onto-mcp 0.3.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/.onto/authority/core-lens-registry.yaml +134 -0
- package/.onto/authority/core-lexicon.yaml +1562 -0
- package/.onto/authority/diagnostic-codes.yaml +94 -0
- package/.onto/domains/accounting/competency_qs.md +384 -0
- package/.onto/domains/accounting/concepts.md +186 -0
- package/.onto/domains/accounting/conciseness_rules.md +160 -0
- package/.onto/domains/accounting/dependency_rules.md +239 -0
- package/.onto/domains/accounting/domain_scope.md +213 -0
- package/.onto/domains/accounting/extension_cases.md +416 -0
- package/.onto/domains/accounting/logic_rules.md +226 -0
- package/.onto/domains/accounting/structure_spec.md +298 -0
- package/.onto/domains/accounting-kr/competency_qs.md +562 -0
- package/.onto/domains/accounting-kr/concepts.md +187 -0
- package/.onto/domains/accounting-kr/conciseness_rules.md +125 -0
- package/.onto/domains/accounting-kr/dependency_rules.md +93 -0
- package/.onto/domains/accounting-kr/domain_scope.md +140 -0
- package/.onto/domains/accounting-kr/extension_cases.md +343 -0
- package/.onto/domains/accounting-kr/logic_rules.md +160 -0
- package/.onto/domains/accounting-kr/structure_spec.md +85 -0
- package/.onto/domains/business/competency_qs.md +263 -0
- package/.onto/domains/business/concepts.md +200 -0
- package/.onto/domains/business/conciseness_rules.md +135 -0
- package/.onto/domains/business/dependency_rules.md +113 -0
- package/.onto/domains/business/domain_scope.md +240 -0
- package/.onto/domains/business/extension_cases.md +249 -0
- package/.onto/domains/business/logic_rules.md +134 -0
- package/.onto/domains/business/structure_spec.md +114 -0
- package/.onto/domains/finance/competency_qs.md +362 -0
- package/.onto/domains/finance/concepts.md +194 -0
- package/.onto/domains/finance/conciseness_rules.md +155 -0
- package/.onto/domains/finance/dependency_rules.md +171 -0
- package/.onto/domains/finance/domain_scope.md +215 -0
- package/.onto/domains/finance/extension_cases.md +350 -0
- package/.onto/domains/finance/logic_rules.md +191 -0
- package/.onto/domains/finance/structure_spec.md +182 -0
- package/.onto/domains/llm-native-development/competency_qs.md +430 -0
- package/.onto/domains/llm-native-development/concepts.md +242 -0
- package/.onto/domains/llm-native-development/conciseness_rules.md +163 -0
- package/.onto/domains/llm-native-development/dependency_rules.md +216 -0
- package/.onto/domains/llm-native-development/domain_scope.md +197 -0
- package/.onto/domains/llm-native-development/extension_cases.md +474 -0
- package/.onto/domains/llm-native-development/logic_rules.md +123 -0
- package/.onto/domains/llm-native-development/prompt_interface.md +49 -0
- package/.onto/domains/llm-native-development/structure_spec.md +245 -0
- package/.onto/domains/market-intelligence/competency_qs.md +274 -0
- package/.onto/domains/market-intelligence/concepts.md +233 -0
- package/.onto/domains/market-intelligence/conciseness_rules.md +165 -0
- package/.onto/domains/market-intelligence/dependency_rules.md +197 -0
- package/.onto/domains/market-intelligence/domain_scope.md +231 -0
- package/.onto/domains/market-intelligence/extension_cases.md +425 -0
- package/.onto/domains/market-intelligence/logic_rules.md +247 -0
- package/.onto/domains/market-intelligence/structure_spec.md +209 -0
- package/.onto/domains/ontology/competency_qs.md +394 -0
- package/.onto/domains/ontology/concepts.md +172 -0
- package/.onto/domains/ontology/conciseness_rules.md +134 -0
- package/.onto/domains/ontology/dependency_rules.md +125 -0
- package/.onto/domains/ontology/domain_scope.md +114 -0
- package/.onto/domains/ontology/extension_cases.md +501 -0
- package/.onto/domains/ontology/logic_rules.md +114 -0
- package/.onto/domains/ontology/problem_framing_profile.md +67 -0
- package/.onto/domains/ontology/structure_spec.md +115 -0
- package/.onto/domains/palantir-foundry/RESEARCH_NOTES.md +911 -0
- package/.onto/domains/palantir-foundry/competency_qs.md +191 -0
- package/.onto/domains/palantir-foundry/competitive_comparison.md +329 -0
- package/.onto/domains/palantir-foundry/concepts.md +197 -0
- package/.onto/domains/palantir-foundry/conciseness_rules.md +245 -0
- package/.onto/domains/palantir-foundry/dependency_rules.md +135 -0
- package/.onto/domains/palantir-foundry/domain_scope.md +395 -0
- package/.onto/domains/palantir-foundry/extension_cases.md +210 -0
- package/.onto/domains/palantir-foundry/logic_rules.md +172 -0
- package/.onto/domains/palantir-foundry/structure_spec.md +291 -0
- package/.onto/domains/software-engineering/competency_qs.md +538 -0
- package/.onto/domains/software-engineering/concepts.md +238 -0
- package/.onto/domains/software-engineering/conciseness_rules.md +167 -0
- package/.onto/domains/software-engineering/dependency_rules.md +216 -0
- package/.onto/domains/software-engineering/domain_scope.md +183 -0
- package/.onto/domains/software-engineering/extension_cases.md +551 -0
- package/.onto/domains/software-engineering/logic_rules.md +240 -0
- package/.onto/domains/software-engineering/problem_framing_profile.md +68 -0
- package/.onto/domains/software-engineering/structure_spec.md +185 -0
- package/.onto/domains/ui-design/competency_qs.md +567 -0
- package/.onto/domains/ui-design/concepts.md +194 -0
- package/.onto/domains/ui-design/conciseness_rules.md +190 -0
- package/.onto/domains/ui-design/dependency_rules.md +323 -0
- package/.onto/domains/ui-design/domain_scope.md +340 -0
- package/.onto/domains/ui-design/extension_cases.md +563 -0
- package/.onto/domains/ui-design/logic_rules.md +349 -0
- package/.onto/domains/ui-design/structure_spec.md +252 -0
- package/.onto/domains/visual-design/competency_qs.md +472 -0
- package/.onto/domains/visual-design/concepts.md +147 -0
- package/.onto/domains/visual-design/conciseness_rules.md +186 -0
- package/.onto/domains/visual-design/dependency_rules.md +282 -0
- package/.onto/domains/visual-design/domain_scope.md +290 -0
- package/.onto/domains/visual-design/extension_cases.md +480 -0
- package/.onto/domains/visual-design/logic_rules.md +232 -0
- package/.onto/domains/visual-design/structure_spec.md +213 -0
- package/.onto/principles/llm-native-development-guideline.md +401 -0
- package/.onto/principles/llm-runtime-interface-principles.md +665 -0
- package/.onto/principles/non-specialist-communication-guideline.md +74 -0
- package/.onto/principles/ontology-as-code-guideline.md +243 -0
- package/.onto/principles/ontology-as-code-naming-charter.md +130 -0
- package/.onto/principles/product-locality-principle.md +129 -0
- package/.onto/principles/productization-charter.md +569 -0
- package/.onto/processes/evolve/material-kind-adapter-contract.md +113 -0
- package/.onto/processes/reconstruct/reconstruct-boundary-contract.md +366 -0
- package/.onto/processes/reconstruct/source-profile-contract.md +107 -0
- package/.onto/processes/reconstruct/source-profiles/code.md +72 -0
- package/.onto/processes/reconstruct/source-profiles/database.md +74 -0
- package/.onto/processes/reconstruct/source-profiles/document.md +71 -0
- package/.onto/processes/reconstruct/source-profiles/spreadsheet.md +79 -0
- package/.onto/processes/review/binding-contract.md +270 -0
- package/.onto/processes/review/execution-preparation-artifacts.md +281 -0
- package/.onto/processes/review/interpretation-contract.md +245 -0
- package/.onto/processes/review/issue-stance-deliberation-contract.md +761 -0
- package/.onto/processes/review/lens-prompt-contract.md +402 -0
- package/.onto/processes/review/lens-registry.md +127 -0
- package/.onto/processes/review/pre-dispatch-contracts.md +428 -0
- package/.onto/processes/review/productized-live-path.md +398 -0
- package/.onto/processes/review/prompt-execution-runner-contract.md +187 -0
- package/.onto/processes/review/record-contract.md +427 -0
- package/.onto/processes/review/record-field-mapping.md +337 -0
- package/.onto/processes/review/review-context-manifest-contract.md +356 -0
- package/.onto/processes/review/review-execution-ux-contract.md +809 -0
- package/.onto/processes/review/review-target-profile-contract.md +259 -0
- package/.onto/processes/review/shared-phenomenon-contract.md +129 -0
- package/.onto/processes/review/synthesize-prompt-contract.md +343 -0
- package/.onto/processes/shared/target-material-kind-contract.md +198 -0
- package/.onto/roles/axiology.md +81 -0
- package/.onto/roles/conciseness.md +36 -0
- package/.onto/roles/coverage.md +34 -0
- package/.onto/roles/dependency.md +37 -0
- package/.onto/roles/evolution.md +35 -0
- package/.onto/roles/logic.md +104 -0
- package/.onto/roles/pragmatics.md +32 -0
- package/.onto/roles/semantics.md +36 -0
- package/.onto/roles/structure.md +33 -0
- package/.onto/roles/synthesize.md +92 -0
- package/AGENTS.md +240 -0
- package/CLAUDE.md +39 -0
- package/README.md +287 -0
- package/bin/onto +92 -0
- package/dist/cli.js +101 -0
- package/dist/core-api/reconstruct-api.js +222 -0
- package/dist/core-api/review-api.js +1271 -0
- package/dist/core-runtime/cli/assemble-review-record.js +431 -0
- package/dist/core-runtime/cli/bootstrap-review-binding.js +186 -0
- package/dist/core-runtime/cli/codex-nested-dispatch.js +226 -0
- package/dist/core-runtime/cli/codex-nested-dispatch.test.js +390 -0
- package/dist/core-runtime/cli/codex-nested-teamlead-executor.js +464 -0
- package/dist/core-runtime/cli/codex-nested-teamlead-executor.test.js +335 -0
- package/dist/core-runtime/cli/codex-review-unit-executor.js +228 -0
- package/dist/core-runtime/cli/complete-review-session.js +64 -0
- package/dist/core-runtime/cli/complexity-assessment.js +153 -0
- package/dist/core-runtime/cli/coordinator-helpers.js +583 -0
- package/dist/core-runtime/cli/coordinator-state-machine-deliberation.test.js +167 -0
- package/dist/core-runtime/cli/coordinator-state-machine.js +794 -0
- package/dist/core-runtime/cli/e2e-codex-multi-agent-fixes.test.js +615 -0
- package/dist/core-runtime/cli/e2e-start-review-session.test.js +312 -0
- package/dist/core-runtime/cli/health.js +44 -0
- package/dist/core-runtime/cli/inline-http-review-unit-executor.js +656 -0
- package/dist/core-runtime/cli/inline-http-review-unit-executor.test.js +567 -0
- package/dist/core-runtime/cli/materialize-review-execution-preparation.js +104 -0
- package/dist/core-runtime/cli/materialize-review-prompt-packets.js +952 -0
- package/dist/core-runtime/cli/migrate-session-roots.js +118 -0
- package/dist/core-runtime/cli/mock-review-unit-executor.js +285 -0
- package/dist/core-runtime/cli/onto-tools.js +369 -0
- package/dist/core-runtime/cli/prepare-review-session.js +272 -0
- package/dist/core-runtime/cli/render-review-final-output.js +350 -0
- package/dist/core-runtime/cli/repo-layout-migration-replace.smoke.test.js +106 -0
- package/dist/core-runtime/cli/review-invoke-auto-resolution.test.js +268 -0
- package/dist/core-runtime/cli/review-invoke-coordinator-topology.test.js +136 -0
- package/dist/core-runtime/cli/review-invoke-resolver-caching.test.js +201 -0
- package/dist/core-runtime/cli/review-invoke-topology-dispatch.test.js +192 -0
- package/dist/core-runtime/cli/review-invoke.js +2030 -0
- package/dist/core-runtime/cli/run-review-prompt-execution.js +2152 -0
- package/dist/core-runtime/cli/session-root-guard.js +168 -0
- package/dist/core-runtime/cli/spawn-watcher.js +173 -0
- package/dist/core-runtime/cli/spawn-watcher.test.js +457 -0
- package/dist/core-runtime/cli/start-review-session.js +68 -0
- package/dist/core-runtime/cli/strip-wrapping-code-fence.js +56 -0
- package/dist/core-runtime/cli/strip-wrapping-code-fence.test.js +79 -0
- package/dist/core-runtime/cli/teamcreate-lens-deliberation-executor.js +412 -0
- package/dist/core-runtime/cli/teamcreate-lens-deliberation-executor.test.js +351 -0
- package/dist/core-runtime/cli/topology-executor-mapping.js +139 -0
- package/dist/core-runtime/cli/topology-executor-mapping.test.js +173 -0
- package/dist/core-runtime/cli/write-review-interpretation.js +81 -0
- package/dist/core-runtime/config/onto-config-cli.js +278 -0
- package/dist/core-runtime/config/onto-config-key-path.js +288 -0
- package/dist/core-runtime/config/onto-config-key-path.test.js +195 -0
- package/dist/core-runtime/config/onto-config-preview.js +108 -0
- package/dist/core-runtime/config/onto-config-preview.test.js +132 -0
- package/dist/core-runtime/discovery/config-chain.js +118 -0
- package/dist/core-runtime/discovery/config-chain.test.js +103 -0
- package/dist/core-runtime/discovery/config-profile.js +199 -0
- package/dist/core-runtime/discovery/config-profile.test.js +233 -0
- package/dist/core-runtime/discovery/host-detection.js +33 -0
- package/dist/core-runtime/discovery/host-detection.test.js +186 -0
- package/dist/core-runtime/discovery/installation-paths.js +21 -0
- package/dist/core-runtime/discovery/installation-paths.test.js +65 -0
- package/dist/core-runtime/discovery/lens-registry.js +60 -0
- package/dist/core-runtime/discovery/lens-registry.test.js +81 -0
- package/dist/core-runtime/discovery/onto-home.js +71 -0
- package/dist/core-runtime/discovery/path-normalization.js +28 -0
- package/dist/core-runtime/discovery/path-normalization.test.js +22 -0
- package/dist/core-runtime/discovery/plugin-path.js +72 -0
- package/dist/core-runtime/discovery/plugin-path.test.js +95 -0
- package/dist/core-runtime/discovery/project-root.js +47 -0
- package/dist/core-runtime/discovery/settings-chain.js +353 -0
- package/dist/core-runtime/discovery/walk-up.js +17 -0
- package/dist/core-runtime/evolve/adapters/code-product/compile/compile-defense.js +344 -0
- package/dist/core-runtime/evolve/adapters/code-product/compile/compile-defense.test.js +915 -0
- package/dist/core-runtime/evolve/adapters/code-product/compile/compile.js +564 -0
- package/dist/core-runtime/evolve/adapters/code-product/compile/compile.test.js +708 -0
- package/dist/core-runtime/evolve/adapters/code-product/parsers/brief-parser.js +165 -0
- package/dist/core-runtime/evolve/adapters/code-product/parsers/brief-parser.test.js +227 -0
- package/dist/core-runtime/evolve/adapters/code-product/validators/validate.js +59 -0
- package/dist/core-runtime/evolve/adapters/code-product/validators/validate.test.js +205 -0
- package/dist/core-runtime/evolve/adapters/methodology/adapter.js +16 -0
- package/dist/core-runtime/evolve/adapters/methodology/adapter.test.js +9 -0
- package/dist/core-runtime/evolve/adapters/methodology/perspectives/authority-consistency.js +298 -0
- package/dist/core-runtime/evolve/adapters/methodology/perspectives/authority-consistency.test.js +70 -0
- package/dist/core-runtime/evolve/adapters/methodology/scope-types/process.js +46 -0
- package/dist/core-runtime/evolve/adapters/methodology/scope-types/process.test.js +73 -0
- package/dist/core-runtime/evolve/adapters/registry.js +47 -0
- package/dist/core-runtime/evolve/adapters/registry.test.js +67 -0
- package/dist/core-runtime/evolve/cli.js +256 -0
- package/dist/core-runtime/evolve/commands/align.js +194 -0
- package/dist/core-runtime/evolve/commands/align.test.js +82 -0
- package/dist/core-runtime/evolve/commands/apply.js +161 -0
- package/dist/core-runtime/evolve/commands/apply.test.js +138 -0
- package/dist/core-runtime/evolve/commands/close.js +39 -0
- package/dist/core-runtime/evolve/commands/close.test.js +99 -0
- package/dist/core-runtime/evolve/commands/defer.js +40 -0
- package/dist/core-runtime/evolve/commands/defer.test.js +134 -0
- package/dist/core-runtime/evolve/commands/draft.js +323 -0
- package/dist/core-runtime/evolve/commands/draft.test.js +178 -0
- package/dist/core-runtime/evolve/commands/e2e-evolve-full-cycle.test.js +208 -0
- package/dist/core-runtime/evolve/commands/error-messages.js +125 -0
- package/dist/core-runtime/evolve/commands/error-messages.test.js +167 -0
- package/dist/core-runtime/evolve/commands/propose-align.js +222 -0
- package/dist/core-runtime/evolve/commands/propose-align.test.js +136 -0
- package/dist/core-runtime/evolve/commands/reconstruct.js +330 -0
- package/dist/core-runtime/evolve/commands/reconstruct.test.js +278 -0
- package/dist/core-runtime/evolve/commands/shared.js +22 -0
- package/dist/core-runtime/evolve/commands/stale-check.js +103 -0
- package/dist/core-runtime/evolve/commands/stale-check.test.js +84 -0
- package/dist/core-runtime/evolve/commands/start.js +887 -0
- package/dist/core-runtime/evolve/commands/start.test.js +396 -0
- package/dist/core-runtime/evolve/config/project-config.js +99 -0
- package/dist/core-runtime/evolve/config/project-config.test.js +170 -0
- package/dist/core-runtime/evolve/renderers/align-packet.js +280 -0
- package/dist/core-runtime/evolve/renderers/align-packet.test.js +332 -0
- package/dist/core-runtime/evolve/renderers/draft-packet.js +303 -0
- package/dist/core-runtime/evolve/renderers/draft-packet.test.js +377 -0
- package/dist/core-runtime/evolve/renderers/format.js +5 -0
- package/dist/core-runtime/evolve/renderers/scope-md.js +237 -0
- package/dist/core-runtime/evolve/renderers/scope-md.test.js +306 -0
- package/dist/core-runtime/govern/cli.js +369 -0
- package/dist/core-runtime/govern/cli.test.js +314 -0
- package/dist/core-runtime/govern/drift-engine.js +103 -0
- package/dist/core-runtime/govern/drift-engine.test.js +319 -0
- package/dist/core-runtime/govern/promote-principle.js +206 -0
- package/dist/core-runtime/govern/promote-principle.test.js +368 -0
- package/dist/core-runtime/govern/queue.js +81 -0
- package/dist/core-runtime/govern/types.js +16 -0
- package/dist/core-runtime/install/cli.js +530 -0
- package/dist/core-runtime/install/detect.js +128 -0
- package/dist/core-runtime/install/detect.test.js +155 -0
- package/dist/core-runtime/install/gitignore-update.js +74 -0
- package/dist/core-runtime/install/gitignore-update.test.js +64 -0
- package/dist/core-runtime/install/install-integration.test.js +373 -0
- package/dist/core-runtime/install/prompts.js +389 -0
- package/dist/core-runtime/install/prompts.test.js +293 -0
- package/dist/core-runtime/install/types.js +26 -0
- package/dist/core-runtime/install/validation.js +295 -0
- package/dist/core-runtime/install/validation.test.js +313 -0
- package/dist/core-runtime/install/writer.js +254 -0
- package/dist/core-runtime/install/writer.test.js +218 -0
- package/dist/core-runtime/learning/extractor.js +461 -0
- package/dist/core-runtime/learning/feedback.js +179 -0
- package/dist/core-runtime/learning/health-report.js +165 -0
- package/dist/core-runtime/learning/health-report.test.js +169 -0
- package/dist/core-runtime/learning/loader.js +388 -0
- package/dist/core-runtime/learning/loader.test.js +102 -0
- package/dist/core-runtime/learning/promote/apply-state.js +240 -0
- package/dist/core-runtime/learning/promote/audit-obligation.js +195 -0
- package/dist/core-runtime/learning/promote/collector.js +432 -0
- package/dist/core-runtime/learning/promote/degraded-state.js +125 -0
- package/dist/core-runtime/learning/promote/domain-doc-proposer.js +166 -0
- package/dist/core-runtime/learning/promote/e2e-promote.test.js +6385 -0
- package/dist/core-runtime/learning/promote/health-snapshot.js +150 -0
- package/dist/core-runtime/learning/promote/insight-reclassifier.js +544 -0
- package/dist/core-runtime/learning/promote/judgment-auditor.js +517 -0
- package/dist/core-runtime/learning/promote/panel-reviewer.js +1158 -0
- package/dist/core-runtime/learning/promote/promote-executor.js +1675 -0
- package/dist/core-runtime/learning/promote/promoter.js +307 -0
- package/dist/core-runtime/learning/promote/retirement.js +122 -0
- package/dist/core-runtime/learning/promote/types.js +23 -0
- package/dist/core-runtime/learning/prompt-sections.js +51 -0
- package/dist/core-runtime/learning/shared/artifact-registry-init.js +45 -0
- package/dist/core-runtime/learning/shared/artifact-registry.js +254 -0
- package/dist/core-runtime/learning/shared/audit-obligation-kernel.js +73 -0
- package/dist/core-runtime/learning/shared/audit-state.js +99 -0
- package/dist/core-runtime/learning/shared/duplicate-check.js +28 -0
- package/dist/core-runtime/learning/shared/llm-caller.js +831 -0
- package/dist/core-runtime/learning/shared/llm-caller.test.js +601 -0
- package/dist/core-runtime/learning/shared/llm-tool-loop.js +393 -0
- package/dist/core-runtime/learning/shared/mode.js +25 -0
- package/dist/core-runtime/learning/shared/paths.js +84 -0
- package/dist/core-runtime/learning/shared/paths.test.js +79 -0
- package/dist/core-runtime/learning/shared/patterns.js +37 -0
- package/dist/core-runtime/learning/shared/recoverability.js +355 -0
- package/dist/core-runtime/learning/shared/recovery-context.js +374 -0
- package/dist/core-runtime/learning/shared/scope.js +1 -0
- package/dist/core-runtime/learning/shared/semantic-classifier.js +94 -0
- package/dist/core-runtime/learning/shared/specs/apply-execution-state-spec.js +42 -0
- package/dist/core-runtime/learning/shared/specs/audit-state-spec.js +37 -0
- package/dist/core-runtime/learning/shared/specs/backup-metadata-spec.js +39 -0
- package/dist/core-runtime/learning/shared/specs/emergency-log-spec.js +41 -0
- package/dist/core-runtime/learning/shared/specs/layout-version-spec.js +38 -0
- package/dist/core-runtime/learning/shared/specs/promote-decisions-spec.js +43 -0
- package/dist/core-runtime/learning/shared/specs/promote-report-spec.js +113 -0
- package/dist/core-runtime/learning/shared/specs/prune-log-spec.js +36 -0
- package/dist/core-runtime/learning/shared/specs/recovery-resolution-spec.js +48 -0
- package/dist/core-runtime/learning/shared/specs/restore-manifest-spec.js +43 -0
- package/dist/core-runtime/learning/shared/specs/spec-helpers.js +64 -0
- package/dist/core-runtime/learning/usage-tracker.js +190 -0
- package/dist/core-runtime/learning/usage-tracker.test.js +176 -0
- package/dist/core-runtime/llm/llm-caller.js +649 -0
- package/dist/core-runtime/llm/llm-tool-loop.js +401 -0
- package/dist/core-runtime/llm/model-switcher.js +62 -0
- package/dist/core-runtime/logger.js +22 -0
- package/dist/core-runtime/onboard/detect-review-axes.js +122 -0
- package/dist/core-runtime/onboard/detect-review-axes.test.js +127 -0
- package/dist/core-runtime/onboard/write-review-block.js +188 -0
- package/dist/core-runtime/onboard/write-review-block.test.js +240 -0
- package/dist/core-runtime/readers/brownfield-builder.js +150 -0
- package/dist/core-runtime/readers/brownfield-builder.test.js +136 -0
- package/dist/core-runtime/readers/code-chunk-collector.js +53 -0
- package/dist/core-runtime/readers/code-chunk-collector.test.js +136 -0
- package/dist/core-runtime/readers/file-utils.js +240 -0
- package/dist/core-runtime/readers/file-utils.test.js +146 -0
- package/dist/core-runtime/readers/lexicon-citation-check.js +93 -0
- package/dist/core-runtime/readers/lexicon-citation-check.test.js +77 -0
- package/dist/core-runtime/readers/mcp-figma.js +30 -0
- package/dist/core-runtime/readers/mcp-figma.test.js +82 -0
- package/dist/core-runtime/readers/mcp-generic.js +31 -0
- package/dist/core-runtime/readers/mcp-generic.test.js +76 -0
- package/dist/core-runtime/readers/ontology-index.js +148 -0
- package/dist/core-runtime/readers/ontology-index.test.js +245 -0
- package/dist/core-runtime/readers/ontology-query.js +168 -0
- package/dist/core-runtime/readers/ontology-query.test.js +311 -0
- package/dist/core-runtime/readers/ontology-resolve.js +48 -0
- package/dist/core-runtime/readers/ontology-resolve.test.js +48 -0
- package/dist/core-runtime/readers/patterns/index.js +7 -0
- package/dist/core-runtime/readers/review-log.js +213 -0
- package/dist/core-runtime/readers/review-log.test.js +313 -0
- package/dist/core-runtime/readers/scan-local.js +102 -0
- package/dist/core-runtime/readers/scan-local.test.js +102 -0
- package/dist/core-runtime/readers/scan-tarball.js +121 -0
- package/dist/core-runtime/readers/scan-tarball.test.js +283 -0
- package/dist/core-runtime/readers/scan-vault.js +34 -0
- package/dist/core-runtime/readers/scan-vault.test.js +81 -0
- package/dist/core-runtime/readers/types.js +42 -0
- package/dist/core-runtime/readers/types.test.js +94 -0
- package/dist/core-runtime/readers/viewpoint-collectors.js +229 -0
- package/dist/core-runtime/reconstruct/artifact-types.js +1 -0
- package/dist/core-runtime/reconstruct/directive-validation.js +123 -0
- package/dist/core-runtime/reconstruct/materialize-preparation.js +251 -0
- package/dist/core-runtime/reconstruct/record.js +198 -0
- package/dist/core-runtime/reconstruct/run.js +578 -0
- package/dist/core-runtime/reconstruct/seed-candidate-validation.js +356 -0
- package/dist/core-runtime/reconstruct/source-observations.js +62 -0
- package/dist/core-runtime/reconstruct/source-profiles.js +73 -0
- package/dist/core-runtime/release-channel/release-channel.js +90 -0
- package/dist/core-runtime/review/artifact-types.js +13 -0
- package/dist/core-runtime/review/citation-audit.js +204 -0
- package/dist/core-runtime/review/citation-audit.test.js +165 -0
- package/dist/core-runtime/review/controlled-lens-deliberation.js +125 -0
- package/dist/core-runtime/review/execution-plan-resolver.js +247 -0
- package/dist/core-runtime/review/execution-plan-resolver.test.js +243 -0
- package/dist/core-runtime/review/execution-topology-resolver-axis-first.test.js +246 -0
- package/dist/core-runtime/review/execution-topology-resolver.js +401 -0
- package/dist/core-runtime/review/execution-topology-resolver.test.js +315 -0
- package/dist/core-runtime/review/failure-records.js +57 -0
- package/dist/core-runtime/review/inline-context-embedder.js +141 -0
- package/dist/core-runtime/review/inline-context-embedder.test.js +154 -0
- package/dist/core-runtime/review/issue-artifact-runtime.js +859 -0
- package/dist/core-runtime/review/legacy-mode-policy.js +88 -0
- package/dist/core-runtime/review/lens-completion-policy.js +17 -0
- package/dist/core-runtime/review/materializers-effort-persist.test.js +79 -0
- package/dist/core-runtime/review/materializers.js +963 -0
- package/dist/core-runtime/review/ontology-path-classifier.js +179 -0
- package/dist/core-runtime/review/ontology-path-classifier.test.js +216 -0
- package/dist/core-runtime/review/packet-boundary-policy.js +215 -0
- package/dist/core-runtime/review/packet-boundary-policy.test.js +107 -0
- package/dist/core-runtime/review/participating-lens-paths.js +61 -0
- package/dist/core-runtime/review/participating-lens-paths.test.js +73 -0
- package/dist/core-runtime/review/review-artifact-utils.js +287 -0
- package/dist/core-runtime/review/review-config-legacy-translate.js +244 -0
- package/dist/core-runtime/review/review-config-legacy-translate.test.js +161 -0
- package/dist/core-runtime/review/review-config-validator.js +289 -0
- package/dist/core-runtime/review/review-config-validator.test.js +236 -0
- package/dist/core-runtime/review/review-execution-profile.js +193 -0
- package/dist/core-runtime/review/review-execution-route.js +52 -0
- package/dist/core-runtime/review/review-progress-contract.js +123 -0
- package/dist/core-runtime/review/review-record-validation.js +251 -0
- package/dist/core-runtime/review/review-result-classification.js +379 -0
- package/dist/core-runtime/review/review-state-machine.js +39 -0
- package/dist/core-runtime/review/route-visibility.js +125 -0
- package/dist/core-runtime/review/shape-pipeline-audit.test.js +311 -0
- package/dist/core-runtime/review/shape-to-topology-id.js +117 -0
- package/dist/core-runtime/review/shape-to-topology-id.test.js +132 -0
- package/dist/core-runtime/review/topology-shape-derivation.js +155 -0
- package/dist/core-runtime/review/topology-shape-derivation.test.js +195 -0
- package/dist/core-runtime/scope-runtime/constants.js +12 -0
- package/dist/core-runtime/scope-runtime/constraint-pool.js +166 -0
- package/dist/core-runtime/scope-runtime/constraint-pool.test.js +674 -0
- package/dist/core-runtime/scope-runtime/domain-validation-log.js +135 -0
- package/dist/core-runtime/scope-runtime/domain-validation-log.test.js +156 -0
- package/dist/core-runtime/scope-runtime/eval-persistence.js +65 -0
- package/dist/core-runtime/scope-runtime/eval-persistence.test.js +84 -0
- package/dist/core-runtime/scope-runtime/event-pipeline.js +64 -0
- package/dist/core-runtime/scope-runtime/event-pipeline.test.js +450 -0
- package/dist/core-runtime/scope-runtime/event-store.js +39 -0
- package/dist/core-runtime/scope-runtime/event-store.test.js +95 -0
- package/dist/core-runtime/scope-runtime/gate-guard.js +348 -0
- package/dist/core-runtime/scope-runtime/gate-guard.test.js +1047 -0
- package/dist/core-runtime/scope-runtime/hash.js +4 -0
- package/dist/core-runtime/scope-runtime/hash.test.js +33 -0
- package/dist/core-runtime/scope-runtime/id.js +4 -0
- package/dist/core-runtime/scope-runtime/id.test.js +17 -0
- package/dist/core-runtime/scope-runtime/reducer.js +297 -0
- package/dist/core-runtime/scope-runtime/reducer.test.js +759 -0
- package/dist/core-runtime/scope-runtime/scope-manager.js +161 -0
- package/dist/core-runtime/scope-runtime/state-machine.js +309 -0
- package/dist/core-runtime/scope-runtime/state-machine.test.js +704 -0
- package/dist/core-runtime/scope-runtime/types.js +116 -0
- package/dist/core-runtime/scope-runtime/types.test.js +69 -0
- package/dist/core-runtime/target-material-kind.js +256 -0
- package/dist/core-runtime/translate/render-for-user.js +169 -0
- package/dist/core-runtime/translate/render-for-user.test.js +122 -0
- package/dist/mcp/server.js +1011 -0
- package/dist/mcp/tool-schemas.js +93 -0
- package/dist/providers/capability-contract.js +1 -0
- package/package.json +68 -0
- package/settings.example.json +33 -0
|
@@ -0,0 +1,398 @@
|
|
|
1
|
+
# Review Productized Live Path
|
|
2
|
+
|
|
3
|
+
> 상태: Active
|
|
4
|
+
> 목적: `검토 (review)`의 `제품화된 실시간 경로 (productized live path)`를 canonical 실행 경로로 고정한다.
|
|
5
|
+
> 기준 문서:
|
|
6
|
+
> - `.onto/processes/review/interpretation-contract.md`
|
|
7
|
+
> - `.onto/processes/review/binding-contract.md`
|
|
8
|
+
> - `.onto/processes/review/lens-prompt-contract.md`
|
|
9
|
+
> - `.onto/processes/review/synthesize-prompt-contract.md`
|
|
10
|
+
> - `.onto/processes/review/record-contract.md`
|
|
11
|
+
> - `.onto/processes/review/record-field-mapping.md`
|
|
12
|
+
> - `.onto/processes/review/execution-preparation-artifacts.md`
|
|
13
|
+
> - `.onto/processes/review/prompt-execution-runner-contract.md`
|
|
14
|
+
> - `.onto/processes/review/issue-stance-deliberation-contract.md`
|
|
15
|
+
> - `.onto/processes/review/review-context-manifest-contract.md`
|
|
16
|
+
> - `.onto/processes/review/pre-dispatch-contracts.md`
|
|
17
|
+
> - `.onto/processes/review/review-execution-ux-contract.md`
|
|
18
|
+
> - `.onto/authority/core-lexicon.yaml`
|
|
19
|
+
|
|
20
|
+
---
|
|
21
|
+
|
|
22
|
+
## 1. Position
|
|
23
|
+
|
|
24
|
+
이 문서는 `검토 (review)`의 현재 canonical live execution truth다.
|
|
25
|
+
|
|
26
|
+
즉 실제 productization target은 이 문서의 순서를 따른다.
|
|
27
|
+
|
|
28
|
+
live path authority는 이 문서와 하위 review contracts가 가진다.
|
|
29
|
+
|
|
30
|
+
---
|
|
31
|
+
|
|
32
|
+
## 2. Canonical Live Path
|
|
33
|
+
|
|
34
|
+
```text
|
|
35
|
+
user request
|
|
36
|
+
-> 호출 해석 (InvocationInterpretation)
|
|
37
|
+
-> 주체자 확인 / 선택 확정
|
|
38
|
+
-> 호출 고정 (InvocationBinding)
|
|
39
|
+
-> execution preparation artifacts
|
|
40
|
+
-> 9개 lens 독립 실행
|
|
41
|
+
-> 통제된 lens 숙의 (controlled lens deliberation)
|
|
42
|
+
-> 종합 단계 (synthesize)
|
|
43
|
+
-> human-readable final output
|
|
44
|
+
-> 리뷰 기록 (ReviewRecord)
|
|
45
|
+
```
|
|
46
|
+
|
|
47
|
+
---
|
|
48
|
+
|
|
49
|
+
## 3. Step-by-Step
|
|
50
|
+
|
|
51
|
+
### 3.1 주체자 요청 수집
|
|
52
|
+
|
|
53
|
+
host가 아래를 받는다.
|
|
54
|
+
|
|
55
|
+
- 주체자 자연어 요청
|
|
56
|
+
- explicit target token
|
|
57
|
+
- explicit domain token
|
|
58
|
+
- explicit execution profile token
|
|
59
|
+
- 현재 repo / selected target context
|
|
60
|
+
|
|
61
|
+
### 3.2 호출 해석 (InvocationInterpretation)
|
|
62
|
+
|
|
63
|
+
`LLM`이 아래를 해석한다.
|
|
64
|
+
|
|
65
|
+
- entrypoint가 `review`인지
|
|
66
|
+
- `검토 대상 범위 (review_target_scope)` 후보가 무엇인지
|
|
67
|
+
- intent가 무엇인지
|
|
68
|
+
- domain recommendation이 필요한지
|
|
69
|
+
- `lens 선택 계획 (LensSelectionPlan)`이 무엇인지
|
|
70
|
+
- `core-axis/full` recommendation이 필요한지
|
|
71
|
+
|
|
72
|
+
prompt-backed path에서도 이 단계의 결과는 최종적으로
|
|
73
|
+
`interpretation.yaml`로 materialize되어야 한다.
|
|
74
|
+
|
|
75
|
+
### 3.3 주체자 확인 / 선택 확정
|
|
76
|
+
|
|
77
|
+
필요한 경우 아래를 주체자와 확정한다.
|
|
78
|
+
|
|
79
|
+
- `도메인 최종 선택 (DomainFinalSelection)`
|
|
80
|
+
- `core-axis/full`
|
|
81
|
+
- explicit override
|
|
82
|
+
|
|
83
|
+
이 단계는 semantic recommendation과 deterministic binding의 중간에서
|
|
84
|
+
주체자가 최종 authority를 행사하는 구간이다.
|
|
85
|
+
|
|
86
|
+
현재 host-facing `review:invoke`의 기본 규칙:
|
|
87
|
+
|
|
88
|
+
- explicit `--domain {name}` / `--no-domain` 이 있으면 그대로 사용
|
|
89
|
+
- configured domain이 하나면 바로 사용
|
|
90
|
+
- configured domain이 여러 개면 interactive selection을 수행
|
|
91
|
+
- interactive selection이 불가능한 non-interactive 환경이면 fail-fast 하고 explicit domain selection을 요구
|
|
92
|
+
- `--domain` 과 `--no-domain` 동시 지정은 parser layer 에서 fail-fast
|
|
93
|
+
|
|
94
|
+
### 3.4 호출 고정 (InvocationBinding)
|
|
95
|
+
|
|
96
|
+
runtime/host가 아래를 고정한다.
|
|
97
|
+
|
|
98
|
+
- resolved target scope
|
|
99
|
+
- final domain value
|
|
100
|
+
- resolved execution realization
|
|
101
|
+
- resolved host runtime
|
|
102
|
+
- resolved review mode
|
|
103
|
+
- resolved lens set
|
|
104
|
+
- session root
|
|
105
|
+
- artifact paths
|
|
106
|
+
|
|
107
|
+
이 단계가 끝나면 prompt-backed path에서도
|
|
108
|
+
적어도 `session_root`와 각 artifact path는 확정되어 있어야 한다.
|
|
109
|
+
|
|
110
|
+
현재 bounded runtime step은 TypeScript core로 구현한다.
|
|
111
|
+
|
|
112
|
+
Preferred repo-local combined entrypoint는 아래다.
|
|
113
|
+
|
|
114
|
+
- `npm run review:invoke -- ...`
|
|
115
|
+
|
|
116
|
+
기본 bounded start step은 아래다.
|
|
117
|
+
|
|
118
|
+
- `npm run review:start-session -- ...`
|
|
119
|
+
|
|
120
|
+
`review:start-session` 내부에서 interpretation, binding, execution preparation,
|
|
121
|
+
prompt packet materialization을 순서대로 수행한다.
|
|
122
|
+
|
|
123
|
+
### 3.5 Execution Preparation Artifacts
|
|
124
|
+
|
|
125
|
+
binding 다음에는 최소 아래 artifact가 materialize되어야 한다.
|
|
126
|
+
|
|
127
|
+
1. `session metadata`
|
|
128
|
+
2. `target snapshot`
|
|
129
|
+
3. `review target materialized input`
|
|
130
|
+
4. `context candidate assembly`
|
|
131
|
+
5. `execution plan`
|
|
132
|
+
6. `actor invocation profiles`
|
|
133
|
+
7. `actor consumer bindings`
|
|
134
|
+
8. `domain binding`
|
|
135
|
+
9. `review value-alignment criteria`
|
|
136
|
+
10. `review context manifest`
|
|
137
|
+
11. `prompt packets`
|
|
138
|
+
12. `execution result`
|
|
139
|
+
|
|
140
|
+
이 단계는 later `ReviewRecord`와 runtime replacement의 bridge다.
|
|
141
|
+
|
|
142
|
+
세부 contract는 `.onto/processes/review/execution-preparation-artifacts.md`를 따른다.
|
|
143
|
+
dispatch 전에 닫혀야 하는 gate와 phase boundary는
|
|
144
|
+
`.onto/processes/review/pre-dispatch-contracts.md`를 따른다.
|
|
145
|
+
|
|
146
|
+
prompt-backed path에서도 실제 파일이 만들어져야 한다.
|
|
147
|
+
즉 이 단계는 단순 개념 설명이 아니라 artifact materialization step이다.
|
|
148
|
+
|
|
149
|
+
`execution plan`은 아래를 deterministic하게 고정한다.
|
|
150
|
+
|
|
151
|
+
- lens별 output seat
|
|
152
|
+
- `synthesis.md` seat
|
|
153
|
+
- `deliberation.md` seat
|
|
154
|
+
- lens별 deliberation response seat
|
|
155
|
+
- `error-log.md` seat
|
|
156
|
+
- `final-output.md` seat
|
|
157
|
+
- `review-record.yaml` seat
|
|
158
|
+
- `execution-result.yaml` seat
|
|
159
|
+
- boundary seat
|
|
160
|
+
- `BoundaryPolicy`
|
|
161
|
+
- `BoundaryPresentation`
|
|
162
|
+
- `BoundaryEnforcementProfile`
|
|
163
|
+
- `EffectiveBoundaryState`
|
|
164
|
+
|
|
165
|
+
`prompt packets`는 아래를 deterministic하게 고정한다.
|
|
166
|
+
|
|
167
|
+
- lens별 prompt handoff text
|
|
168
|
+
- `synthesize` prompt handoff text
|
|
169
|
+
- 각 packet이 읽어야 할 artifact path
|
|
170
|
+
- 각 packet이 써야 할 output path
|
|
171
|
+
- packet은 가급적 lightweight handoff여야 하며, authoritative artifact의 전체 본문을 과도하게 embedded하지 않는다
|
|
172
|
+
- `맥락 격리 추론 단위 (ContextIsolatedReasoningUnit)`는 packet이 가리키는 artifact file을 직접 읽는다
|
|
173
|
+
- synthesize runtime packet은 participating lens output의 seat/ref를 넘기되, 가능하면 본문 전체를 다시 중복 embedded하지 않는다
|
|
174
|
+
- packet은 boundary policy와 effective boundary state를 함께 제시해, 허용된 탐색 공간과 강제 강도를 분명히 해야 한다
|
|
175
|
+
|
|
176
|
+
### 3.6 9개 lens 독립 실행
|
|
177
|
+
|
|
178
|
+
각 lens는 **맥락 격리 추론 단위 (ContextIsolatedReasoningUnit)** 로 실행된다.
|
|
179
|
+
|
|
180
|
+
가능하면 host는 자유 텍스트 대신 TS core가 materialize한 prompt packet을 사용해야 한다.
|
|
181
|
+
기본 bounded dispatch step은 아래다.
|
|
182
|
+
|
|
183
|
+
- `npm run review:run-prompt-execution -- ...`
|
|
184
|
+
|
|
185
|
+
이 step은 실행 종료 시 `execution-result.yaml`을 반드시 materialize해야 한다.
|
|
186
|
+
|
|
187
|
+
canonical requirement:
|
|
188
|
+
|
|
189
|
+
1. 각 lens는 자기 전용 맥락을 가진다
|
|
190
|
+
2. Round 1에서는 다른 lens 결과를 보지 않는다
|
|
191
|
+
3. 메인 `LLM` 콘텍스트는 per-lens detailed reasoning을 직접 모두 담지 않는다
|
|
192
|
+
4. 각 실행 단위는 packet에 제시된 `BoundaryPolicy`와 `EffectiveBoundaryState`를 hard constraint로 읽는다
|
|
193
|
+
5. 경계 안에서 충분한 근거를 얻지 못하면 broad search로 타협하지 않고 degraded/uncertain output을 남긴다
|
|
194
|
+
6. lens dispatch는 병렬 실행이 기본이다
|
|
195
|
+
7. runtime은 선택된 lens 전체를 dispatch한다. host가 자체 제한으로 이를 수행할 수 없으면 fail-loud하게 중단한다
|
|
196
|
+
|
|
197
|
+
현재 repo-local TS bounded path에서 연결된 `ReviewExecutionProfile` mode는 아래다.
|
|
198
|
+
|
|
199
|
+
- `main-workers`: main이 teamlead 역할을 수행하고 worker가 lens를 실행한다.
|
|
200
|
+
- `nested-workers`: worker teamlead가 nested worker lens를 실행한다. `teamlead.llm`은 outer worker에, `lens.llm`은 inner lens worker에 각각 적용된다.
|
|
201
|
+
- `synthesize.llm`: deliberation 이후 별도 synthesize unit에 적용된다. synthesize는 모든 lens output, issue artifacts, deliberation, problem framing을 통합하므로 높은 effort가 기본적으로 적합하다.
|
|
202
|
+
|
|
203
|
+
worker executor는 profile resolution에서 아래 중 하나로 고정된다.
|
|
204
|
+
|
|
205
|
+
- `codex`: host-bound OAuth 또는 Codex worker path.
|
|
206
|
+
- `direct_call`: `api_key` 또는 `local` provider path.
|
|
207
|
+
- `mock`: `ONTO_LLM_MOCK=1` 또는 explicit test override.
|
|
208
|
+
|
|
209
|
+
중요한 점은 host-specific naming이 아니라:
|
|
210
|
+
|
|
211
|
+
- `lens별 독립성`
|
|
212
|
+
- `메인 콘텍스트 보존`
|
|
213
|
+
- `독립 의미 검증`
|
|
214
|
+
|
|
215
|
+
이 유지되는 것이다.
|
|
216
|
+
|
|
217
|
+
packet materialization만 단독으로 디버깅해야 할 때는 아래 내부 bounded step을 쓸 수 있다.
|
|
218
|
+
|
|
219
|
+
- `npm run review:materialize-prompt-packets -- ...`
|
|
220
|
+
|
|
221
|
+
현재 TS bounded runner의 lens 병렬성은 선택된 lens 수와 같다. full review는
|
|
222
|
+
9개 lens를 모두 병렬 dispatch하고, core-axis 또는 명시적 lens 선택은 해당 lens
|
|
223
|
+
전체를 병렬 dispatch한다.
|
|
224
|
+
|
|
225
|
+
Round 1 이후에는 `lens-completion-barrier.yaml`이 작성되어야 한다.
|
|
226
|
+
canonical review path는 선택된 lens 전원이 완료되고, 선택된 lens 수가 최소 2개일 때만
|
|
227
|
+
issue artifacts, controlled deliberation, synthesize로 진행한다.
|
|
228
|
+
|
|
229
|
+
### 3.7 통제된 lens 숙의 (Controlled Lens Deliberation)
|
|
230
|
+
|
|
231
|
+
Round 1 lens 결과가 나온 뒤, review는 반드시 통제된 lens 숙의 단계를 거친다.
|
|
232
|
+
|
|
233
|
+
이 단계의 목적은 `synthesize`가 혼자 충돌을 판정하는 것이 아니라,
|
|
234
|
+
서로 다른 관점의 lens가 teamlead가 제한한 context 안에서 자기 입장을 재평가하고,
|
|
235
|
+
teamlead deliberation result가 합의/조건부 합의/지속 이견을 명시하게 하는 것이다.
|
|
236
|
+
|
|
237
|
+
canonical requirement:
|
|
238
|
+
|
|
239
|
+
1. 각 lens deliberation response는 fresh bounded context에서 실행된다.
|
|
240
|
+
2. 입력은 해당 lens의 Round 1 결과와 다른 participating lens 결과로 제한된다.
|
|
241
|
+
3. lens는 최종 종합을 수행하지 않고 자기 관점의 유지/수정/양보/지속 이견만 기록한다.
|
|
242
|
+
4. teamlead-controlled deliberation result는 모든 lens response를 읽고 `deliberation.md`를 작성한다.
|
|
243
|
+
5. `deliberation.md`는 synthesize보다 앞선 authoritative conflict-resolution artifact다.
|
|
244
|
+
6. `synthesize`는 이 결과를 소비하며, 독자적으로 새 resolution을 만들지 않는다.
|
|
245
|
+
|
|
246
|
+
MCP/TS runtime에서는 이 의미론을 provider 독립적인 controlled deliberation packet으로 실현한다.
|
|
247
|
+
중요한 것은 기능명이 아니라 `분리된 관점 + 제한 context + teamlead 통제 + 기록 가능한 resolution`이다.
|
|
248
|
+
|
|
249
|
+
#### 3.7.1 Issue-Stance Deliberation Target
|
|
250
|
+
|
|
251
|
+
다음 runtime target은 `.onto/processes/review/issue-stance-deliberation-contract.md`를 따른다.
|
|
252
|
+
|
|
253
|
+
목표 구조:
|
|
254
|
+
|
|
255
|
+
```text
|
|
256
|
+
Round 1 lens outputs
|
|
257
|
+
-> finding-ledger.yaml
|
|
258
|
+
-> finding-relation-graph.yaml
|
|
259
|
+
-> issue-ledger.yaml
|
|
260
|
+
-> issue-stance-matrix.yaml
|
|
261
|
+
-> deliberation-plan.yaml
|
|
262
|
+
-> issue-scoped controlled deliberation
|
|
263
|
+
-> deliberation.md
|
|
264
|
+
-> common spine + domain problem framing profile
|
|
265
|
+
-> problem-framing.yaml
|
|
266
|
+
-> synthesize
|
|
267
|
+
```
|
|
268
|
+
|
|
269
|
+
핵심 invariant:
|
|
270
|
+
|
|
271
|
+
1. 모든 surface finding은 finding ledger에 등록된다.
|
|
272
|
+
2. finding 간 관계를 검토해 root-cause issue cluster를 만든다.
|
|
273
|
+
3. 모든 participating lens는 모든 root-cause issue에 대해 stance를 남긴다.
|
|
274
|
+
4. material conflict가 있는 issue만 deliberation에 진입한다.
|
|
275
|
+
5. deliberation은 서로 다른 입장의 이유와 root-cause 해석 차이, stance 유지/변경 여부를 확인한다.
|
|
276
|
+
6. issue별 결론은 `no-deliberation-needed`, `resolved`, `narrowed`, `unresolved-with-reason` 중 하나다.
|
|
277
|
+
7. review closure는 `problem-framing.yaml`에서 공통 spine과 선택된 domain profile 기반 분류를 기록한다.
|
|
278
|
+
8. domain별 분류 축은 `.onto/domains/{domain}/problem_framing_profile.md`가 소유한다.
|
|
279
|
+
9. `synthesize`는 issue status와 common spine/domain axes classification을 변경하지 않고 보존적으로 렌더링한다.
|
|
280
|
+
|
|
281
|
+
### 3.8 종합 단계 (synthesize)
|
|
282
|
+
|
|
283
|
+
`synthesize`는 lens finding을 읽고 아래를 정리한다.
|
|
284
|
+
|
|
285
|
+
- consensus
|
|
286
|
+
- conditional consensus
|
|
287
|
+
- disagreement
|
|
288
|
+
- overlooked premises
|
|
289
|
+
- immediate actions
|
|
290
|
+
- recommendations
|
|
291
|
+
- final review result
|
|
292
|
+
|
|
293
|
+
중요:
|
|
294
|
+
|
|
295
|
+
- `synthesize`는 새 독립 관점을 만들지 않는다
|
|
296
|
+
- `synthesize`는 deliberation actor가 아니다
|
|
297
|
+
- 충돌 resolution의 authority는 `deliberation.md`다
|
|
298
|
+
- `New Perspectives`는 `axiology`가 제시하는 영역이다
|
|
299
|
+
- synthesize는 `axiology`가 제시한 추가 관점이 있으면 그것을 보존/배치할 수는 있지만, 스스로 invent하면 안 된다
|
|
300
|
+
|
|
301
|
+
### 3.9 리뷰 기록 (ReviewRecord)
|
|
302
|
+
|
|
303
|
+
`review`의 primary output은 `리뷰 기록 (ReviewRecord)`다.
|
|
304
|
+
|
|
305
|
+
최종 human-readable output은 먼저 render될 수 있지만,
|
|
306
|
+
later `learn/govern`가 읽을 canonical artifact는 `ReviewRecord`여야 한다.
|
|
307
|
+
|
|
308
|
+
즉:
|
|
309
|
+
|
|
310
|
+
- lens markdown
|
|
311
|
+
- deliberation markdown
|
|
312
|
+
- synthesis markdown
|
|
313
|
+
|
|
314
|
+
은 최종적으로 `ReviewRecord`의 source/human-readable layer로 내려간다.
|
|
315
|
+
|
|
316
|
+
세부 contract는 `.onto/processes/review/record-contract.md`를 따른다.
|
|
317
|
+
|
|
318
|
+
prompt-backed path에서는 `final-output.md`를 먼저 render한 뒤,
|
|
319
|
+
team lead 또는 bounded TS step이 마지막에 `review-record.yaml`을 actual aggregate로 assemble해야 한다.
|
|
320
|
+
|
|
321
|
+
Preferred repo-local combined completion 포함 entrypoint는 아래다.
|
|
322
|
+
|
|
323
|
+
- `npm run review:invoke -- ...`
|
|
324
|
+
|
|
325
|
+
기본 bounded completion step은 아래다.
|
|
326
|
+
|
|
327
|
+
- `npm run review:complete-session -- ...`
|
|
328
|
+
|
|
329
|
+
필요하면 아래 분해된 step도 내부 bounded step으로 사용할 수 있다.
|
|
330
|
+
|
|
331
|
+
- `npm run review:render-final-output -- ...`
|
|
332
|
+
- `npm run review:finalize-session -- ...`
|
|
333
|
+
|
|
334
|
+
### 3.10 Human-Readable Final Output
|
|
335
|
+
|
|
336
|
+
주체자에게 보여주는 최종 review output은
|
|
337
|
+
`ReviewRecord`와 synthesis result를 기반으로 render된 결과다.
|
|
338
|
+
|
|
339
|
+
즉 사람이 읽는 결과와
|
|
340
|
+
later system handoff artifact를 분리한다.
|
|
341
|
+
|
|
342
|
+
기본 bounded render step은 아래다.
|
|
343
|
+
|
|
344
|
+
- `npm run review:render-final-output -- ...`
|
|
345
|
+
|
|
346
|
+
중요:
|
|
347
|
+
|
|
348
|
+
- degraded case가 발생하면 `review:run-prompt-execution`은 `degradation-summary.yaml`을 구조화 source로 기록하고 `error-log.md`에는 실행 로그를 남긴다
|
|
349
|
+
- `review:complete-session`은 `final-output.md`와 `review-record.yaml`을 모두 필수 산출물로 취급한다
|
|
350
|
+
- 필수 artifact가 없으면 해당 단계는 즉시 실패한다
|
|
351
|
+
|
|
352
|
+
### 3.11 Execution UX Presentation Target
|
|
353
|
+
|
|
354
|
+
Review 진행 UX의 design target은 `.onto/processes/review/review-execution-ux-contract.md`를 따른다.
|
|
355
|
+
|
|
356
|
+
이 UX contract는 final output만이 아니라 setup, pre-dispatch readiness, lens
|
|
357
|
+
execution, issue construction, controlled deliberation, synthesize, halted partial
|
|
358
|
+
result 전 구간에서 주체자가 판단 가능한 상태를 받도록 하는 presentation target이다.
|
|
359
|
+
|
|
360
|
+
현재 active runtime truth는 기존 artifacts와 CLI/MCP 결과가 소유한다.
|
|
361
|
+
`onto.review_status`는 artifact-backed `llmPresentation.progress`를 반환하며,
|
|
362
|
+
그 안에는 polling liveness state, latest review signal, progress stepper가
|
|
363
|
+
포함된다. halted partial일 때는 `llmPresentation.halt`도 반환한다. completed result는
|
|
364
|
+
`llmPresentation.finalResult`, `final-output.md`, `review-record.yaml`에서 같은
|
|
365
|
+
result classification projection을 노출한다.
|
|
366
|
+
Progress step id/label/total truth는 runtime review progress contract와
|
|
367
|
+
`review-run-manifest.yaml.execution_contract`가 소유한다.
|
|
368
|
+
|
|
369
|
+
Severity/result classification은 runtime active projection이다. Projection source는
|
|
370
|
+
`finding-ledger.yaml`, `issue-ledger.yaml`, `problem-framing.yaml`,
|
|
371
|
+
`execution-result.yaml`이다.
|
|
372
|
+
|
|
373
|
+
Target rules:
|
|
374
|
+
|
|
375
|
+
- `severity`가 finding의 materiality boundary를 포함한다
|
|
376
|
+
- `blocker`, `high`, `medium`은 material issue로 파생된다
|
|
377
|
+
- `low`, `info`는 non-material finding 또는 evidence observation이다
|
|
378
|
+
- 모든 material issue는 affected purpose, failure condition, impact, evidence refs를 가져야 한다
|
|
379
|
+
- 시작 시점에는 환경, 방식, non-secret 모델/profile, 도메인, target, review direction을 짧게 제시해야 한다
|
|
380
|
+
- 긴 실행 중에는 stepwise/progress-bar 형태로 진행 상태와 새로 수집된 review 정보를 함께 제시해야 하며, 새 정보가 없더라도 bounded liveness update를 제공해야 한다
|
|
381
|
+
- issue-stage artifact가 생성되면 progress update는 finding/issue count, highest severity, material issue count 같은 새 review signal을 함께 제시해야 한다
|
|
382
|
+
- 중간 finding-like update는 `lens_local`, `issue_candidate`, `deliberation_pending`, `deliberated`, `finalized` 같은 interim signal status를 표시해야 한다
|
|
383
|
+
- halted partial result는 halt identity와 produced/absent artifact truth를 먼저 보여줘야 한다
|
|
384
|
+
- CLI가 보이지 않는 MCP/host 환경에서는 `onto.review_status` polling을 기본 경로로 host LLM presentation input을 갱신해야 한다
|
|
385
|
+
- 별도 HTML/UI 구현은 요구하지 않는다. CLI, MCP, `final-output.md`, `review-record.yaml`은 같은 bounded facts를 제시해야 한다
|
|
386
|
+
|
|
387
|
+
---
|
|
388
|
+
|
|
389
|
+
## 4. Immediate Follow-up
|
|
390
|
+
|
|
391
|
+
이 문서의 productized live path는 현재 MCP `onto.review`에서
|
|
392
|
+
`review:invoke`를 통해 `review:start-session`, `review:run-prompt-execution`,
|
|
393
|
+
`review:complete-session` bounded steps를 유지한다.
|
|
394
|
+
|
|
395
|
+
남은 follow-up은 live path 자체의 구조 변경이 아니라 운영/확장 품질이다.
|
|
396
|
+
|
|
397
|
+
1. provider credentials/endpoints가 의도적으로 준비된 환경에서 provider별 live conformance를 수행한다.
|
|
398
|
+
2. replay, stale artifact, partial worker, duplicate-dispatch semantics가 닫힌 뒤 explicit resume command를 설계한다.
|
|
@@ -0,0 +1,187 @@
|
|
|
1
|
+
# Review Prompt Execution Runner Contract
|
|
2
|
+
|
|
3
|
+
> 상태: Active
|
|
4
|
+
> 목적: `execution-plan.yaml`과 `prompt-packets/*.prompt.md`를 읽고, 각 `맥락 격리 추론 단위 (ContextIsolatedReasoningUnit)`를 deterministic하게 dispatch하는 bounded runtime contract를 고정한다.
|
|
5
|
+
> 기준 문서:
|
|
6
|
+
> - `.onto/processes/review/productized-live-path.md`
|
|
7
|
+
> - `.onto/processes/review/lens-prompt-contract.md`
|
|
8
|
+
> - `.onto/processes/review/synthesize-prompt-contract.md`
|
|
9
|
+
> - `.onto/processes/review/record-contract.md`
|
|
10
|
+
> - `.onto/processes/review/pre-dispatch-contracts.md`
|
|
11
|
+
> - `.onto/authority/core-lexicon.yaml`
|
|
12
|
+
|
|
13
|
+
---
|
|
14
|
+
|
|
15
|
+
## 1. Position
|
|
16
|
+
|
|
17
|
+
이 contract는 runtime이 `LLM` 대신 의미 판단을 수행한다는 뜻이 아니다.
|
|
18
|
+
|
|
19
|
+
runtime은 아래만 한다.
|
|
20
|
+
|
|
21
|
+
1. `execution-plan.yaml`을 읽는다
|
|
22
|
+
2. lens/deliberation/synthesize prompt packet seat를 확인한다
|
|
23
|
+
3. 각 lens packet을 병렬로 외부 실행 단위에 deterministic하게 전달한다
|
|
24
|
+
4. 각 output seat에 실제 결과 파일이 생성되었는지 검사한다
|
|
25
|
+
5. 기준 미달이면 `fail-close` 한다
|
|
26
|
+
6. `EffectiveBoundaryState`를 `error-log.md`와 degraded 판단의 구조적 basis로 남긴다
|
|
27
|
+
7. Round 1 완료 후 `lens-completion-barrier.yaml`을 작성한다
|
|
28
|
+
|
|
29
|
+
추가로 synthesize dispatch 직전에는 runtime이
|
|
30
|
+
participating lens output과 controlled deliberation result의 seat/ref를 synthesize runtime packet에 반영할 수 있다.
|
|
31
|
+
이건 새로운 semantic 판단이 아니라,
|
|
32
|
+
이미 존재하는 lens 결과 seat를 declared handoff에 맞게 전달하는 deterministic handoff다.
|
|
33
|
+
|
|
34
|
+
즉 이 runner는:
|
|
35
|
+
|
|
36
|
+
- `결정론적 계약 실행기 (deterministic contract executor)`
|
|
37
|
+
- `구조 적합성 게이트 (structural conformance gate)`
|
|
38
|
+
|
|
39
|
+
역할만 가진다.
|
|
40
|
+
|
|
41
|
+
---
|
|
42
|
+
|
|
43
|
+
## 2. Inputs
|
|
44
|
+
|
|
45
|
+
최소 입력:
|
|
46
|
+
|
|
47
|
+
1. `project_root`
|
|
48
|
+
2. `session_root`
|
|
49
|
+
3. `execution-plan.yaml`
|
|
50
|
+
4. `prompt-packets/{lens}.prompt.md`
|
|
51
|
+
5. `prompt-packets/{lens}.deliberation.prompt.md`
|
|
52
|
+
6. `prompt-packets/teamlead.deliberation.prompt.md`
|
|
53
|
+
7. `prompt-packets/synthesize.prompt.md`
|
|
54
|
+
8. executor realization
|
|
55
|
+
- `worker`
|
|
56
|
+
- `direct-call`
|
|
57
|
+
9. host runtime
|
|
58
|
+
- `codex`
|
|
59
|
+
- `openai`
|
|
60
|
+
- `anthropic`
|
|
61
|
+
- `grok`
|
|
62
|
+
- `lmstudio`
|
|
63
|
+
- `standalone`
|
|
64
|
+
10. selected lens count
|
|
65
|
+
|
|
66
|
+
중요:
|
|
67
|
+
|
|
68
|
+
- runtime은 packet 내용을 해석하지 않는다
|
|
69
|
+
- packet과 output seat를 외부 실행 단위에 전달만 한다
|
|
70
|
+
- lens dispatch order는 deterministic하게 유지하되, 실행은 bounded parallel이어야 한다
|
|
71
|
+
|
|
72
|
+
---
|
|
73
|
+
|
|
74
|
+
## 3. Outputs
|
|
75
|
+
|
|
76
|
+
최소 출력:
|
|
77
|
+
|
|
78
|
+
1. `round1/{lens}.md`
|
|
79
|
+
2. `deliberation/round1/{lens}-deliberation.md`
|
|
80
|
+
3. `deliberation.md`
|
|
81
|
+
4. `synthesis.md`
|
|
82
|
+
5. `execution-result.yaml`
|
|
83
|
+
6. `error-log.md`
|
|
84
|
+
|
|
85
|
+
원칙:
|
|
86
|
+
|
|
87
|
+
- lens output seat는 `execution-plan.yaml`이 고정한다
|
|
88
|
+
- deliberation output seat는 `execution-plan.yaml`이 고정한다
|
|
89
|
+
- synthesize output seat도 `execution-plan.yaml`이 고정한다
|
|
90
|
+
- `execution-result.yaml`은 actual execution truth의 canonical seat다
|
|
91
|
+
- `lens-completion-barrier.yaml`은 downstream stage 진입 gate다
|
|
92
|
+
- degraded case / partial failure는 `error-log.md`에 기록해야 한다
|
|
93
|
+
- `error-log.md`는 최소 한 번 `EffectiveBoundaryState`를 기록해야 한다
|
|
94
|
+
- `error-log.md`는 runner progress seat도 겸할 수 있다
|
|
95
|
+
- runner는 seat를 바꾸면 안 된다
|
|
96
|
+
- `execution-result.yaml`은 최소 아래를 담아야 한다
|
|
97
|
+
- planned/participating/degraded/excluded lens ids
|
|
98
|
+
- per-unit started/completed timestamps
|
|
99
|
+
- per-unit duration
|
|
100
|
+
- deliberation execution status
|
|
101
|
+
- synthesize execution status
|
|
102
|
+
- halt reason
|
|
103
|
+
|
|
104
|
+
---
|
|
105
|
+
|
|
106
|
+
## 4. Canonical Bounded Step
|
|
107
|
+
|
|
108
|
+
현재 TS core bounded step:
|
|
109
|
+
|
|
110
|
+
```bash
|
|
111
|
+
npm run review:run-prompt-execution -- \
|
|
112
|
+
--project-root {project_root} \
|
|
113
|
+
--session-root {session_root} \
|
|
114
|
+
--executor-bin {executor_bin} \
|
|
115
|
+
--executor-arg {executor_arg}
|
|
116
|
+
```
|
|
117
|
+
|
|
118
|
+
옵션:
|
|
119
|
+
|
|
120
|
+
- `--synthesize-executor-bin`
|
|
121
|
+
- `--synthesize-executor-arg`
|
|
122
|
+
|
|
123
|
+
를 통해 synthesize만 다른 realization으로 분리할 수 있다.
|
|
124
|
+
|
|
125
|
+
현재 repo-local actual realization 예:
|
|
126
|
+
|
|
127
|
+
```bash
|
|
128
|
+
npm run review:run-prompt-execution -- \
|
|
129
|
+
--project-root {project_root} \
|
|
130
|
+
--session-root {session_root} \
|
|
131
|
+
--executor-bin npm \
|
|
132
|
+
--executor-arg=run \
|
|
133
|
+
--executor-arg=review:codex-unit-executor \
|
|
134
|
+
--executor-arg=--
|
|
135
|
+
```
|
|
136
|
+
|
|
137
|
+
현재 구현에서 prompt execution runner를 통해 실행되는 execution profile:
|
|
138
|
+
|
|
139
|
+
- `worker + codex` (Codex CLI 경로)
|
|
140
|
+
- `worker + direct-call` (API/local provider 경로)
|
|
141
|
+
- `worker + mock` (conformance/test 경로)
|
|
142
|
+
|
|
143
|
+
원칙:
|
|
144
|
+
|
|
145
|
+
- 병렬 실행은 필수다
|
|
146
|
+
- runtime은 선택된 lens 전체를 dispatch한다
|
|
147
|
+
- host adapter가 선택된 lens 전체 dispatch를 보장할 수 없으면 fail-loud하게 중단한다
|
|
148
|
+
- controlled lens deliberation은 participating lens outputs가 확정된 뒤에만 시작한다
|
|
149
|
+
- synthesize는 `deliberation.md`가 생성된 뒤에만 시작한다
|
|
150
|
+
- issue artifacts, controlled deliberation, synthesize는
|
|
151
|
+
`.onto/processes/review/pre-dispatch-contracts.md §6`의 lens completion barrier가
|
|
152
|
+
`downstream_allowed=true`일 때만 시작한다
|
|
153
|
+
|
|
154
|
+
---
|
|
155
|
+
|
|
156
|
+
## 5. What The Runner Must Not Do
|
|
157
|
+
|
|
158
|
+
이 runner는 아래를 하면 안 된다.
|
|
159
|
+
|
|
160
|
+
1. lens 순서를 semantic하게 재조정
|
|
161
|
+
2. packet 내용을 해석해서 수정
|
|
162
|
+
3. synthesize 결과를 임의로 보정
|
|
163
|
+
4. missing output을 추론으로 보완
|
|
164
|
+
5. output이 비어 있는데도 통과
|
|
165
|
+
|
|
166
|
+
즉 이 단계는 semantic execution engine이 아니라
|
|
167
|
+
semantic execution dispatch engine이다.
|
|
168
|
+
|
|
169
|
+
packet은 가능하면 lightweight해야 한다.
|
|
170
|
+
runtime은 packet을 giant embedded payload로 만들기보다,
|
|
171
|
+
authoritative artifact path와 output seat를 고정하는 쪽을 우선한다.
|
|
172
|
+
|
|
173
|
+
또한 runner는 boundary seat를 semantic하게 재해석하지 않는다.
|
|
174
|
+
다만 아래는 해야 한다.
|
|
175
|
+
|
|
176
|
+
1. `EffectiveBoundaryState`를 log basis로 남긴다
|
|
177
|
+
2. 경계 제약 아래에서 output이 생성되지 않았을 때 degraded/fail-close 경로를 탄다
|
|
178
|
+
|
|
179
|
+
---
|
|
180
|
+
|
|
181
|
+
## 6. Immediate Follow-up
|
|
182
|
+
|
|
183
|
+
다음 단계는 아래다.
|
|
184
|
+
|
|
185
|
+
1. `review:start-session -> review:run-prompt-execution -> review:complete-session`를 `onto.review`의 canonical bounded runtime path로 유지한다
|
|
186
|
+
2. 실제 host realization이 이 contract를 따르도록 연결한다
|
|
187
|
+
3. provider별 controlled deliberation conformance test를 추가한다
|