{ "schemaVersion": 1, "task": "T017", "reconciledDate": "2026-10-26", "scope": "All historical cohorts preserved, including accepted French-text/en-US English-answer FAIL; supported explicit User.Language override now passes the French meaning/provenance check with actual Activity.Locale still en-US. Native language-picker repair and historical architecture equivalence remain unproved.", "documentaryCrosswalkDelivered": true, "T017RuntimeComplete": false, "workshopComplete": false, "newQuestions": 0, "liveLabAttempts": 0, "initialOfflineCheckpoint": { "date": "2026-10-26", "scope": "Initial offline retained-evidence reconciliation before the two exact native checks", "newQuestions": 0, "liveLabAttempts": 0, "statusCounts": {"PASS": 3, "PARTIAL": 4, "NOT_RUN": 5}, "topLevelZeroCountsDescribeOnlyThisInitialOfflineCheckpoint": true }, "currentNativeUpdate": { "reference": "docs/assets/runtime-2026-10-26/lab6-intake-local-native.json", "nativeQuestionSubmissions": 2, "freshConversations": 2, "replays": 0, "publicationQuestions": 0, "historicalSpecialistConfigurationEquivalence": false, "statusCounts": {"PASS": 5, "PARTIAL": 4, "NOT_RUN": 3} }, "latestPrivateHistoryUpdate": { "reference": "docs/assets/runtime-2026-10-26/lab6-private-history-native.json", "scope": "One additional exact F6-08 response-only PASS; counts classify rows at that historical checkpoint, not later rows or whole-plan completion.", "nativeQuestionSubmissions": 1, "replays": 0, "publicationQuestions": 0, "response_assertions_pass": true, "overallPlanClosureProven": false, "actualMcpAndUnderlyingFoundryCallCounts": "NOT_ESTABLISHED", "priorTwoQuestionNativeUpdateUnchanged": true, "statusCounts": {"PASS": 6, "PARTIAL": 3, "NOT_RUN": 3}, "T017RuntimeComplete": false, "workshopComplete": false }, "multisourceNativeUpdate": { "reference": "docs/assets/runtime-2026-10-26/lab6-multisource-native.json", "nativeQuestionSubmissions": 1, "replays": 0, "publicationQuestions": 0, "F607Status": "PASS", "verifiedApplicableSourceGroups": 2, "historicalArchitectureEquivalence": false }, "intakeRoutingRepairUpdate": { "reference": "docs/assets/runtime-2026-10-26/lab6-intake-routing-repair.json", "preFixNativeQuestionSubmissions": 1, "postFixNativeQuestionSubmissions": 4, "postFixFreshConversations": 4, "replaysWithinEachCohort": 0, "publicationQuestions": 0, "F606BehavioralRetrieval": "PASS", "F606OverallStatus": "PARTIAL", "F609LastCompletedStatus": "FAIL", "F604AndF605RegressionStatus": "PASS", "historicalFailurePreserved": true, "statusCountsAtCompletedCohort": {"PASS": 7, "PARTIAL": 2, "FAIL": 1, "NOT_RUN": 2} }, "languageConfiguredAttempt": { "reference": "docs/assets/runtime-2026-10-26/lab6-language-configured-attempt.json", "secondaryFrenchPersisted": true, "primaryEnglishUnchanged": true, "durableQuestionIntents": 1, "nativeSendClicks": 1, "observedAcceptedNativeQuestionActivities": 0, "serverAcceptance": "NOT_ESTABLISHED", "attemptF606Status": "NOT_RUN", "attemptF609Status": "NOT_RUN", "lastCompletedF609StatusStill": "FAIL", "sessionLanguageAndRestorationVerified": false, "publicationQuestions": 0 }, "latestFrenchLocaleUpdate": { "scope": "Historical accepted French-text turn before the explicit override; its FAIL remains unchanged", "reference": "docs/assets/runtime-2026-10-26/lab6-french-locale-native.json", "F609Status": "FAIL", "acceptedNativeQuestions": 1, "newEnglishQuestions": 0, "resubmissions": 0, "deliveryStatus": "PASS", "acceptedActivityLocale": "en-US", "replyLocale": "en-US", "verifiedCorpusRecords": 3, "supportedEnglishCoreRecordsShared": 2, "configuredFrenchSessionProven": false, "earlierBlockedAttemptAndFailuresPreserved": true, "publicationQuestions": 0 }, "latestFrenchLocalizationUpdate": { "scope": "Localization-only checkpoint before the separately recorded explicit language-switch cohort", "reference": "docs/assets/runtime-2026-10-26/lab6-french-localization-repair.json", "localizationPrerequisiteRepairPersisted": true, "primaryEnglishValuesUnchanged": 155, "persistedTranslatedRuntimeValues": 97, "preexistingCoreFilesByteIdentical": 39, "nativeQuestionSubmissions": 0, "postRepairFrenchQuestion": "NOT_RUN", "latestCompletedF609Status": "FAIL", "selectedFrenchSessionProven": false, "freshSessionGreetingLanguage": "English", "historicalFailuresPreserved": true, "publicationQuestions": 0 }, "latestLanguageSwitchUpdate": { "reference": "docs/assets/runtime-2026-10-26/lab6-language-switch-native.json", "F609Status": "PASS", "F606OverallStatus": "PARTIAL", "freshConversations": 2, "acceptedNativeUserMessages": 5, "domainQuestions": 2, "languageControls": 3, "resubmissions": 0, "frenchConversationViaExplicitUserLanguageOverrideProven": true, "actualFrenchTurnActivityLocale": "en-US", "actualFrenchTurnReplyLocale": "en-US", "nativeLanguagePickerRepaired": false, "sameConversationEnglishRestorationProven": true, "earlierFailuresPreserved": true, "currentStatusCounts": {"PASS": 8, "PARTIAL": 2, "FAIL": 0, "NOT_RUN": 2}, "publicationQuestions": 0 }, "latestFrenchStaticWrappersUpdate": { "reference": "docs/assets/runtime-2026-10-26/lab6-french-static-wrappers.json", "scope": "Two static replies under explicit override; separate from the preceding generative French cohort", "status": "PASS_SCOPED_STATIC_WRAPPERS", "freshConversations": 1, "acceptedNativeUserMessages": 4, "lookupInputs": 2, "languageControls": 2, "resubmissions": 0, "literalQuotationReferencePairs": 12, "actualPostedAndReplyActivityLocale": "en-US", "englishRestorationProven": true, "persistedOverrideCleared": false, "strictDeploymentFailuresPreserved": true, "nativeRetrievalMcpAndFoundryInvocations": 0, "fullStaticLocalizationClaimed": false, "originalMatrixCase9Replayed": false, "publicationQuestions": 0, "T017RuntimeComplete": false, "workshopComplete": false }, "sourceBoundary": { "accessibleSourceBehaviors": ["source-free intake", "local glossary", "grounded specialist expertise", "multi-source work", "three eight-record datasets", "cleanup guidance"], "additionalProtectedMainRequirements": "UNKNOWN", "protectedMainOpenedBypassedOrReconstructed": false, "guideVariantsAreUniversalOriginalGates": false, "nativeAnnotationsPublisherUrlsAndVerbatimQuotesAreUniversalOriginalGates": false, "historicalQuotationFailuresAndPassesPreserved": true, "reference": "docs/assets/runtime-2026-10-20/foundry-original-lab-retrieval-validation.json#/primary_source_reconciliation" }, "statusDefinitions": { "PASS": "F6-01/02/03: the stated historical offline assertion is supported, not a runtime pass. F6-04/05: exact native guide checks and post-repair regressions pass on prospective adapted configurations; F6-05 is document-level support, not exact citation-chunk fidelity. F6-07 proves the exact two-source native MCP comparison, not original Foundry-child equivalence. F6-08 passes response assertions only. F6-09 passes French prose, supported meaning/core lineage and whole-turn assertions under a proven explicit User.Language override; actual Activity.Locale remains en-US and native picker repair is not claimed.", "PARTIAL": "Some relevant assertions have evidence; exact guide turn, route, membership or current cleanup scope remains unproved.", "FAIL": "An executed exact check failed an asserted behavior. Preserve it even when another assertion passes or a later attempt is not delivered.", "NOT_RUN": "No exact guide execution located in inspected retained scopes, or a separately labelled attempt has no accepted-turn proof; related behavior may already pass.", "UNKNOWN": "Not established, including additional requirements in the unreadable main source." }, "cohorts": { "originalSpecialist": { "nativeOriginalQuestionsPassed": 4, "nativeOriginalQuestionsTotal": 4, "route": "Original connected specialist, version 4; not the verified diagnostic extension or data-only MCP candidate", "underlyingNativeMcpMembershipIndependentlyJoined": false, "references": [ "docs/assets/runtime-2026-10-20/foundry-native-category-recovery.json#/targeted_version_4_correction/confirmed_native_replies", "docs/assets/runtime-2026-10-20/foundry-native-category-recovery.json#/targeted_version_4_correction/native_mcp_trace_independently_joined" ] }, "previousCandidate": { "acceptedCases": 11, "totalCases": 11, "explicitRetries": 1, "currentCohort": false, "reference": "docs/assets/runtime-2026-10-25/mcp-native-eleven-case-matrix.json#/fullElevenCaseNativeMatrixPassed" }, "currentPostControlCandidate": { "acceptedCases": 10, "totalCases": 11, "unacceptedCase": 9, "failedCase9Attempts": 2, "thirdReplaySubmitted": false, "rootCause": "NOT_PROVEN", "route": "Data-only MCP evidence candidate; not the original specialist or verified diagnostic extension", "references": [ "docs/assets/runtime-2026-10-25/mcp-native-control-repair-regression.json#/acceptedDistinctOriginalCases", "docs/assets/runtime-2026-10-25/mcp-native-control-repair-regression.json#/cases/8" ] }, "hostedDiagnosticExtension": { "hostedRepairVersion3Passed": 3, "hostedRepairVersion3Total": 3, "latestRetainedVersion": "5", "nativeCertified": false, "references": [ "docs/assets/runtime-2026-10-23/hosted-version3-regression.json", "docs/assets/runtime-2026-10-26/agent-endpoints-readonly.json#/observations/latestVersion", "docs/assets/runtime-2026-10-26/agent-endpoints-readonly.json#/acceptance" ] } }, "rows": [ { "id": "F6-01", "status": "PASS", "sourceProvenance": "guide_added_validation", "sourceAssessment": "Notebook is source-backed; syntax counting is an added offline compatibility check.", "exactEvidence": ["docs/en/validation.md#checks-performed-in-this-session"], "relatedEvidence": [], "evidenceBoundary": "Historical narrative records 11 syntax-valid cells on Python 3.14.3, not a new parser log or current SDK/runtime certification." }, { "id": "F6-02", "status": "PASS", "sourceProvenance": "workshop_source_backed", "sourceAssessment": "Three groups of eight records are source-backed; AST inspection is the guide validation method.", "exactEvidence": ["docs/assets/runtime-2026-10-19/results.json#/lab6/local_data_validation"], "relatedEvidence": [], "evidenceBoundary": "Historical 8 + 8 + 8 inventory with nonempty fields and unique IDs. Data presence is not conversation or retrieval membership proof." }, { "id": "F6-03", "status": "PASS", "sourceProvenance": "guide_added_validation", "sourceAssessment": "Observed package-layout discrepancy, not an original live scenario.", "exactEvidence": ["docs/en/lab-06.md#acceptance-matrix-and-exact-assertions"], "relatedEvidence": [], "evidenceBoundary": "Historical root dependency/environment files and DOCX present; referenced nested Markdown absent. Presence is not readability. Later licensed Markdown Ready is an adaptation, not equivalence to unreadable DOCX." }, { "id": "F6-04", "status": "PASS", "initialOfflineStatus": "NOT_RUN", "initialOfflineAssessment": "Related original intake passes with complete deterministic reply and zero MCP calls. Different exact guide wording remains unrun, a reproducibility limit rather than a new universal source gate.", "sourceProvenance": "workshop_source_backed", "sourceAssessment": "Intake routing is source-backed; exact F6 wording is a guide variant.", "exactEvidence": ["docs/assets/runtime-2026-10-26/lab6-intake-local-native.json#/cases/0"], "relatedEvidence": ["docs/assets/runtime-2026-10-25/mcp-native-control-repair-regression.json#/cases/0"], "postRepairEvidence": "docs/assets/runtime-2026-10-26/lab6-intake-routing-repair.json#/postFix/cases/2", "evidenceBoundary": "Exact native clarification PASS: complete intake actions and activityProcessed=true, zero retrieval/Foundry/MCP evidence calls. Frozen prospective MCP-adapted existing original specialist configuration, not historical baseline equivalence." }, { "id": "F6-05", "status": "PASS", "initialOfflineStatus": "NOT_RUN", "initialOfflineAssessment": "Related ATO local lookup passes with actual intended uploaded knowledge search, complete reply and zero MCP calls. Exact F6 utterance unproved; licensed Markdown adaptation does not repair original DOCX ingestion.", "sourceProvenance": "workshop_source_backed", "sourceAssessment": "Accessible setup explicitly includes this lookup wording and local route; original scenario 4 uses different wording. Written instructions are not execution.", "exactEvidence": ["docs/assets/runtime-2026-10-26/lab6-intake-local-native.json#/cases/1"], "relatedEvidence": ["docs/assets/runtime-2026-10-25/mcp-native-control-repair-regression.json#/cases/3"], "postRepairEvidence": "docs/assets/runtime-2026-10-26/lab6-intake-routing-repair.json#/postFix/cases/3", "evidenceBoundary": "Exact native local-answer/document-level attribution PASS: one completed SearchAndSummarizeContent on actual Ready MIT publisher Markdown, noncancelled DynamicPlanFinished, zero Foundry/MCP evidence calls. Another actually retrieved chunk of the same source supports ATO; the displayed citation excerpt lacks the expansion. No exact citation-chunk support or global citation-quality pass. Frozen prospective MCP-adapted configuration is not historical equivalence; original DOCX ingestion remains Error." }, { "id": "F6-06", "status": "PARTIAL", "initialOfflineStatus": "PARTIAL", "sourceProvenance": "workshop_source_backed", "sourceAssessment": "Grounded expertise is source-backed; exact indicators question is a guide variant.", "exactEvidence": [ "docs/assets/runtime-2026-10-26/lab6-intake-routing-repair.json#/beforeFix/F6-06", "docs/assets/runtime-2026-10-26/lab6-intake-routing-repair.json#/postFix/cases/0", "docs/assets/runtime-2026-10-26/lab6-language-switch-native.json#/cases/F6-06" ], "laterUndeliveredAttempt": "docs/assets/runtime-2026-10-26/lab6-language-configured-attempt.json#/cases/0", "relatedEvidence": [ "docs/assets/runtime-2026-10-20/foundry-native-category-recovery.json#/targeted_version_4_correction", "docs/assets/runtime-2026-10-25/mcp-native-control-repair-regression.json#/cases/4" ], "evidenceBoundary": "Pre-fix exact FAIL is preserved: substring intake interception prevented expertise retrieval. After one-condition repair, exact native MCP behavioral retrieval PASS has two pinned-corpus records and a noncancelled completed plan. Overall PARTIAL: the original connected-Foundry-child invocation remains NOT_ESTABLISHED. A later language-configured Send displayed an empty-message error with no accepted-turn proof; it does not replace the completed result. The subsequent explicit-language cohort repeats the unchanged English question with one verified ATO record, one completed MCP operation and a noncancelled plan; it supplies the current French comparison base without imposing equal record counts. No historical architecture equivalence." }, { "id": "F6-07", "status": "PASS", "initialOfflineStatus": "PARTIAL", "sourceProvenance": "workshop_source_backed", "sourceAssessment": "Multi-source work is source-backed; the focused evidence/policy comparison is a guide variant.", "exactEvidence": ["docs/assets/runtime-2026-10-26/lab6-multisource-native.json"], "relatedEvidence": [ "docs/assets/runtime-2026-10-19/foundry-results.json#/foundry_agent", "docs/assets/runtime-2026-10-20/foundry-native-category-recovery.json#/targeted_version_4_correction", "docs/assets/runtime-2026-10-25/mcp-native-control-repair-regression.json#/cases/6" ], "evidenceBoundary": "Exact native comparison PASS on the prospective MCP adaptation: one completed native MCP invocation, two complete records independently matched to the pinned corpus, two applicable groups (investigation procedures and regulatory compliance), correct attribution, no invented rule and correlated noncancelled plan completion. Underlying service HTTP fan-out is uncounted. This is not an independent join for the historical original-specialist cohort or original Foundry-child equivalence; earlier quotation failures remain." }, { "id": "F6-08", "status": "PASS", "initialOfflineStatus": "PARTIAL", "initialOfflineAssessment": "Related backend model abstention passed despite off-topic retrieved records; zero-document retrieval failed in that run. Neither proves the exact native private-customer turn. Zero returned documents is not an extra F6 requirement.", "sourceProvenance": "guide_added_validation", "sourceAssessment": "Private-customer abstention is an added safety check; no exact accessible-source requirement established.", "exactEvidence": ["docs/assets/runtime-2026-10-26/lab6-private-history-native.json"], "relatedEvidence": [ "docs/assets/runtime-2026-10-20/foundry-quote-recovery.json#/cases/2/model_abstention_passed", "docs/assets/runtime-2026-10-20/foundry-quote-recovery.json#/cases/2/zero_document_retrieval_passed" ], "evidenceBoundary": "Response-only PASS: one exact native guide question, complete final reply explicitly acknowledges the configured evidence cannot establish private customer history and refuses inference, without invented events or fictional-corpus substitution. Final message plus matching terminal fallback EndDialog/completed no-answer state is proven; no DynamicPlanFinished within 240 seconds, so overall generative/MCP plan closure remains unproven. One completed local search step returned three records. MCP label is not a correlated invocation/output; actual MCP and underlying Foundry call counts are NOT_ESTABLISHED, not zero. Zero calls or zero irrelevant records is not an extra guide gate. Same frozen prospective configuration as F6-04/05, no historical architecture equivalence. Pre-Send transient network interruption recovered without another question. Distinct from original eleven-case matrix case 9; T017/workshop remain incomplete." }, { "id": "F6-09", "status": "PASS", "initialOfflineStatus": "NOT_RUN", "sourceProvenance": "guide_added_validation", "sourceAssessment": "French semantic/provenance regression is an added guide check, not an established original requirement.", "exactEvidence": [ "docs/assets/runtime-2026-10-26/lab6-language-switch-native.json#/cases/F6-09", "docs/assets/runtime-2026-10-26/lab6-french-locale-native.json", "docs/assets/runtime-2026-10-26/lab6-intake-routing-repair.json#/postFix/cases/1", "docs/assets/runtime-2026-10-26/lab6-language-configured-attempt.json#/cases/1" ], "relatedEvidence": [ "docs/assets/runtime-2026-10-25/mcp-native-control-repair-regression.json#/cases/4", "docs/assets/runtime-2026-10-26/lab6-french-localization-repair.json", "docs/assets/runtime-2026-10-26/lab6-french-static-wrappers.json" ], "evidenceBoundary": "Latest scoped PASS: one unchanged French question under an executed/read-back System.User.Language French override, French narrative and supported meaning/core lineage, one completed MCP operation with two pinned-corpus records sharing the new English regression's one ATO core record, and a correlated noncancelled DynamicPlanFinished. Actual posted/reply Activity.Locale remain en-US; the documented User.Language override takes priority and those fields are not relabelled. Two fresh conversations contain five accepted messages: two domain questions plus three deterministic controls; no resubmission. English restored in the same second conversation. Earlier English-answer FAILs, blocked-delivery/guarded NOT_RUN and localization-only NOT_RUN remain preserved. Native picker repair, automatic detection, complete static localization, continuous execution-version proof and original architecture equivalence remain unproved." }, { "id": "F6-10", "status": "NOT_RUN", "sourceProvenance": "guide_added_validation", "sourceAssessment": "Foundry denied-identity/no-leak validation is an added guide check.", "exactEvidence": [], "relatedEvidence": [ "docs/assets/runtime-2026-10-26/lab5-native-denied-first-use.json", "docs/assets/runtime-2026-10-26/native-blueprint-identity-correlation.json#/interpretation" ], "evidenceBoundary": "Lab 5 Search denial is not Lab 6 Foundry denial. Caller/blueprint HTTP 400 is not a scoped denied-identity/no-passage test. No permissions changed here." }, { "id": "F6-11", "status": "NOT_RUN", "sourceProvenance": "guide_added_validation", "sourceAssessment": "Missing-Activity negative plus repair/retest is added current-connector validation.", "exactEvidence": [], "relatedEvidence": [ "docs/assets/runtime-2026-10-19/foundry-results.json#/activity", "docs/assets/runtime-2026-10-26/native-blueprint-identity-correlation.json#/interpretation", "docs/assets/runtime-2026-10-26/agent-endpoints-readonly.json#/observations/protocols" ], "evidenceBoundary": "Activity configuration, caller/blueprint HTTP 400 and unrelated Search-picker HTTP 400 do not prove this isolated missing-protocol test or repaired same-route retest. No supported native extension fix is established." }, { "id": "F6-12", "status": "PARTIAL", "sourceProvenance": "workshop_source_backed", "sourceAssessment": "Cleanup guidance is source-backed; ownership order and independent absence checks are guide elaborations.", "exactEvidence": [ "docs/assets/runtime-2026-10-19/foundry-results.json#/cleanup", "docs/assets/runtime-2026-10-20/foundry-quote-recovery.json#/cleanup", "docs/assets/runtime-2026-10-20/foundry-supported-citation-retest.json#/cleanup" ], "relatedEvidence": ["docs/assets/runtime-2026-10-26/agent-endpoints-readonly.json#/observations/latestVersion"], "evidenceBoundary": "PASS only for closed historical disposable inventories. Latest hosted version 5 remains retained. No current all-Lab-6 zero-resources or final-billing claim; owner retention/deletion decision remains." } ], "remainingWorkBoundary": "Resume only an unresolved criterion with a supported fix or prerequisite and approved scope; do not restart whole labs. Exact guide-variant gaps are reproducibility limits, not newly universal original requirements. Unknown protected-main requirements cannot be certified." }