SILVER/NODEDC_1C
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

ARCH: усилить metadata ambiguity choice sets в lane arbitration

Browse Source
This commit is contained in:
dctouch
2026-04-22 09:45:49 +03:00
parent d9454fcdef
commit c328c52c9b
6 changed files with 282 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files
llm_normalizer/backend/dist/services/assistantMcpDiscoveryTurnInputAdapter.js
Vendored
+61 -14
View File
@@ -228,9 +228,22 @@ function collectFollowupDiscoverySeed(followupContext) {
dateScope,
metadataRouteFamily: toNonEmptyString(followupContext?.previous_discovery_metadata_route_family),
metadataSelectedEntitySet: toNonEmptyString(followupContext?.previous_discovery_metadata_selected_entity_set),
metadataAmbiguityDetected: followupContext?.previous_discovery_metadata_ambiguity_detected === true
metadataAmbiguityDetected: followupContext?.previous_discovery_metadata_ambiguity_detected === true,
metadataAmbiguityEntitySets: collectEntityCandidates(followupContext?.previous_discovery_metadata_ambiguity_entity_sets)
};
}
function metadataEntitySetsSuggestDocumentLane(values) {
return values.some((value) => /(?:документ|document|invoice|waybill|накладн|счет[- ]?фактур|акт)/iu.test(value));
}
function metadataEntitySetsSuggestMovementLane(values) {
return values.some((value) => /(?:регистр|register|movement|движени|операц|проводк|bank)/iu.test(value));
}
function metadataAmbiguityCollapsesToDocumentLane(values) {
return values.length > 0 && metadataEntitySetsSuggestDocumentLane(values) && !metadataEntitySetsSuggestMovementLane(values);
}
function metadataAmbiguityCollapsesToMovementLane(values) {
return values.length > 0 && metadataEntitySetsSuggestMovementLane(values) && !metadataEntitySetsSuggestDocumentLane(values);
}
function hasLifecycleSignal(text) {
return /(?:сколько\s+лет|как\s+давно|давно\s+ли|возраст|перв(?:ая|ый)\s+актив|когда\s+начал|когда\s+появ|lifecycle|activity\s+duration|business\s+age|how\s+long)/iu.test(text);
}
@@ -257,10 +270,10 @@ function hasMetadataObjectHint(text) {
return /(?:\u0440\u0435\u0433\u0438\u0441\u0442\u0440(?:\u044b)?|\u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442(?:\u044b)?|\u0441\u043f\u0440\u0430\u0432\u043e\u0447\u043d\u0438\u043a(?:\u0438)?|\u043f\u043e\u043b(?:\u0435|\u044f)|registers?|documents?|catalogs?|fields?)/iu.test(text);
}
function hasDocumentEvidenceFollowupSignal(text) {
return /(?:\u043f\u043e\s+\u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442(?:\u0430\u043c|\u044b)?|\u0434\u0430\u0432\u0430\u0439\s+\u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442(?:\u044b)?|\u0438\u0449\u0438\s+\u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442(?:\u044b)?|\u043f\u043e\u043a\u0430\u0436\u0438\s+\u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442(?:\u044b)?|document(?:s)?\s+(?:then|next)?|(?:then|next)\s+documents?|go\s+to\s+documents?)/iu.test(text);
return /(?:\u043f\u043e\s+\u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442(?:\u0430\u043c|\u044b)?|\u0434\u0430\u0432\u0430\u0439\s+\u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442(?:\u044b)?|\u0438\u0449\u0438\s+\u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442(?:\u044b)?|\u043f\u043e\u043a\u0430\u0436\u0438\s+\u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442(?:\u044b)?|(?:\u043f\u043e\u043a\u0430\u0436\u0438|\u043a\u0430\u043a\u0438\u0435|\u0441\u043f\u0438\u0441\u043e\u043a|\u0434\u0430\u0439|\u0438\u0449\u0438)\s+(?:\u0441\u0447(?:[еe]т|\u0435\u0442)[-\u2011 ]?\u0444\u0430\u043a\u0442\u0443\u0440(?:\u044b|\u0430)?|\u043d\u0430\u043a\u043b\u0430\u0434\u043d(?:\u044b\u0435|\u0430\u044f)?|\u0430\u043a\u0442(?:\u044b)?|\u0440\u0435\u0430\u043b\u0438\u0437\u0430\u0446(?:\u0438\u0438|\u0438\u044e)|invoice(?:s)?|bill(?:s)?|waybill(?:s)?)|document(?:s)?\s+(?:then|next)?|(?:then|next)\s+documents?|go\s+to\s+documents?)/iu.test(text);
}
function hasMovementEvidenceFollowupSignal(text) {
return /(?:\u043f\u043e\s+\u0434\u0432\u0438\u0436\u0435\u043d(?:\u0438\u044f\u043c|\u0438\u044f)?|\u0434\u0430\u0432\u0430\u0439\s+\u0434\u0432\u0438\u0436\u0435\u043d(?:\u0438\u044f|\u0438\u0435)|\u0438\u0449\u0438\s+\u0434\u0432\u0438\u0436\u0435\u043d(?:\u0438\u044f|\u0438\u0435)|\u043f\u043e\u043a\u0430\u0436\u0438\s+\u0434\u0432\u0438\u0436\u0435\u043d(?:\u0438\u044f|\u0438\u0435)|\u0431\u0430\u043d\u043a\u043e\u0432\u0441\u043a(?:\u0438\u0435|\u0438\u0439)\s+\u0434\u0432\u0438\u0436\u0435\u043d(?:\u0438\u044f|\u0438\u0435)|movement(?:s)?\s+(?:then|next)?|(?:then|next)\s+movements?|go\s+to\s+movements?)/iu.test(text);
return /(?:\u043f\u043e\s+\u0434\u0432\u0438\u0436\u0435\u043d(?:\u0438\u044f\u043c|\u0438\u044f)?|\u0434\u0430\u0432\u0430\u0439\s+\u0434\u0432\u0438\u0436\u0435\u043d(?:\u0438\u044f|\u0438\u0435)|\u0438\u0449\u0438\s+\u0434\u0432\u0438\u0436\u0435\u043d(?:\u0438\u044f|\u0438\u0435)|\u043f\u043e\u043a\u0430\u0436\u0438\s+\u0434\u0432\u0438\u0436\u0435\u043d(?:\u0438\u044f|\u0438\u0435)|\u0431\u0430\u043d\u043a\u043e\u0432\u0441\u043a(?:\u0438\u0435|\u0438\u0439)\s+\u0434\u0432\u0438\u0436\u0435\u043d(?:\u0438\u044f|\u0438\u0435)|(?:\u043f\u043e\u043a\u0430\u0436\u0438|\u043a\u0430\u043a\u0438\u0435|\u0441\u043f\u0438\u0441\u043e\u043a|\u0434\u0430\u0439|\u0438\u0449\u0438)\s+(?:\u043f\u043b\u0430\u0442[еe]\u0436(?:\u0438|\u0438)?|\u043e\u043f\u0435\u0440\u0430\u0446(?:\u0438\u0438|\u0438\u044e)|\u043f\u0440\u043e\u0432\u043e\u0434\u043a(?:\u0438|\u0430)|\u0441\u043f\u0438\u0441\u0430\u043d(?:\u0438\u044f|\u0438\u0435)|\u043f\u043e\u0441\u0442\u0443\u043f\u043b\u0435\u043d(?:\u0438\u044f|\u0438\u0435)|payment(?:s)?|transaction(?:s)?|operation(?:s)?|posting(?:s)?|bank\s+operation(?:s)?)|movement(?:s)?\s+(?:then|next)?|(?:then|next)\s+movements?|go\s+to\s+movements?)/iu.test(text);
}
function hasMetadataDownstreamContinuationSignal(text) {
return /(?:\u0434\u0430\u0432\u0430\u0439\s+\u0434\u0430\u043b\u044c\u0448\u0435|\u0438\u0434(?:\u0435|\u0451)\u043c\s+\u0434\u0430\u043b\u044c\u0448\u0435|\u043f\u043e\u0448\u043b(?:\u0438|\u0451\u043c)\s+\u0434\u0430\u043b\u044c\u0448\u0435|\u043f\u0440\u043e\u0434\u043e\u043b\u0436\u0430\u0439|\u0438\u0449\u0438\s+\u0434\u0430\u043b\u044c\u0448\u0435|\u0438\u0449\u0438\s+\u0434\u0430\u043d\u043d\u044b\u0435|\u043f\u043e\u043a\u0430\u0436\u0438\s+\u0434\u0430\u043d\u043d\u044b\u0435|\u043f\u043e\u043a\u0430\u0436\u0438\s+\u0441\u0442\u0440\u043e\u043a\u0438|\u0433\u043b\u0443\u0431\u0436\u0435|\u0447\u0442\u043e\s+\u0434\u0430\u043b\u044c\u0448\u0435|continue|go\s+ahead|go\s+deeper|look\s+deeper|drill\s+down|show\s+(?:data|rows))/iu.test(text);
@@ -351,6 +364,8 @@ function buildAssistantMcpDiscoveryTurnInput(input) {
const monthlyAggregationSignal = hasMonthlyAggregationSignal(rawText);
const explicitDateScopeLiteralDetected = hasExplicitDateScopeLiteral(rawText);
const rawDateScope = collectDateScopeFromRawText(rawText);
const metadataDocumentHintSignal = hasDocumentEvidenceFollowupSignal(rawText);
const metadataMovementHintSignal = hasMovementEvidenceFollowupSignal(rawText);
const rawDomain = toNonEmptyString(assistantTurnMeaning?.asked_domain_family);
const rawAction = toNonEmptyString(assistantTurnMeaning?.asked_action_family);
const rawAggregationAxis = toNonEmptyString(assistantTurnMeaning?.asked_aggregation_axis);
@@ -373,26 +388,28 @@ function buildAssistantMcpDiscoveryTurnInput(input) {
followupSeed.counterparty &&
!rawLifecycleSignal &&
!rawValueFlowSignal &&
hasDocumentEvidenceFollowupSignal(rawText));
metadataDocumentHintSignal);
const metadataAmbiguityResolvedDocumentFollowupApplicable = Boolean(followupSeed.pilotScope === "metadata_inspection_v1" &&
followupSeed.metadataAmbiguityDetected &&
(followupSeed.metadataAmbiguityEntitySets.length === 0 ||
metadataEntitySetsSuggestDocumentLane(followupSeed.metadataAmbiguityEntitySets)) &&
followupSeed.counterparty &&
!rawLifecycleSignal &&
!rawValueFlowSignal &&
hasDocumentEvidenceFollowupSignal(rawText));
metadataDocumentHintSignal);
const metadataGroundedMovementFollowupApplicable = Boolean(followupSeed.pilotScope === "metadata_inspection_v1" &&
followupSeed.metadataRouteFamily === "movement_evidence" &&
!followupSeed.metadataAmbiguityDetected &&
followupSeed.counterparty &&
!rawLifecycleSignal &&
!rawValueFlowSignal &&
hasMovementEvidenceFollowupSignal(rawText));
metadataMovementHintSignal);
const metadataAmbiguityResolvedMovementFollowupApplicable = Boolean(followupSeed.pilotScope === "metadata_inspection_v1" &&
followupSeed.metadataAmbiguityDetected &&
(followupSeed.metadataAmbiguityEntitySets.length === 0 ||
metadataEntitySetsSuggestMovementLane(followupSeed.metadataAmbiguityEntitySets)) &&
followupSeed.counterparty &&
!rawLifecycleSignal &&
!rawValueFlowSignal &&
hasMovementEvidenceFollowupSignal(rawText));
metadataMovementHintSignal);
const metadataGroundedLaneContinuationApplicable = Boolean(followupSeed.pilotScope === "metadata_inspection_v1" &&
(followupSeed.metadataRouteFamily === "document_evidence" ||
followupSeed.metadataRouteFamily === "movement_evidence") &&
@@ -401,15 +418,37 @@ function buildAssistantMcpDiscoveryTurnInput(input) {
!rawLifecycleSignal &&
!rawValueFlowSignal &&
!rawMetadataSignal &&
!hasDocumentEvidenceFollowupSignal(rawText) &&
!hasMovementEvidenceFollowupSignal(rawText) &&
!metadataDocumentHintSignal &&
!metadataMovementHintSignal &&
hasMetadataDownstreamContinuationSignal(rawText));
const metadataAmbiguityCollapsedDocumentLaneContinuationApplicable = Boolean(followupSeed.pilotScope === "metadata_inspection_v1" &&
followupSeed.metadataAmbiguityDetected &&
metadataAmbiguityCollapsesToDocumentLane(followupSeed.metadataAmbiguityEntitySets) &&
followupSeed.counterparty &&
!rawLifecycleSignal &&
!rawValueFlowSignal &&
!rawMetadataSignal &&
!metadataDocumentHintSignal &&
!metadataMovementHintSignal &&
hasMetadataDownstreamContinuationSignal(rawText));
const metadataAmbiguityCollapsedMovementLaneContinuationApplicable = Boolean(followupSeed.pilotScope === "metadata_inspection_v1" &&
followupSeed.metadataAmbiguityDetected &&
metadataAmbiguityCollapsesToMovementLane(followupSeed.metadataAmbiguityEntitySets) &&
followupSeed.counterparty &&
!rawLifecycleSignal &&
!rawValueFlowSignal &&
!rawMetadataSignal &&
!metadataDocumentHintSignal &&
!metadataMovementHintSignal &&
hasMetadataDownstreamContinuationSignal(rawText));
const metadataGroundedDocumentLaneApplicable = metadataGroundedDocumentFollowupApplicable ||
metadataAmbiguityResolvedDocumentFollowupApplicable ||
(metadataGroundedLaneContinuationApplicable && followupSeed.metadataRouteFamily === "document_evidence");
(metadataGroundedLaneContinuationApplicable && followupSeed.metadataRouteFamily === "document_evidence") ||
metadataAmbiguityCollapsedDocumentLaneContinuationApplicable;
const metadataGroundedMovementLaneApplicable = metadataGroundedMovementFollowupApplicable ||
metadataAmbiguityResolvedMovementFollowupApplicable ||
(metadataGroundedLaneContinuationApplicable && followupSeed.metadataRouteFamily === "movement_evidence");
(metadataGroundedLaneContinuationApplicable && followupSeed.metadataRouteFamily === "movement_evidence") ||
metadataAmbiguityCollapsedMovementLaneContinuationApplicable;
const effectiveMetadataFollowupSeedApplicable = metadataFollowupSeedApplicable &&
!metadataGroundedDocumentLaneApplicable &&
!metadataGroundedMovementLaneApplicable;
@@ -437,7 +476,9 @@ function buildAssistantMcpDiscoveryTurnInput(input) {
const lifecycleSignal = rawLifecycleSignal || seededDomain === "counterparty_lifecycle";
const bidirectionalValueFlowSignal = !lifecycleSignal &&
(rawBidirectionalValueFlowSignal || seededAction === "net_value_flow");
const valueFlowSignal = !lifecycleSignal && (rawValueFlowSignal || seededDomain === "counterparty_value");
const valueFlowSignal = !lifecycleSignal &&
!metadataGroundedMovementLaneApplicable &&
(rawValueFlowSignal || seededDomain === "counterparty_value");
const payoutSignal = valueFlowSignal &&
!bidirectionalValueFlowSignal &&
(rawPayoutSignal || seededAction === "payout");
@@ -616,6 +657,12 @@ function buildAssistantMcpDiscoveryTurnInput(input) {
if (metadataGroundedLaneContinuationApplicable) {
pushReason(reasonCodes, "mcp_discovery_metadata_grounded_lane_continuation");
}
if (metadataAmbiguityCollapsedDocumentLaneContinuationApplicable) {
pushReason(reasonCodes, "mcp_discovery_metadata_ambiguity_collapsed_to_document_lane");
}
if (metadataAmbiguityCollapsedMovementLaneContinuationApplicable) {
pushReason(reasonCodes, "mcp_discovery_metadata_ambiguity_collapsed_to_movement_lane");
}
if (unsupported) {
pushReason(reasonCodes, "mcp_discovery_unsupported_but_understood_turn");
}