Open-World: удержать бизнес-обзор от мусорных anchors

This commit is contained in:
2026-05-05 15:43:47 +03:00
parent ee69c7bc8d
commit d797b3fa99
13 changed files with 424 additions and 19 deletions
@@ -56,7 +56,17 @@ function hasUnicodeLikelyCounterpartyAfterBy(text) {
"\u0431\u0430\u043d\u043a\u0435",
"\u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442\u0430\u043c",
"\u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442\u044b",
"\u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442\u0443"
"\u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442\u0443",
"\u0434\u0430\u043d\u043d\u044b\u043c",
"\u044d\u0442\u0438\u043c",
"\u044d\u0442\u0438\u043c\u0438",
"\u0438\u0442\u043e\u0433\u0443",
"\u0438\u0442\u043e\u0433\u0430\u043c",
"\u0432\u0441\u0435\u043c\u0443",
"\u0432\u0441\u0435\u0439",
"\u0432\u0441\u0435\u043c",
"\u0432\u044b\u0432\u043e\u0434\u0443",
"\u0432\u044b\u0432\u043e\u0434\u0430\u043c"
]);
return !stopWords.has(token);
}
@@ -1152,6 +1152,30 @@ function isLikelyCounterpartyToken(rawToken) {
"этомуже",
"томуже"
]);
for (const semanticStopWord of [
"данным",
"этим",
"этими",
"итогу",
"итогам",
"всему",
"всей",
"всем",
"выводу",
"выводам",
"аудиту",
"прокси",
"покажи",
"показать",
"выведи",
"вывести",
"дай",
"дать",
"какие",
"список"
]) {
stopWords.add(semanticStopWord);
}
return !stopWords.has(token);
}
function hasPartyAnchorMention(text) {
@@ -1658,6 +1682,18 @@ function resolveUnicodeAddressIntentBridge(text) {
"сальдо",
"дате",
"периоду",
"данным",
"этим",
"этими",
"итогу",
"итогам",
"всему",
"всей",
"всем",
"выводу",
"выводам",
"аудиту",
"прокси",
"складу",
"товару",
"этому",
@@ -95,11 +95,48 @@ function isActionVerbEntityCandidate(value) {
}
return new Set(["провести", "показать", "посмотреть", "смотреть", "найти", "искать", "анализ"]).has((0, addressTextRepair_1.normalizeRussianComparableText)(value));
}
function isGarbageSemanticAnchorCandidate(value) {
if (!value) {
return false;
}
const text = compactLower((0, addressTextRepair_1.repairAddressMojibakeText)(value));
if (!text || text.length < 2) {
return true;
}
const compact = (0, addressTextRepair_1.normalizeRussianComparableText)(text);
if (new Set([
"данным",
"этим",
"этими",
"итогу",
"итогам",
"всему",
"всей",
"всем",
"выводу",
"выводам",
"аудиту",
"прокси",
"proxy",
"summary",
"overall"
]).has(compact)) {
return true;
}
if (/^(?:по\s+)?(?:этим|этими)\s+данн\p{L}*$/iu.test(text) ||
/^(?:и\s+)?кто\s+(?:главн\p{L}*|основн\p{L}*|крупн\p{L}*)\s+(?:клиент|покупател|поставщик|контрагент)(?:\s+в)?$/iu.test(text) ||
/^(?:что|чего)\s+(?:подтвержден\p{L}*|не\s+хватает)/iu.test(text) ||
/^(?:можно\s+ли|если\s+нет|дай\s+proxy|дай\s+прокси)/iu.test(text)) {
return true;
}
return false;
}
function isInvalidEntityCandidate(value) {
return Boolean(value &&
(isReferentialEntityPlaceholder(value) ||
isValueFlowPredicateEntityCandidate(value) ||
isActionVerbEntityCandidate(value)));
isActionVerbEntityCandidate(value) ||
isGarbageSemanticAnchorCandidate(value)));
}
function normalizeFollowupCounterpartyCandidate(value) {
const text = candidateValue(value);
@@ -288,6 +325,13 @@ function mapPilotScopeToFollowupMeaning(pilotScope) {
unsupported: "1c_metadata_surface"
};
}
if (pilotScope === "business_overview_route_template_v1") {
return {
domain: "business_overview",
action: "broad_evaluation",
unsupported: "broad_business_evaluation"
};
}
return {
domain: null,
action: null,
@@ -568,6 +612,23 @@ function hasBusinessOverviewSignal(text) {
}
return /(?:\u0431\u0438\u0437\u043d\u0435\u0441[-\s]?\u043e\u0431\u0437\u043e\u0440|\u0431\u0438\u0437\u043d\u0435\u0441[-\s]?\u0430\u0443\u0434\u0438\u0442|\u043f\u043e\u043b\u043d\w*\s+\u0430\u043d\u0430\u043b\u0438\u0437\s+(?:\u043a\u043e\u043c\u043f\u0430\u043d|\u0431\u0438\u0437\u043d\u0435\u0441|\u0434\u0435\u044f\u0442\u0435\u043b)|\u0441\u0432\u043e\u0434\u043d\w*\s+\u0430\u043d\u0430\u043b\u0438\u0437\s+(?:\u043a\u043e\u043c\u043f\u0430\u043d|\u0431\u0438\u0437\u043d\u0435\u0441|\u0434\u0435\u044f\u0442\u0435\u043b)|\u043a\u0430\u043a\s+\u0442\u044b\s+\u043e\u0446\u0435\u043d(?:\u0438\u0448\u044c|\u0438)\s+\u0434\u0435\u044f\u0442\u0435\u043b\u044c\u043d\u043e\u0441\u0442|\u043a\u043e\u043c\u043f\u0430\u043d(?:\u0438\u0438|\u0438\u044e|\u0438\u044f)\s+\u0432\s+\u0446\u0435\u043b\u043e\u043c|company\s+(?:analysis|overview)|business\s+(?:overview|audit)|llm[-\s]?audit|бизнес[-\s]?обзор|бизнес[-\s]?аудит)/iu.test(text);
}
function hasBusinessOverviewContinuationSignal(text) {
if (!text) {
return false;
}
const normalized = compactLower(text);
const hasEvidenceContinuationCue = /(?:по\s+этим\s+данн|на\s+базе\s+этих\s+данн|из\s+этого|по\s+итогам|что\s+подтвержден|чего\s+не\s+хватает|что\s+пока\s+нельзя|что\s+можно\s+сказать|какой\s+вывод|итогов|резюм|вывод)/iu.test(normalized);
const hasAnalystContinuationCue = /(?:можно\s+ли|если\s+нет|proxy|прокси|аудит|оцен|что\s+думаешь|нормальн\p{L}*\s+прибыл|прибыл|марж|рентаб|ликвидн|просроч|качество\s+долг|риск|налогов\p{L}*\s+вывод)/iu.test(normalized);
const hasTaxContinuationCue = /(?:ндс|vat)[\s\S]{0,120}(?:позици|период|основан|не\s+хватает|налогов\p{L}*\s+вывод)|(?:позици|налогов\p{L}*\s+вывод)[\s\S]{0,80}(?:ндс|vat)/iu.test(normalized);
return hasEvidenceContinuationCue || hasAnalystContinuationCue || hasTaxContinuationCue;
}
function hasBusinessOverviewFollowupSeed(followupSeed) {
return (followupSeed.pilotScope === "business_overview_route_template_v1" ||
followupSeed.domain === "business_overview" ||
followupSeed.action === "broad_evaluation" ||
followupSeed.unsupported === "broad_business_evaluation" ||
followupSeed.loopSelectedChainId === "business_overview");
}
function hasValueFlowSignal(text) {
return /(?:оборот|выручк|оплат|плат[её]ж|заплат|перечисл|списан|расход|исходящ|входящ|получ(?:ил|ено|ен)|поступил|поступлен|денежн[а-яёa-z0-9_-]*\s+поток|(?<!\p{L})заработ(?:ал|али|ало|аем|ает|ать|ано|ок)(?!\p{L})|supplier|value[-\s]?flow|turnover|revenue|payment|payout|outflow|cash\s+flow|\bearn(?:ed|ing|ings)?\b)/iu.test(text);
}
@@ -912,7 +973,8 @@ function buildAssistantMcpDiscoveryTurnInput(input) {
const rawEntitySourceText = repairedUserText ?? rawUserText ?? repairedEffectiveText ?? rawEffectiveText ?? rawSignalSourceText;
const rawText = compactLower(rawSignalSourceText);
const rawReferentialDocumentExclusionSignal = hasReferentialDocumentExclusionFollowupSignal(repairedUserText ?? rawUserText ?? "");
const rawBusinessOverviewSignal = hasBusinessOverviewSignal(rawText);
const businessOverviewContinuationSignal = hasBusinessOverviewFollowupSeed(followupSeed) && hasBusinessOverviewContinuationSignal(rawText);
const rawBusinessOverviewSignal = hasBusinessOverviewSignal(rawText) || businessOverviewContinuationSignal;
const rawLifecycleSignal = !rawBusinessOverviewSignal && hasLifecycleSignal(rawText);
const rawBidirectionalValueFlowSignal = !rawBusinessOverviewSignal && !rawLifecycleSignal && hasBidirectionalValueFlowSignal(rawText);
const rawValueFlowSignal = !rawBusinessOverviewSignal &&
@@ -1744,6 +1806,9 @@ function buildAssistantMcpDiscoveryTurnInput(input) {
if (businessOverviewSignal) {
pushReason(reasonCodes, "mcp_discovery_broad_business_evaluation_route_candidate");
}
if (businessOverviewContinuationSignal) {
pushReason(reasonCodes, "mcp_discovery_business_overview_continuation_from_followup_context");
}
if (!(valueFlowOrganizationStaysScope && normalizedPredecomposeCounterparty === explicitOrganizationScope) &&
normalizedPredecomposeCounterparty) {
pushReason(reasonCodes, "mcp_discovery_counterparty_from_predecompose");