().toISOString(),
|
|
112
|
+
);
|
|
113
|
+
return json({ evaluationId, releaseId });
|
|
114
|
+
}
|
|
115
|
+
}
|
|
@@ -0,0 +1,194 @@
|
|
|
1
|
+
import { test } from "node:test";
|
|
2
|
+
import assert from "node:assert/strict";
|
|
3
|
+
import { mkdtempSync, rmSync } from "node:fs";
|
|
4
|
+
import { join } from "node:path";
|
|
5
|
+
import { tmpdir } from "node:os";
|
|
6
|
+
import { runtime, request, input } from "./runtime-test-fixtures.js";
|
|
7
|
+
import { validateStructuredOutput } from "./output-validation.js";
|
|
8
|
+
function evidence(runIds: string[]) {
|
|
9
|
+
return {
|
|
10
|
+
runIds,
|
|
11
|
+
evaluationSetKey: "analysis.eval.copilot-general",
|
|
12
|
+
evaluationSetVersion: "1",
|
|
13
|
+
runnerVersion: "copilot-eval-test",
|
|
14
|
+
runtimePin: {
|
|
15
|
+
key: "test-runtime",
|
|
16
|
+
identity: "test-deployment",
|
|
17
|
+
releaseDigest: `sha256:${"a".repeat(64)}`,
|
|
18
|
+
offersDigest: `sha256:${"b".repeat(64)}`,
|
|
19
|
+
},
|
|
20
|
+
resultDigest: `sha256:${"c".repeat(64)}`,
|
|
21
|
+
metrics: { passed: 1, scenarioCount: runIds.length, assertionRate: 1 },
|
|
22
|
+
};
|
|
23
|
+
}
|
|
24
|
+
async function completed(
|
|
25
|
+
handler: ReturnType<typeof runtime>,
|
|
26
|
+
key: string,
|
|
27
|
+
operationKey = "test.operation",
|
|
28
|
+
) {
|
|
29
|
+
const body = input();
|
|
30
|
+
body.operationKey = operationKey;
|
|
31
|
+
const start = await handler(request("/api/agent/runs", "POST", body, key));
|
|
32
|
+
assert.equal(start?.status, 200);
|
|
33
|
+
const run = (await start!.json()) as { runId: string; eventsUrl: string };
|
|
34
|
+
const events = await handler(request(run.eventsUrl));
|
|
35
|
+
assert.match(await events!.text(), /event: run.completed/);
|
|
36
|
+
return run;
|
|
37
|
+
}
|
|
38
|
+
test("evaluation records actual DSH runs and persists idempotent external metrics across restart", async () => {
|
|
39
|
+
const directory = mkdtempSync(join(tmpdir(), "dsh-evaluation-")),
|
|
40
|
+
database = join(directory, "runtime.sqlite");
|
|
41
|
+
let handler = runtime({ AGENT_RUNTIME_DATABASE_PATH: database });
|
|
42
|
+
try {
|
|
43
|
+
const run = await completed(handler, "evaluation-run"),
|
|
44
|
+
body = evidence([run.runId]);
|
|
45
|
+
const response = await handler(
|
|
46
|
+
request("/api/agent/evaluations", "POST", body),
|
|
47
|
+
);
|
|
48
|
+
assert.equal(response?.status, 200);
|
|
49
|
+
const recorded = await response!.json();
|
|
50
|
+
assert.match((recorded as { evaluationId: string }).evaluationId, /^eval_/);
|
|
51
|
+
assert.equal((recorded as { releaseId: string }).releaseId, "test-release");
|
|
52
|
+
await handler.close();
|
|
53
|
+
handler = runtime({ AGENT_RUNTIME_DATABASE_PATH: database });
|
|
54
|
+
const replay = await handler(
|
|
55
|
+
request("/api/agent/evaluations", "POST", body),
|
|
56
|
+
);
|
|
57
|
+
assert.deepEqual(await replay!.json(), recorded);
|
|
58
|
+
const conflict = await handler(
|
|
59
|
+
request("/api/agent/evaluations", "POST", {
|
|
60
|
+
...body,
|
|
61
|
+
metrics: { passed: 0 },
|
|
62
|
+
}),
|
|
63
|
+
);
|
|
64
|
+
assert.equal(conflict?.status, 409);
|
|
65
|
+
} finally {
|
|
66
|
+
await handler.close();
|
|
67
|
+
rmSync(directory, { recursive: true });
|
|
68
|
+
}
|
|
69
|
+
});
|
|
70
|
+
test("evaluation enforces auth, terminal state, binding consistency and bounded request shape", async () => {
|
|
71
|
+
const handler = runtime();
|
|
72
|
+
try {
|
|
73
|
+
const body = input();
|
|
74
|
+
body.userPrompt = "hold-model";
|
|
75
|
+
const start = await handler(
|
|
76
|
+
request("/api/agent/runs", "POST", body, "pending-evaluation"),
|
|
77
|
+
);
|
|
78
|
+
const pending = (await start!.json()) as {
|
|
79
|
+
runId: string;
|
|
80
|
+
eventsUrl: string;
|
|
81
|
+
};
|
|
82
|
+
const inProgress = await handler(
|
|
83
|
+
request("/api/agent/evaluations", "POST", evidence([pending.runId])),
|
|
84
|
+
);
|
|
85
|
+
assert.equal(inProgress?.status, 409);
|
|
86
|
+
await handler(
|
|
87
|
+
request(`/api/agent/runs/${pending.runId}/cancel`, "POST", {
|
|
88
|
+
reason: "evaluation cancelled",
|
|
89
|
+
}),
|
|
90
|
+
);
|
|
91
|
+
const events = await handler(request(pending.eventsUrl));
|
|
92
|
+
assert.match(await events!.text(), /event: run.cancelled/);
|
|
93
|
+
const cancelled = await handler(
|
|
94
|
+
request("/api/agent/evaluations", "POST", {
|
|
95
|
+
...evidence([pending.runId]),
|
|
96
|
+
metrics: { passed: 0, cancelled: 1 },
|
|
97
|
+
}),
|
|
98
|
+
);
|
|
99
|
+
assert.equal(cancelled?.status, 200);
|
|
100
|
+
const forbidden = await handler(
|
|
101
|
+
new Request("http://runtime/api/agent/evaluations", {
|
|
102
|
+
method: "POST",
|
|
103
|
+
body: JSON.stringify(evidence([pending.runId])),
|
|
104
|
+
}),
|
|
105
|
+
);
|
|
106
|
+
assert.equal(forbidden?.status, 403);
|
|
107
|
+
const unknown = await handler(
|
|
108
|
+
request("/api/agent/evaluations", "POST", evidence(["run_missing"])),
|
|
109
|
+
);
|
|
110
|
+
assert.equal(unknown?.status, 404);
|
|
111
|
+
const excessive = await handler(
|
|
112
|
+
request(
|
|
113
|
+
"/api/agent/evaluations",
|
|
114
|
+
"POST",
|
|
115
|
+
evidence(Array.from({ length: 101 }, (_, index) => `run_${index}`)),
|
|
116
|
+
),
|
|
117
|
+
);
|
|
118
|
+
assert.equal(excessive?.status, 400);
|
|
119
|
+
const duplicate = await handler(
|
|
120
|
+
request(
|
|
121
|
+
"/api/agent/evaluations",
|
|
122
|
+
"POST",
|
|
123
|
+
evidence([pending.runId, pending.runId]),
|
|
124
|
+
),
|
|
125
|
+
);
|
|
126
|
+
assert.equal(duplicate?.status, 400);
|
|
127
|
+
const other = await completed(
|
|
128
|
+
handler,
|
|
129
|
+
"other-operation",
|
|
130
|
+
"test.other-operation",
|
|
131
|
+
);
|
|
132
|
+
const mixed = await handler(
|
|
133
|
+
request(
|
|
134
|
+
"/api/agent/evaluations",
|
|
135
|
+
"POST",
|
|
136
|
+
evidence([pending.runId, other.runId]),
|
|
137
|
+
),
|
|
138
|
+
);
|
|
139
|
+
assert.equal(mixed?.status, 409);
|
|
140
|
+
} finally {
|
|
141
|
+
await handler.close();
|
|
142
|
+
}
|
|
143
|
+
});
|
|
144
|
+
test("structured-output diagnostics preserve safe schema kind and path without provider content", () => {
|
|
145
|
+
const logs: string[] = [],
|
|
146
|
+
context = {
|
|
147
|
+
runId: "run_test",
|
|
148
|
+
operationKey: "analysis.copilot.global",
|
|
149
|
+
outputSchemaKey: "copilot.output.v1",
|
|
150
|
+
};
|
|
151
|
+
assert.throws(
|
|
152
|
+
() =>
|
|
153
|
+
validateStructuredOutput(
|
|
154
|
+
"private unstructured model content",
|
|
155
|
+
{ type: "object" },
|
|
156
|
+
context,
|
|
157
|
+
(value) => logs.push(value),
|
|
158
|
+
),
|
|
159
|
+
/格式校验/,
|
|
160
|
+
);
|
|
161
|
+
assert.equal(JSON.parse(logs[0]!).validationKind, "json-syntax");
|
|
162
|
+
assert.throws(
|
|
163
|
+
() =>
|
|
164
|
+
validateStructuredOutput(
|
|
165
|
+
'{"answer":123}',
|
|
166
|
+
{ type: "object", properties: { answer: { type: "string" } } },
|
|
167
|
+
context,
|
|
168
|
+
(value) => logs.push(value),
|
|
169
|
+
),
|
|
170
|
+
/格式校验/,
|
|
171
|
+
);
|
|
172
|
+
assert.deepEqual(
|
|
173
|
+
{
|
|
174
|
+
kind: JSON.parse(logs[1]!).validationKind,
|
|
175
|
+
path: JSON.parse(logs[1]!).instancePath,
|
|
176
|
+
},
|
|
177
|
+
{ kind: "schema-instance", path: "/answer" },
|
|
178
|
+
);
|
|
179
|
+
assert.throws(
|
|
180
|
+
() =>
|
|
181
|
+
validateStructuredOutput(
|
|
182
|
+
"{}",
|
|
183
|
+
{ type: "not-a-schema-type" },
|
|
184
|
+
context,
|
|
185
|
+
(value) => logs.push(value),
|
|
186
|
+
),
|
|
187
|
+
/格式校验/,
|
|
188
|
+
);
|
|
189
|
+
assert.equal(JSON.parse(logs[2]!).validationKind, "schema-definition");
|
|
190
|
+
assert.doesNotMatch(
|
|
191
|
+
logs.join("\n"),
|
|
192
|
+
/private unstructured model content|"answer":123/,
|
|
193
|
+
);
|
|
194
|
+
});
|
|
@@ -0,0 +1,71 @@
|
|
|
1
|
+
import { SafeError } from "./contracts.js";
|
|
2
|
+
import { terminal, type RuntimeStore } from "./store.js";
|
|
3
|
+
export function publicEvent(event: Record<string, unknown>) {
|
|
4
|
+
const {
|
|
5
|
+
attestation: _attestation,
|
|
6
|
+
sourcePreview: _preview,
|
|
7
|
+
sourceSnapshotDigest: _digest,
|
|
8
|
+
...value
|
|
9
|
+
} = event;
|
|
10
|
+
return value;
|
|
11
|
+
}
|
|
12
|
+
export function eventResponse(
|
|
13
|
+
request: Request,
|
|
14
|
+
id: string,
|
|
15
|
+
store: RuntimeStore,
|
|
16
|
+
) {
|
|
17
|
+
const url = new URL(request.url),
|
|
18
|
+
query = url.searchParams.get("after") ?? "0",
|
|
19
|
+
header = request.headers.get("Last-Event-ID") ?? "0";
|
|
20
|
+
if (!/^\d+$/.test(query) || !/^\d+$/.test(header))
|
|
21
|
+
throw new SafeError("INVALID_INPUT", "事件游标无效", false, 400);
|
|
22
|
+
let after = Math.max(Number(query), Number(header));
|
|
23
|
+
if (!Number.isSafeInteger(after))
|
|
24
|
+
throw new SafeError("INVALID_INPUT", "事件游标无效", false, 400);
|
|
25
|
+
const encoder = new TextEncoder();
|
|
26
|
+
let cleanup: (() => void) | undefined;
|
|
27
|
+
let cancelled = false;
|
|
28
|
+
const body = new ReadableStream<Uint8Array>({
|
|
29
|
+
start(sink) {
|
|
30
|
+
let done = false;
|
|
31
|
+
const close = () => {
|
|
32
|
+
if (done) return;
|
|
33
|
+
done = true;
|
|
34
|
+
unsubscribe();
|
|
35
|
+
clearInterval(heartbeat);
|
|
36
|
+
request.signal.removeEventListener("abort", close);
|
|
37
|
+
if (!cancelled) sink.close();
|
|
38
|
+
};
|
|
39
|
+
const flush = () => {
|
|
40
|
+
if (done) return;
|
|
41
|
+
for (const event of store.events(id, after)) {
|
|
42
|
+
after = event.sequence;
|
|
43
|
+
sink.enqueue(
|
|
44
|
+
encoder.encode(
|
|
45
|
+
`id: ${event.sequence}\nevent: ${event.type}\ndata: ${JSON.stringify(publicEvent(event))}\n\n`,
|
|
46
|
+
),
|
|
47
|
+
);
|
|
48
|
+
}
|
|
49
|
+
if (terminal(store.get(id).status)) close();
|
|
50
|
+
};
|
|
51
|
+
const unsubscribe = store.subscribe(id, flush),
|
|
52
|
+
heartbeat = setInterval(() => {
|
|
53
|
+
if (!done) sink.enqueue(encoder.encode(": heartbeat\n\n"));
|
|
54
|
+
}, 15000);
|
|
55
|
+
cleanup = close;
|
|
56
|
+
request.signal.addEventListener("abort", close, { once: true });
|
|
57
|
+
flush();
|
|
58
|
+
},
|
|
59
|
+
cancel() {
|
|
60
|
+
cancelled = true;
|
|
61
|
+
cleanup?.();
|
|
62
|
+
},
|
|
63
|
+
});
|
|
64
|
+
return new Response(body, {
|
|
65
|
+
headers: {
|
|
66
|
+
"content-type": "text/event-stream",
|
|
67
|
+
"cache-control": "no-cache, no-transform",
|
|
68
|
+
"x-accel-buffering": "no",
|
|
69
|
+
},
|
|
70
|
+
});
|
|
71
|
+
}
|
|
@@ -0,0 +1,194 @@
|
|
|
1
|
+
import { SafeError, type ProviderPolicy } from "./contracts.js";
|
|
2
|
+
import type { RuntimeStore } from "./store.js";
|
|
3
|
+
const detectors = [
|
|
4
|
+
/\bBearer\s+[A-Za-z0-9._~+/=-]{8,}/gi,
|
|
5
|
+
/\b(?:api[_-]?key|secret|password|token)\s*[:=]\s*\S+/gi,
|
|
6
|
+
/\b1[3-9]\d{9}\b/g,
|
|
7
|
+
/\b[A-Z0-9._%+-]+@[A-Z0-9.-]+\.[A-Z]{2,}\b/gi,
|
|
8
|
+
/\b\d{17}[\dXx]\b/g,
|
|
9
|
+
/\b(?:\d[ -]?){16,30}\b/g,
|
|
10
|
+
];
|
|
11
|
+
export function governedValue(
|
|
12
|
+
value: unknown,
|
|
13
|
+
action: "redact" | "reject",
|
|
14
|
+
): unknown {
|
|
15
|
+
if (typeof value === "string") {
|
|
16
|
+
let result = value;
|
|
17
|
+
for (const pattern of detectors) {
|
|
18
|
+
pattern.lastIndex = 0;
|
|
19
|
+
if (pattern.test(result)) {
|
|
20
|
+
if (action === "reject")
|
|
21
|
+
throw new SafeError(
|
|
22
|
+
"FORBIDDEN",
|
|
23
|
+
"Provider 请求被字段治理策略拒绝",
|
|
24
|
+
false,
|
|
25
|
+
403,
|
|
26
|
+
);
|
|
27
|
+
pattern.lastIndex = 0;
|
|
28
|
+
result = result.replace(pattern, "[REDACTED]");
|
|
29
|
+
}
|
|
30
|
+
}
|
|
31
|
+
return result;
|
|
32
|
+
}
|
|
33
|
+
if (Array.isArray(value))
|
|
34
|
+
return value.map((item) => governedValue(item, action));
|
|
35
|
+
if (value && typeof value === "object")
|
|
36
|
+
return Object.fromEntries(
|
|
37
|
+
Object.entries(value).map(([key, item]) => [
|
|
38
|
+
key,
|
|
39
|
+
governedValue(item, action),
|
|
40
|
+
]),
|
|
41
|
+
);
|
|
42
|
+
return value;
|
|
43
|
+
}
|
|
44
|
+
export function governProvider(
|
|
45
|
+
messages: Record<string, unknown>[],
|
|
46
|
+
tools: Record<string, unknown>[],
|
|
47
|
+
outputSchema: Record<string, unknown> | null | undefined,
|
|
48
|
+
policy: ProviderPolicy,
|
|
49
|
+
audit: RuntimeStore,
|
|
50
|
+
metadata: Record<string, unknown>,
|
|
51
|
+
) {
|
|
52
|
+
const rules = new Map(
|
|
53
|
+
policy.envelopeFields.map((rule) => [rule.field, rule]),
|
|
54
|
+
);
|
|
55
|
+
let redacted = false;
|
|
56
|
+
const classified = new Set<string>();
|
|
57
|
+
function apply(field: string, value: unknown): unknown {
|
|
58
|
+
const rule = rules.get(field);
|
|
59
|
+
if (!rule)
|
|
60
|
+
throw new SafeError(
|
|
61
|
+
"FORBIDDEN",
|
|
62
|
+
`Provider 未授权发送字段:${field}`,
|
|
63
|
+
false,
|
|
64
|
+
403,
|
|
65
|
+
);
|
|
66
|
+
classified.add(rule.classification);
|
|
67
|
+
const governed = governedValue(value, rule.dlpAction);
|
|
68
|
+
redacted ||= JSON.stringify(governed) !== JSON.stringify(value);
|
|
69
|
+
return governed;
|
|
70
|
+
}
|
|
71
|
+
try {
|
|
72
|
+
if (rules.size !== policy.envelopeFields.length)
|
|
73
|
+
throw new SafeError("FORBIDDEN", "Provider 字段规则重复", false, 403);
|
|
74
|
+
const roles: Record<string, string> = {
|
|
75
|
+
system: "instructions",
|
|
76
|
+
developer: "instructions",
|
|
77
|
+
user: "user-prompt",
|
|
78
|
+
assistant: "provider-output",
|
|
79
|
+
tool: "tool-results",
|
|
80
|
+
};
|
|
81
|
+
const governedMessages = messages.map((message) => {
|
|
82
|
+
const field = roles[String(message.role)];
|
|
83
|
+
if (!field)
|
|
84
|
+
throw new SafeError("FORBIDDEN", "Provider 消息角色无效", false, 403);
|
|
85
|
+
if (Array.isArray(message.content)) {
|
|
86
|
+
const content = message.content.map((raw: unknown) => {
|
|
87
|
+
const part = raw as Record<string, unknown>;
|
|
88
|
+
if (part.type === "text") return apply(field, part);
|
|
89
|
+
const image = part.image_url as Record<string, unknown> | undefined;
|
|
90
|
+
if (
|
|
91
|
+
part.type !== "image_url" ||
|
|
92
|
+
typeof image?.url !== "string" ||
|
|
93
|
+
!/^data:image\/(png|jpeg|webp);base64,[A-Za-z0-9+/=]+$/.test(
|
|
94
|
+
image.url,
|
|
95
|
+
)
|
|
96
|
+
)
|
|
97
|
+
throw new SafeError("FORBIDDEN", "图片输入结构无效", false, 403);
|
|
98
|
+
const rule = rules.get("artifact-content");
|
|
99
|
+
if (!rule)
|
|
100
|
+
throw new SafeError(
|
|
101
|
+
"FORBIDDEN",
|
|
102
|
+
"Provider 未授权发送图片",
|
|
103
|
+
false,
|
|
104
|
+
403,
|
|
105
|
+
);
|
|
106
|
+
classified.add(rule.classification);
|
|
107
|
+
return part;
|
|
108
|
+
});
|
|
109
|
+
const { content: _content, ...metadata } = message;
|
|
110
|
+
return {
|
|
111
|
+
...(apply(field, metadata) as Record<string, unknown>),
|
|
112
|
+
content,
|
|
113
|
+
};
|
|
114
|
+
}
|
|
115
|
+
return apply(field, message) as Record<string, unknown>;
|
|
116
|
+
});
|
|
117
|
+
const governedTools = tools.length
|
|
118
|
+
? (apply("tool-definitions", tools) as Record<string, unknown>[])
|
|
119
|
+
: [];
|
|
120
|
+
const schema = outputSchema
|
|
121
|
+
? (apply("output-schema", outputSchema) as Record<string, unknown>)
|
|
122
|
+
: undefined;
|
|
123
|
+
const bytes = Buffer.byteLength(
|
|
124
|
+
JSON.stringify({
|
|
125
|
+
messages: governedMessages,
|
|
126
|
+
tools: governedTools,
|
|
127
|
+
schema,
|
|
128
|
+
}),
|
|
129
|
+
);
|
|
130
|
+
audit.audit({
|
|
131
|
+
...metadata,
|
|
132
|
+
classified: [...classified].sort(),
|
|
133
|
+
sentBytes: bytes,
|
|
134
|
+
decision: redacted ? "redact" : "allow",
|
|
135
|
+
});
|
|
136
|
+
return { messages: governedMessages, tools: governedTools, schema, bytes };
|
|
137
|
+
} catch (error) {
|
|
138
|
+
audit.audit({
|
|
139
|
+
...metadata,
|
|
140
|
+
classified: [...classified].sort(),
|
|
141
|
+
sentBytes: 0,
|
|
142
|
+
decision: "reject",
|
|
143
|
+
});
|
|
144
|
+
throw error;
|
|
145
|
+
}
|
|
146
|
+
}
|
|
147
|
+
export async function* sseJson(
|
|
148
|
+
response: Response,
|
|
149
|
+
byteLimit: number,
|
|
150
|
+
signal?: AbortSignal,
|
|
151
|
+
): AsyncGenerator<Record<string, unknown>> {
|
|
152
|
+
if (!response.body)
|
|
153
|
+
throw new SafeError("PROVIDER_UNAVAILABLE", "模型响应为空", true, 503);
|
|
154
|
+
const reader = response.body.getReader();
|
|
155
|
+
const abort = () => {
|
|
156
|
+
void reader.cancel().catch(() => {});
|
|
157
|
+
};
|
|
158
|
+
signal?.addEventListener("abort", abort, { once: true });
|
|
159
|
+
if (signal?.aborted) abort();
|
|
160
|
+
const decoder = new TextDecoder();
|
|
161
|
+
let buffer = "",
|
|
162
|
+
bytes = 0;
|
|
163
|
+
try {
|
|
164
|
+
for (;;) {
|
|
165
|
+
const next = await reader.read();
|
|
166
|
+
if (next.done) break;
|
|
167
|
+
bytes += next.value.byteLength;
|
|
168
|
+
if (bytes > byteLimit)
|
|
169
|
+
throw new SafeError("BUDGET_EXCEEDED", "模型响应超过上限");
|
|
170
|
+
buffer += decoder.decode(next.value, { stream: true });
|
|
171
|
+
let lineEnd: number;
|
|
172
|
+
while ((lineEnd = buffer.indexOf("\n")) >= 0) {
|
|
173
|
+
const line = buffer.slice(0, lineEnd).trim();
|
|
174
|
+
buffer = buffer.slice(lineEnd + 1);
|
|
175
|
+
if (!line.startsWith("data:")) continue;
|
|
176
|
+
const value = line.slice(5).trim();
|
|
177
|
+
if (value === "[DONE]") return;
|
|
178
|
+
if (value) {
|
|
179
|
+
const parsed: unknown = JSON.parse(value);
|
|
180
|
+
if (!parsed || typeof parsed !== "object" || Array.isArray(parsed))
|
|
181
|
+
throw new SafeError(
|
|
182
|
+
"PROVIDER_UNAVAILABLE",
|
|
183
|
+
"模型返回协议无效",
|
|
184
|
+
true,
|
|
185
|
+
);
|
|
186
|
+
yield parsed as Record<string, unknown>;
|
|
187
|
+
}
|
|
188
|
+
}
|
|
189
|
+
}
|
|
190
|
+
} finally {
|
|
191
|
+
signal?.removeEventListener("abort", abort);
|
|
192
|
+
await reader.cancel();
|
|
193
|
+
}
|
|
194
|
+
}
|