ARCH: заземлить metadata surface в следующий MCP lane

This commit is contained in:
2026-04-21 22:33:46 +03:00
parent d66e2bfb01
commit 6d9c1568c3
9 changed files with 491 additions and 9 deletions
@@ -122,6 +122,12 @@ export interface AssistantMcpDiscoveryDerivedMetadataSurface {
matched_rows: number;
available_entity_sets: string[];
matched_objects: string[];
selected_entity_set: string | null;
selected_surface_objects: string[];
downstream_route_family: "document_evidence" | "movement_evidence" | "catalog_drilldown" | null;
recommended_next_primitive: "query_documents" | "query_movements" | "drilldown_related_objects" | null;
ambiguity_detected: boolean;
ambiguity_entity_sets: string[];
available_fields: string[];
known_limitations: string[];
inference_basis: "confirmed_1c_metadata_surface_rows";
@@ -741,6 +747,101 @@ function metadataAvailableFields(rows: Array<Record<string, unknown>>): string[]
return result;
}
function normalizeMetadataEntitySetToken(value: string): string {
return String(value ?? "")
.toLowerCase()
.replace(/[\s_.-]+/g, "");
}
function metadataMatchesRequestedType(entitySet: string, requestedMetaType: string): boolean {
const entityToken = normalizeMetadataEntitySetToken(entitySet);
const requestedToken = normalizeMetadataEntitySetToken(requestedMetaType);
return entityToken.includes(requestedToken) || requestedToken.includes(entityToken);
}
function metadataRouteFamilyForEntitySet(
entitySet: string
): "document_evidence" | "movement_evidence" | "catalog_drilldown" | null {
const token = normalizeMetadataEntitySetToken(entitySet);
if (token.includes("документ") || token.includes("document")) {
return "document_evidence";
}
if (
token.includes("регистрнакопления") ||
token.includes("регистсведений") ||
token.includes("регистрсведений") ||
token.includes("accumulationregister") ||
token.includes("informationregister")
) {
return "movement_evidence";
}
if (token.includes("справочник") || token.includes("catalog") || token.includes("directory")) {
return "catalog_drilldown";
}
return null;
}
function metadataNextPrimitiveForRouteFamily(
routeFamily: "document_evidence" | "movement_evidence" | "catalog_drilldown" | null
): "query_documents" | "query_movements" | "drilldown_related_objects" | null {
if (routeFamily === "document_evidence") {
return "query_documents";
}
if (routeFamily === "movement_evidence") {
return "query_movements";
}
if (routeFamily === "catalog_drilldown") {
return "drilldown_related_objects";
}
return null;
}
function selectMetadataEntityGrounding(
availableEntitySets: string[],
requestedMetaTypes: string[]
): {
selectedEntitySet: string | null;
ambiguityDetected: boolean;
ambiguityEntitySets: string[];
} {
const requestedMatches = availableEntitySets.filter((entitySet) =>
requestedMetaTypes.some((requestedMetaType) => metadataMatchesRequestedType(entitySet, requestedMetaType))
);
if (requestedMatches.length === 1) {
return {
selectedEntitySet: requestedMatches[0] ?? null,
ambiguityDetected: false,
ambiguityEntitySets: []
};
}
if (requestedMatches.length > 1) {
return {
selectedEntitySet: null,
ambiguityDetected: true,
ambiguityEntitySets: requestedMatches
};
}
if (availableEntitySets.length === 1) {
return {
selectedEntitySet: availableEntitySets[0] ?? null,
ambiguityDetected: false,
ambiguityEntitySets: []
};
}
return {
selectedEntitySet: null,
ambiguityDetected: availableEntitySets.length > 1,
ambiguityEntitySets: availableEntitySets
};
}
function metadataObjectsForEntitySet(entitySet: string | null, matchedObjects: string[]): string[] {
if (!entitySet) {
return [];
}
return matchedObjects.filter((item) => item.startsWith(`${entitySet}.`) || item.includes(entitySet));
}
function deriveMetadataSurface(
result: AddressMcpMetadataRowsResult | null,
metadataScope: string | null,
@@ -761,14 +862,30 @@ function deriveMetadataSurface(
pushUnique(availableEntitySets, entitySet);
}
}
const grounding = selectMetadataEntityGrounding(availableEntitySets, requestedMetaTypes);
const downstreamRouteFamily = grounding.selectedEntitySet
? metadataRouteFamilyForEntitySet(grounding.selectedEntitySet)
: null;
const knownLimitations: string[] = [];
if (grounding.ambiguityDetected && grounding.ambiguityEntitySets.length > 0) {
knownLimitations.push(
`Exact downstream metadata surface remains ambiguous across: ${grounding.ambiguityEntitySets.join(", ")}`
);
}
return {
metadata_scope: metadataScope,
requested_meta_types: requestedMetaTypes,
matched_rows: result.rows.length,
available_entity_sets: availableEntitySets,
matched_objects: matchedObjects,
selected_entity_set: grounding.selectedEntitySet,
selected_surface_objects: metadataObjectsForEntitySet(grounding.selectedEntitySet, matchedObjects),
downstream_route_family: downstreamRouteFamily,
recommended_next_primitive: metadataNextPrimitiveForRouteFamily(downstreamRouteFamily),
ambiguity_detected: grounding.ambiguityDetected,
ambiguity_entity_sets: grounding.ambiguityEntitySets,
available_fields: metadataAvailableFields(result.rows),
known_limitations: [],
known_limitations: knownLimitations,
inference_basis: "confirmed_1c_metadata_surface_rows"
};
}
@@ -787,17 +904,37 @@ function buildMetadataConfirmedFacts(
if (surface.available_entity_sets.length > 0) {
facts.push(`Available metadata object sets: ${surface.available_entity_sets.join(", ")}`);
}
if (surface.selected_entity_set) {
facts.push(`Selected metadata entity set: ${surface.selected_entity_set}`);
}
if (surface.selected_surface_objects.length > 0) {
facts.push(`Selected metadata objects: ${surface.selected_surface_objects.slice(0, 8).join(", ")}`);
}
if (surface.available_fields.length > 0) {
facts.push(`Available metadata fields/sections: ${surface.available_fields.slice(0, 12).join(", ")}`);
}
return facts;
}
function buildMetadataInferredFacts(
surface: AssistantMcpDiscoveryDerivedMetadataSurface | null
): string[] {
if (!surface || !surface.selected_entity_set || !surface.downstream_route_family || !surface.recommended_next_primitive) {
return [];
}
return [
`A likely next checked lane may be inferred as ${surface.downstream_route_family} from the confirmed metadata surface`
];
}
function buildMetadataUnknownFacts(
surface: AssistantMcpDiscoveryDerivedMetadataSurface | null,
metadataScope: string | null
): string[] {
if (surface) {
if (surface.ambiguity_detected && surface.ambiguity_entity_sets.length > 0) {
return [...surface.known_limitations];
}
if (surface.available_fields.length > 0) {
return [];
}
@@ -1466,6 +1603,7 @@ export async function executeAssistantMcpDiscoveryPilot(
plan: planner.discovery_plan,
probeResults,
confirmedFacts: buildMetadataConfirmedFacts(derivedMetadataSurface),
inferredFacts: buildMetadataInferredFacts(derivedMetadataSurface),
unknownFacts: buildMetadataUnknownFacts(derivedMetadataSurface, metadataScope),
sourceRowsSummary,
queryLimitations,