Починить восстановление кириллицы в автопрогонах

2026-05-10 08:38:52 +03:00 · 2026-05-10 08:38:52 +03:00 · 3be06b5f93
parent 9b02083493
commit 3be06b5f93
18 changed files with 618 additions and 74 deletions
--- a/llm_normalizer/backend/dist/routes/autoRuns.js
+++ b/llm_normalizer/backend/dist/routes/autoRuns.js
@ -102,7 +102,7 @@ function parseAutoGenTitle(value) {
    if (!title) {
        return null;
    }
-    return title.slice(0, 160);
+    return repairAutogenMojibake(title).slice(0, 160);
 }
 function parseManualCaseDecision(value, fallback = "needs_dialog_policy_fix") {
    const normalized = toStringSafe(value);
@ -1186,7 +1186,9 @@ function textMojibakeScore(value) {
    const hardMarkers = (source.match(/[Ѓѓ‚„…†‡€‰‹ЉЊЌЋЏ‘’“”•–—™љ›њќћџ]/g) ?? []).length;
    const pairMarkers = (source.match(/(?:Р.|С.|Ð.|Ñ.)/g) ?? []).length;
    const doubleEncodedMarkers = (source.match(/(?:Г[Ђ-џ]|В[Ђ-џ]|Ã.|Â.)/gu) ?? []).length;
-    return cyrillic + latin - hardMarkers * 3 - pairMarkers * 2 - doubleEncodedMarkers * 2;
+    const replacement = (source.match(/\uFFFD/g) ?? []).length;
+    const c1Controls = (source.match(/[\u0080-\u009f]/g) ?? []).length;
+    return cyrillic + latin - replacement * 8 - c1Controls * 5 - hardMarkers * 3 - pairMarkers * 2 - doubleEncodedMarkers * 2;
 }
 function looksLikeMojibake(value) {
    const source = String(value ?? "");
@ -1201,16 +1203,36 @@ function looksLikeMojibake(value) {
    }
    return (source.match(/(?:Г[Ђ-џ]|В[Ђ-џ]|Ã.|Â.)/gu) ?? []).length >= 2;
 }
+function encodeWin1251MojibakeBytes(value) {
+    const chunks = [];
+    for (const char of String(value ?? "")) {
+        const code = char.codePointAt(0) ?? 0;
+        if (code >= 0x80 && code <= 0x9f) {
+            chunks.push(Buffer.from([code]));
+            continue;
+        }
+        chunks.push(iconv_lite_1.default.encode(char, "win1251"));
+    }
+    return Buffer.concat(chunks);
+}
+function decodeUtf8FromWin1251Mojibake(value) {
+    return encodeWin1251MojibakeBytes(value).toString("utf8");
+}
+function repairKnownReplacementDamagedAutogenText(value) {
+    return String(value ?? "")
+        .replace(/\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422[\uFFFD?]+\u0412\u0410/giu, "\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422\u0418\u0412\u0410")
+        .replace(/\u041e\u0411\u0429[\uFFFD?]+\u0419/giu, "\u041e\u0411\u0429\u0418\u0419");
+}
 function repairAutogenMojibake(value) {
-    const source = String(value ?? "");
-    if (!looksLikeMojibake(source)) {
+    const source = repairKnownReplacementDamagedAutogenText(String(value ?? ""));
+    if (!looksLikeMojibake(source) && !/[\u0080-\u009f\uFFFD]/.test(source)) {
        return source;
    }
    let candidate = source;
    for (let pass = 0; pass < 3; pass += 1) {
        let improved = false;
        try {
-            const fromWin1251 = iconv_lite_1.default.encode(candidate, "win1251").toString("utf8");
+            const fromWin1251 = decodeUtf8FromWin1251Mojibake(candidate);
            if (textMojibakeScore(fromWin1251) > textMojibakeScore(candidate)) {
                candidate = fromWin1251;
                improved = true;
@ -1229,11 +1251,16 @@ function repairAutogenMojibake(value) {
        catch {
            // ignore
        }
+        const repairedKnownText = repairKnownReplacementDamagedAutogenText(candidate);
+        if (repairedKnownText !== candidate) {
+            candidate = repairedKnownText;
+            improved = true;
+        }
        if (!improved) {
            break;
        }
    }
-    return candidate;
+    return repairKnownReplacementDamagedAutogenText(candidate);
 }
 function sanitizeGeneratedQuestion(value) {
    return repairAutogenMojibake(String(value ?? ""))
@ -1432,7 +1459,8 @@ function extractQuestionsFromAutogenOutput(rawText) {
 }
 exports.__autoRunsQuestionTestUtils = {
    splitQuestionCandidates,
-    extractQuestionsFromAutogenOutput
+    extractQuestionsFromAutogenOutput,
+    repairAutogenMojibake
 };
 async function generateQwenSeedQuestionsLive(input) {
    const seedExamples = collectCanonicalQuestions(40);
--- a/llm_normalizer/backend/dist/routes/eval.js
+++ b/llm_normalizer/backend/dist/routes/eval.js
@ -11,6 +11,7 @@ const nanoid_1 = require("nanoid");
 const express_1 = require("express");
 const config_1 = require("../config");
 const http_1 = require("../utils/http");
+const addressTextRepair_1 = require("../services/addressTextRepair");
 const ASYNC_JOBS = new Map();
 const MAX_ASYNC_JOBS = 80;
 function toRecord(value) {
@ -30,7 +31,7 @@ function toArray(value) {
    return Array.isArray(value) ? value : [];
 }
 function normalizeQuestionChunk(value) {
-    return String(value ?? "")
+    return (0, addressTextRepair_1.repairAddressMojibakeText)(String(value ?? ""))
        .replace(/\r/g, " ")
        .replace(/\t/g, " ")
        .replace(/\s+/g, " ")
@ -92,7 +93,7 @@ function normalizeRuntimeQuestionList(items) {
    return normalized.filter((item) => item.length > 0);
 }
 function splitQuestionCandidate(raw) {
-    const normalized = String(raw ?? "").replace(/\r/g, "\n").trim();
+    const normalized = (0, addressTextRepair_1.repairAddressMojibakeText)(String(raw ?? "")).replace(/\r/g, "\n").trim();
    if (!normalized) {
        return [];
    }
@ -455,7 +456,8 @@ function buildEvalRouter(services) {
            }
            const questions = normalizeRuntimeQuestions(body.questions);
            const scenarioQuestions = normalizeRuntimeQuestions(body.scenarioQuestions, { dedupe: false, splitCandidates: false });
-            const scenarioTitle = toStringSafe(body.scenarioTitle);
+            const scenarioTitleRaw = toStringSafe(body.scenarioTitle);
+            const scenarioTitle = scenarioTitleRaw ? (0, addressTextRepair_1.repairAddressMojibakeText)(scenarioTitleRaw) : null;
            const jobId = `job-${(0, nanoid_1.nanoid)(10)}`;
            const runId = `assistant-stage1-${(0, nanoid_1.nanoid)(10)}`;
            const runtimeCaseSetFile = scenarioQuestions.length > 0
--- a/llm_normalizer/backend/dist/services/addressTextRepair.js
+++ b/llm_normalizer/backend/dist/services/addressTextRepair.js
@ -6,44 +6,68 @@ Object.defineProperty(exports, "__esModule", { value: true });
 exports.repairAddressMojibakeText = repairAddressMojibakeText;
 exports.normalizeRussianComparableText = normalizeRussianComparableText;
 const iconv_lite_1 = __importDefault(require("iconv-lite"));
+const MOJIBAKE_CONTINUATION_CLASS = "[\\u0080-\\u00bf\\u0401-\\u040f\\u0451-\\u045f\\u2018-\\u201e\\u2020-\\u2022\\u2013-\\u2014\\u2122\\u20ac]";
+const MOJIBAKE_PAIR_PATTERN = new RegExp(`(?:[\\u0420\\u0421]${MOJIBAKE_CONTINUATION_CLASS})`, "gu");
 function compactWhitespace(value) {
    return value.replace(/\s+/g, " ").trim();
 }
+function countMatches(value, pattern) {
+    return (String(value ?? "").match(pattern) ?? []).length;
+}
 function textMojibakeScore(value) {
    const source = String(value ?? "");
-    const cyrillic = (source.match(/[\u0400-\u04ff]/g) ?? []).length;
-    const latin = (source.match(/[A-Za-z]/g) ?? []).length;
-    const replacement = (source.match(/[<5B>]/g) ?? []).length;
-    const pairMarkers = (source.match(/(?:Р.|С.|Ð.|Ñ.)/g) ?? []).length;
-    const doubleEncodedMarkers = (source.match(/(?:Р“[Р-џ]|Р’[Р-џ]|Ã.|Â.)/gu) ?? []).length;
-    return cyrillic + latin - replacement * 3 - pairMarkers * 2 - doubleEncodedMarkers * 2;
+    const cyrillic = countMatches(source, /[\u0400-\u04ff]/g);
+    const latin = countMatches(source, /[A-Za-z]/g);
+    const replacement = countMatches(source, /\uFFFD/g);
+    const c1Controls = countMatches(source, /[\u0080-\u009f]/g);
+    const pairMarkers = countMatches(source, MOJIBAKE_PAIR_PATTERN);
+    const doubleEncodedMarkers = countMatches(source, /(?:\u0420[\u00a0-\u00bf]\u0421|\u0413[\u0080-\u00bf]|\u00c3.|\u00c2.)/gu);
+    return cyrillic + latin - replacement * 8 - c1Controls * 5 - pairMarkers * 3 - doubleEncodedMarkers * 2;
 }
 function looksLikeAddressMojibake(value) {
    const source = String(value ?? "");
    if (!source.trim()) {
        return false;
    }
-    if (/[<EFBFBD>]/.test(source)) {
+    if (/[\u0080-\u009f\uFFFD]/.test(source)) {
        return true;
    }
-    if ((source.match(/(?:Р.|С.|Ð.|Ñ.)/g) ?? []).length >= 2) {
+    if (countMatches(source, MOJIBAKE_PAIR_PATTERN) >= 2) {
        return true;
    }
-    if ((source.match(/(?:Р“[Р-џ]|Р’[Р-џ]|Ã.|Â.)/gu) ?? []).length >= 2) {
-        return true;
+    return countMatches(source, /(?:\u0420[\u00a0-\u00bf]\u0421|\u0413[\u0080-\u00bf]|\u00c3.|\u00c2.)/gu) >= 2;
+}
+function encodeWin1251MojibakeBytes(value) {
+    const chunks = [];
+    for (const char of String(value ?? "")) {
+        const code = char.codePointAt(0) ?? 0;
+        if (code >= 0x80 && code <= 0x9f) {
+            chunks.push(Buffer.from([code]));
+            continue;
+        }
+        chunks.push(iconv_lite_1.default.encode(char, "win1251"));
    }
-    return false;
+    return Buffer.concat(chunks);
+}
+function decodeUtf8FromWin1251Mojibake(value) {
+    return encodeWin1251MojibakeBytes(value).toString("utf8");
+}
+function repairKnownReplacementDamagedRussianText(value) {
+    return String(value ?? "")
+        .replace(/\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422[\uFFFD?]+\u0412\u0410/giu, "\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422\u0418\u0412\u0410")
+        .replace(/\u041e\u0411\u0429[\uFFFD?]+\u0419/giu, "\u041e\u0411\u0429\u0418\u0419");
 }
 function repairAddressMojibakeText(value) {
    const source = String(value ?? "");
-    if (!looksLikeAddressMojibake(source)) {
-        return source;
+    const sourceWithKnownRepairs = repairKnownReplacementDamagedRussianText(source);
+    if (!looksLikeAddressMojibake(sourceWithKnownRepairs)) {
+        return sourceWithKnownRepairs;
    }
-    let candidate = source;
+    let candidate = sourceWithKnownRepairs;
    for (let pass = 0; pass < 3; pass += 1) {
        let improved = false;
        try {
-            const fromWin1251 = iconv_lite_1.default.encode(candidate, "win1251").toString("utf8");
+            const fromWin1251 = decodeUtf8FromWin1251Mojibake(candidate);
            if (textMojibakeScore(fromWin1251) > textMojibakeScore(candidate)) {
                candidate = fromWin1251;
                improved = true;
@ -62,12 +86,17 @@ function repairAddressMojibakeText(value) {
        catch {
            // Ignore decode failures and keep the current candidate.
        }
+        const repairedKnownText = repairKnownReplacementDamagedRussianText(candidate);
+        if (repairedKnownText !== candidate) {
+            candidate = repairedKnownText;
+            improved = true;
+        }
        if (!improved) {
            break;
        }
    }
-    return candidate;
+    return repairKnownReplacementDamagedRussianText(candidate);
 }
 function normalizeRussianComparableText(value) {
-    return compactWhitespace(repairAddressMojibakeText(String(value ?? "")).toLowerCase()).replace(/ё/g, "е");
+    return compactWhitespace(repairAddressMojibakeText(String(value ?? "")).toLowerCase()).replace(/\u0451/g, "\u0435");
 }
--- a/llm_normalizer/backend/dist/services/assistantLivingModePolicy.js
+++ b/llm_normalizer/backend/dist/services/assistantLivingModePolicy.js
@ -209,7 +209,10 @@ function createAssistantLivingModePolicy(deps) {
        if (hasAffectiveReactionCue) {
            return false;
        }
-        return normalized.length <= 36 && !/[?]/.test(String(userMessage ?? ""));
+        const rawQuestionProbe = String(userMessage ?? "")
+            .replace(/\uFFFD\?/g, "\uFFFD")
+            .replace(/пїЅ\?/giu, "пїЅ");
+        return normalized.length <= 36 && !/[?]/.test(rawQuestionProbe);
    }
    function hasAssistantDataScopeMetaQuestionSignal(text) {
        const repaired = repairAddressMojibake(String(text ?? ""));
--- a/llm_normalizer/backend/dist/services/assistantOrganizationMatcher.js
+++ b/llm_normalizer/backend/dist/services/assistantOrganizationMatcher.js
@ -182,18 +182,25 @@ function scoreOrganizationMentionInMessage(message, organization) {
                variantScore = Math.max(variantScore, variant.length * 5);
                continue;
            }
+            let fuzzyTokenScore = 0;
            const fuzzyMatched = messageTokens.some((messageToken) => {
                if (messageToken === variant) {
+                    fuzzyTokenScore = Math.max(fuzzyTokenScore, variant.length * 5);
                    return true;
                }
                if (messageToken.length >= 5 && variant.length >= 5) {
-                    return messageToken.startsWith(variant) || variant.startsWith(messageToken);
+                    const prefixMatched = messageToken.startsWith(variant) || variant.startsWith(messageToken);
+                    if (prefixMatched) {
+                        const prefixLength = Math.min(messageToken.length, variant.length);
+                        fuzzyTokenScore = Math.max(fuzzyTokenScore, variant.length * (prefixLength >= 7 ? 5 : 3));
+                    }
+                    return prefixMatched;
                }
                return false;
            });
            if (fuzzyMatched) {
                matched = true;
-                variantScore = Math.max(variantScore, Math.max(20, variant.length * 3));
+                variantScore = Math.max(variantScore, Math.max(20, fuzzyTokenScore || variant.length * 3));
            }
        }
        if (matched) {
--- a/llm_normalizer/backend/dist/services/assistantService.js
+++ b/llm_normalizer/backend/dist/services/assistantService.js
@ -1944,7 +1944,9 @@ function textMojibakeScoreForAddress(value) {
    const hardMarkers = (source.match(/[Ѓѓ‚„…†‡€‰‹ЉЊЌЋЏ\uFFFD?’“”•–—™љ›њќћџ]/g) ?? []).length;
    const pairMarkers = (source.match(/(?:Р.|С.|Ð.|Ñ.)/g) ?? []).length;
    const doubleEncodedMarkers = (source.match(/(?:Г[Ђ-џ]|В[Ђ-џ]|Ã.|Â.)/gu) ?? []).length;
-    return cyrillic + latin - hardMarkers * 3 - pairMarkers * 2 - doubleEncodedMarkers * 2;
+    const replacement = (source.match(/\uFFFD/g) ?? []).length;
+    const c1Controls = (source.match(/[\u0080-\u009f]/g) ?? []).length;
+    return cyrillic + latin - replacement * 8 - c1Controls * 5 - hardMarkers * 3 - pairMarkers * 2 - doubleEncodedMarkers * 2;
 }
 function looksLikeMojibakeForAddress(value) {
    const source = String(value ?? "");
@ -1962,16 +1964,36 @@ function looksLikeMojibakeForAddress(value) {
    }
    return false;
 }
+function encodeWin1251MojibakeBytesForAddress(value) {
+    const chunks = [];
+    for (const char of String(value ?? "")) {
+        const code = char.codePointAt(0) ?? 0;
+        if (code >= 0x80 && code <= 0x9f) {
+            chunks.push(Buffer.from([code]));
+            continue;
+        }
+        chunks.push(iconv_lite_1.default.encode(char, "win1251"));
+    }
+    return Buffer.concat(chunks);
+}
+function decodeUtf8FromWin1251MojibakeForAddress(value) {
+    return encodeWin1251MojibakeBytesForAddress(value).toString("utf8");
+}
+function repairKnownReplacementDamagedAddressText(value) {
+    return String(value ?? "")
+        .replace(/\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422[\uFFFD?]+\u0412\u0410/giu, "\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422\u0418\u0412\u0410")
+        .replace(/\u041e\u0411\u0429[\uFFFD?]+\u0419/giu, "\u041e\u0411\u0429\u0418\u0419");
+}
 function repairAddressMojibake(value) {
-    const source = String(value ?? "");
-    if (!looksLikeMojibakeForAddress(source)) {
+    const source = repairKnownReplacementDamagedAddressText(String(value ?? ""));
+    if (!looksLikeMojibakeForAddress(source) && !/[\u0080-\u009f\uFFFD]/.test(source)) {
        return source;
    }
    let candidate = source;
    for (let pass = 0; pass < 3; pass += 1) {
        let improved = false;
        try {
-            const fromWin1251 = iconv_lite_1.default.encode(candidate, "win1251").toString("utf8");
+            const fromWin1251 = decodeUtf8FromWin1251MojibakeForAddress(candidate);
            if (textMojibakeScoreForAddress(fromWin1251) > textMojibakeScoreForAddress(candidate)) {
                candidate = fromWin1251;
                improved = true;
@ -1986,11 +2008,16 @@ function repairAddressMojibake(value) {
            }
        }
        catch (_error) { }
+        const repairedKnownText = repairKnownReplacementDamagedAddressText(candidate);
+        if (repairedKnownText !== candidate) {
+            candidate = repairedKnownText;
+            improved = true;
+        }
        if (!improved) {
            break;
        }
    }
-    return candidate;
+    return repairKnownReplacementDamagedAddressText(candidate);
 }
 function sanitizeOutgoingAssistantText(value, fallback = "Не смог сформировать читаемый ответ. Уточните запрос.") {
    const repaired = repairAddressMojibake(String(value ?? ""));
@ -4311,18 +4338,25 @@ function scoreOrganizationMentionInMessage(message, organization) {
                variantScore = Math.max(variantScore, variant.length * 5);
                continue;
            }
+            let fuzzyTokenScore = 0;
            const fuzzyMatched = messageTokens.some((messageToken) => {
                if (messageToken === variant) {
+                    fuzzyTokenScore = Math.max(fuzzyTokenScore, variant.length * 5);
                    return true;
                }
                if (messageToken.length >= 5 && variant.length >= 5) {
-                    return messageToken.startsWith(variant) || variant.startsWith(messageToken);
+                    const prefixMatched = messageToken.startsWith(variant) || variant.startsWith(messageToken);
+                    if (prefixMatched) {
+                        const prefixLength = Math.min(messageToken.length, variant.length);
+                        fuzzyTokenScore = Math.max(fuzzyTokenScore, variant.length * (prefixLength >= 7 ? 5 : 3));
+                    }
+                    return prefixMatched;
                }
                return false;
            });
            if (fuzzyMatched) {
                matched = true;
-                variantScore = Math.max(variantScore, Math.max(20, variant.length * 3));
+                variantScore = Math.max(variantScore, Math.max(20, fuzzyTokenScore || variant.length * 3));
            }
        }
        if (matched) {
--- a/llm_normalizer/backend/src/routes/autoRuns.ts
+++ b/llm_normalizer/backend/src/routes/autoRuns.ts
@ -321,7 +321,7 @@ function parseAutoGenTitle(value: unknown): string | null {
  if (!title) {
    return null;
  }
-  return title.slice(0, 160);
+  return repairAutogenMojibake(title).slice(0, 160);
 }

 function parseManualCaseDecision(value: unknown, fallback: ManualCaseDecision = "needs_dialog_policy_fix"): ManualCaseDecision {
@ -1504,7 +1504,9 @@ function textMojibakeScore(value: string): number {
  const hardMarkers = (source.match(/[Ѓѓ‚„…†‡€‰‹ЉЊЌЋЏ‘’“”•–—™љ›њќћџ]/g) ?? []).length;
  const pairMarkers = (source.match(/(?:Р.|С.|Ð.|Ñ.)/g) ?? []).length;
  const doubleEncodedMarkers = (source.match(/(?:Г[Ђ-џ]|В[Ђ-џ]|Ã.|Â.)/gu) ?? []).length;
-  return cyrillic + latin - hardMarkers * 3 - pairMarkers * 2 - doubleEncodedMarkers * 2;
+  const replacement = (source.match(/\uFFFD/g) ?? []).length;
+  const c1Controls = (source.match(/[\u0080-\u009f]/g) ?? []).length;
+  return cyrillic + latin - replacement * 8 - c1Controls * 5 - hardMarkers * 3 - pairMarkers * 2 - doubleEncodedMarkers * 2;
 }

 function looksLikeMojibake(value: string): boolean {
@ -1521,16 +1523,39 @@ function looksLikeMojibake(value: string): boolean {
  return (source.match(/(?:Г[Ђ-џ]|В[Ђ-џ]|Ã.|Â.)/gu) ?? []).length >= 2;
 }

+function encodeWin1251MojibakeBytes(value: string): Buffer {
+  const chunks: Buffer[] = [];
+  for (const char of String(value ?? "")) {
+    const code = char.codePointAt(0) ?? 0;
+    if (code >= 0x80 && code <= 0x9f) {
+      chunks.push(Buffer.from([code]));
+      continue;
+    }
+    chunks.push(iconv.encode(char, "win1251"));
+  }
+  return Buffer.concat(chunks);
+}
+
+function decodeUtf8FromWin1251Mojibake(value: string): string {
+  return encodeWin1251MojibakeBytes(value).toString("utf8");
+}
+
+function repairKnownReplacementDamagedAutogenText(value: string): string {
+  return String(value ?? "")
+    .replace(/\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422[\uFFFD?]+\u0412\u0410/giu, "\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422\u0418\u0412\u0410")
+    .replace(/\u041e\u0411\u0429[\uFFFD?]+\u0419/giu, "\u041e\u0411\u0429\u0418\u0419");
+}
+
 function repairAutogenMojibake(value: string): string {
-  const source = String(value ?? "");
-  if (!looksLikeMojibake(source)) {
+  const source = repairKnownReplacementDamagedAutogenText(String(value ?? ""));
+  if (!looksLikeMojibake(source) && !/[\u0080-\u009f\uFFFD]/.test(source)) {
    return source;
  }
  let candidate = source;
  for (let pass = 0; pass < 3; pass += 1) {
    let improved = false;
    try {
-      const fromWin1251 = iconv.encode(candidate, "win1251").toString("utf8");
+      const fromWin1251 = decodeUtf8FromWin1251Mojibake(candidate);
      if (textMojibakeScore(fromWin1251) > textMojibakeScore(candidate)) {
        candidate = fromWin1251;
        improved = true;
@ -1547,11 +1572,16 @@ function repairAutogenMojibake(value: string): string {
    } catch {
      // ignore
    }
+    const repairedKnownText = repairKnownReplacementDamagedAutogenText(candidate);
+    if (repairedKnownText !== candidate) {
+      candidate = repairedKnownText;
+      improved = true;
+    }
    if (!improved) {
      break;
    }
  }
-  return candidate;
+  return repairKnownReplacementDamagedAutogenText(candidate);
 }

 function sanitizeGeneratedQuestion(value: string): string {
@ -1775,7 +1805,8 @@ function extractQuestionsFromAutogenOutput(rawText: string): string[] {

 export const __autoRunsQuestionTestUtils = {
  splitQuestionCandidates,
-  extractQuestionsFromAutogenOutput
+  extractQuestionsFromAutogenOutput,
+  repairAutogenMojibake
 };

 async function generateQwenSeedQuestionsLive(input: {
--- a/llm_normalizer/backend/src/routes/eval.ts
+++ b/llm_normalizer/backend/src/routes/eval.ts
@ -7,6 +7,7 @@ import type { AppServices } from "../serverContext";
 import { ApiError, ok } from "../utils/http";
 import type { EvalRunMode, NormalizeRequestPayload } from "../types/normalizer";
 import type { EvalTarget } from "../types/assistantEval";
+import { repairAddressMojibakeText } from "../services/addressTextRepair";

 type EvalAsyncStatus = "queued" | "running" | "completed" | "failed" | "canceled";

@ -67,7 +68,7 @@ function toArray(value: unknown): unknown[] {
 }

 function normalizeQuestionChunk(value: string): string {
-  return String(value ?? "")
+  return repairAddressMojibakeText(String(value ?? ""))
    .replace(/\r/g, " ")
    .replace(/\t/g, " ")
    .replace(/\s+/g, " ")
@ -136,7 +137,7 @@ function normalizeRuntimeQuestionList(items: string[]): string[] {
 }

 function splitQuestionCandidate(raw: string): string[] {
-  const normalized = String(raw ?? "").replace(/\r/g, "\n").trim();
+  const normalized = repairAddressMojibakeText(String(raw ?? "")).replace(/\r/g, "\n").trim();
  if (!normalized) {
    return [];
  }
@ -533,7 +534,8 @@ export function buildEvalRouter(services: AppServices): Router {
      }
      const questions = normalizeRuntimeQuestions(body.questions);
      const scenarioQuestions = normalizeRuntimeQuestions(body.scenarioQuestions, { dedupe: false, splitCandidates: false });
-      const scenarioTitle = toStringSafe(body.scenarioTitle);
+      const scenarioTitleRaw = toStringSafe(body.scenarioTitle);
+      const scenarioTitle = scenarioTitleRaw ? repairAddressMojibakeText(scenarioTitleRaw) : null;

      const jobId = `job-${nanoid(10)}`;
      const runId = `assistant-stage1-${nanoid(10)}`;
--- a/llm_normalizer/backend/src/services/addressTextRepair.ts
+++ b/llm_normalizer/backend/src/services/addressTextRepair.ts
@ -1,17 +1,26 @@
 import iconv from "iconv-lite";

+const MOJIBAKE_CONTINUATION_CLASS =
+  "[\\u0080-\\u00bf\\u0401-\\u040f\\u0451-\\u045f\\u2018-\\u201e\\u2020-\\u2022\\u2013-\\u2014\\u2122\\u20ac]";
+const MOJIBAKE_PAIR_PATTERN = new RegExp(`(?:[\\u0420\\u0421]${MOJIBAKE_CONTINUATION_CLASS})`, "gu");
+
 function compactWhitespace(value: string): string {
  return value.replace(/\s+/g, " ").trim();
 }

+function countMatches(value: string, pattern: RegExp): number {
+  return (String(value ?? "").match(pattern) ?? []).length;
+}
+
 function textMojibakeScore(value: string): number {
  const source = String(value ?? "");
-  const cyrillic = (source.match(/[\u0400-\u04ff]/g) ?? []).length;
-  const latin = (source.match(/[A-Za-z]/g) ?? []).length;
-  const replacement = (source.match(/[<5B>]/g) ?? []).length;
-  const pairMarkers = (source.match(/(?:Р.|С.|Ð.|Ñ.)/g) ?? []).length;
-  const doubleEncodedMarkers = (source.match(/(?:Р“[Р-џ]|Р’[Р-џ]|Ã.|Â.)/gu) ?? []).length;
-  return cyrillic + latin - replacement * 3 - pairMarkers * 2 - doubleEncodedMarkers * 2;
+  const cyrillic = countMatches(source, /[\u0400-\u04ff]/g);
+  const latin = countMatches(source, /[A-Za-z]/g);
+  const replacement = countMatches(source, /\uFFFD/g);
+  const c1Controls = countMatches(source, /[\u0080-\u009f]/g);
+  const pairMarkers = countMatches(source, MOJIBAKE_PAIR_PATTERN);
+  const doubleEncodedMarkers = countMatches(source, /(?:\u0420[\u00a0-\u00bf]\u0421|\u0413[\u0080-\u00bf]|\u00c3.|\u00c2.)/gu);
+  return cyrillic + latin - replacement * 8 - c1Controls * 5 - pairMarkers * 3 - doubleEncodedMarkers * 2;
 }

 function looksLikeAddressMojibake(value: string): boolean {
@ -19,30 +28,51 @@ function looksLikeAddressMojibake(value: string): boolean {
  if (!source.trim()) {
    return false;
  }
-  if (/[<EFBFBD>]/.test(source)) {
+  if (/[\u0080-\u009f\uFFFD]/.test(source)) {
    return true;
  }
-  if ((source.match(/(?:Р.|С.|Ð.|Ñ.)/g) ?? []).length >= 2) {
+  if (countMatches(source, MOJIBAKE_PAIR_PATTERN) >= 2) {
    return true;
  }
-  if ((source.match(/(?:Р“[Р-џ]|Р’[Р-џ]|Ã.|Â.)/gu) ?? []).length >= 2) {
-    return true;
+  return countMatches(source, /(?:\u0420[\u00a0-\u00bf]\u0421|\u0413[\u0080-\u00bf]|\u00c3.|\u00c2.)/gu) >= 2;
+}
+
+function encodeWin1251MojibakeBytes(value: string): Buffer {
+  const chunks: Buffer[] = [];
+  for (const char of String(value ?? "")) {
+    const code = char.codePointAt(0) ?? 0;
+    if (code >= 0x80 && code <= 0x9f) {
+      chunks.push(Buffer.from([code]));
+      continue;
+    }
+    chunks.push(iconv.encode(char, "win1251"));
  }
-  return false;
+  return Buffer.concat(chunks);
+}
+
+function decodeUtf8FromWin1251Mojibake(value: string): string {
+  return encodeWin1251MojibakeBytes(value).toString("utf8");
+}
+
+function repairKnownReplacementDamagedRussianText(value: string): string {
+  return String(value ?? "")
+    .replace(/\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422[\uFFFD?]+\u0412\u0410/giu, "\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422\u0418\u0412\u0410")
+    .replace(/\u041e\u0411\u0429[\uFFFD?]+\u0419/giu, "\u041e\u0411\u0429\u0418\u0419");
 }

 export function repairAddressMojibakeText(value: string): string {
  const source = String(value ?? "");
-  if (!looksLikeAddressMojibake(source)) {
-    return source;
+  const sourceWithKnownRepairs = repairKnownReplacementDamagedRussianText(source);
+  if (!looksLikeAddressMojibake(sourceWithKnownRepairs)) {
+    return sourceWithKnownRepairs;
  }

-  let candidate = source;
+  let candidate = sourceWithKnownRepairs;
  for (let pass = 0; pass < 3; pass += 1) {
    let improved = false;

    try {
-      const fromWin1251 = iconv.encode(candidate, "win1251").toString("utf8");
+      const fromWin1251 = decodeUtf8FromWin1251Mojibake(candidate);
      if (textMojibakeScore(fromWin1251) > textMojibakeScore(candidate)) {
        candidate = fromWin1251;
        improved = true;
@ -61,14 +91,20 @@ export function repairAddressMojibakeText(value: string): string {
      // Ignore decode failures and keep the current candidate.
    }

+    const repairedKnownText = repairKnownReplacementDamagedRussianText(candidate);
+    if (repairedKnownText !== candidate) {
+      candidate = repairedKnownText;
+      improved = true;
+    }
+
    if (!improved) {
      break;
    }
  }

-  return candidate;
+  return repairKnownReplacementDamagedRussianText(candidate);
 }

 export function normalizeRussianComparableText(value: unknown): string {
-  return compactWhitespace(repairAddressMojibakeText(String(value ?? "")).toLowerCase()).replace(/ё/g, "е");
+  return compactWhitespace(repairAddressMojibakeText(String(value ?? "")).toLowerCase()).replace(/\u0451/g, "\u0435");
 }
--- a/llm_normalizer/backend/src/services/assistantLivingModePolicy.ts
+++ b/llm_normalizer/backend/src/services/assistantLivingModePolicy.ts
@ -279,7 +279,10 @@ export function createAssistantLivingModePolicy(deps: AssistantLivingModePolicyD
      if (hasAffectiveReactionCue) {
          return false;
      }
-      return normalized.length <= 36 && !/[?]/.test(String(userMessage ?? ""));
+      const rawQuestionProbe = String(userMessage ?? "")
+          .replace(/\uFFFD\?/g, "\uFFFD")
+          .replace(/пїЅ\?/giu, "пїЅ");
+      return normalized.length <= 36 && !/[?]/.test(rawQuestionProbe);
  }

  function hasAssistantDataScopeMetaQuestionSignal(text) {
--- a/llm_normalizer/backend/src/services/assistantOrganizationMatcher.ts
+++ b/llm_normalizer/backend/src/services/assistantOrganizationMatcher.ts
@ -189,18 +189,25 @@ export function scoreOrganizationMentionInMessage(message: unknown, organization
        variantScore = Math.max(variantScore, variant.length * 5);
        continue;
      }
+      let fuzzyTokenScore = 0;
      const fuzzyMatched = messageTokens.some((messageToken) => {
        if (messageToken === variant) {
+          fuzzyTokenScore = Math.max(fuzzyTokenScore, variant.length * 5);
          return true;
        }
        if (messageToken.length >= 5 && variant.length >= 5) {
-          return messageToken.startsWith(variant) || variant.startsWith(messageToken);
+          const prefixMatched = messageToken.startsWith(variant) || variant.startsWith(messageToken);
+          if (prefixMatched) {
+            const prefixLength = Math.min(messageToken.length, variant.length);
+            fuzzyTokenScore = Math.max(fuzzyTokenScore, variant.length * (prefixLength >= 7 ? 5 : 3));
+          }
+          return prefixMatched;
        }
        return false;
      });
      if (fuzzyMatched) {
        matched = true;
-        variantScore = Math.max(variantScore, Math.max(20, variant.length * 3));
+        variantScore = Math.max(variantScore, Math.max(20, fuzzyTokenScore || variant.length * 3));
      }
    }
    if (matched) {
--- a/llm_normalizer/backend/src/services/assistantService.ts
+++ b/llm_normalizer/backend/src/services/assistantService.ts
@ -1900,7 +1900,9 @@ function textMojibakeScoreForAddress(value) {
    const hardMarkers = (source.match(/[Ѓѓ‚„…†‡€‰‹ЉЊЌЋЏ\uFFFD?’“”•–—™љ›њќћџ]/g) ?? []).length;
    const pairMarkers = (source.match(/(?:Р.|С.|Ð.|Ñ.)/g) ?? []).length;
    const doubleEncodedMarkers = (source.match(/(?:Г[Ђ-џ]|В[Ђ-џ]|Ã.|Â.)/gu) ?? []).length;
-    return cyrillic + latin - hardMarkers * 3 - pairMarkers * 2 - doubleEncodedMarkers * 2;
+    const replacement = (source.match(/\uFFFD/g) ?? []).length;
+    const c1Controls = (source.match(/[\u0080-\u009f]/g) ?? []).length;
+    return cyrillic + latin - replacement * 8 - c1Controls * 5 - hardMarkers * 3 - pairMarkers * 2 - doubleEncodedMarkers * 2;
 }
 function looksLikeMojibakeForAddress(value) {
    const source = String(value ?? "");
@ -1918,16 +1920,36 @@ function looksLikeMojibakeForAddress(value) {
    }
    return false;
 }
+function encodeWin1251MojibakeBytesForAddress(value) {
+    const chunks = [];
+    for (const char of String(value ?? "")) {
+        const code = char.codePointAt(0) ?? 0;
+        if (code >= 0x80 && code <= 0x9f) {
+            chunks.push(Buffer.from([code]));
+            continue;
+        }
+        chunks.push(iconv.encode(char, "win1251"));
+    }
+    return Buffer.concat(chunks);
+}
+function decodeUtf8FromWin1251MojibakeForAddress(value) {
+    return encodeWin1251MojibakeBytesForAddress(value).toString("utf8");
+}
+function repairKnownReplacementDamagedAddressText(value) {
+    return String(value ?? "")
+        .replace(/\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422[\uFFFD?]+\u0412\u0410/giu, "\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422\u0418\u0412\u0410")
+        .replace(/\u041e\u0411\u0429[\uFFFD?]+\u0419/giu, "\u041e\u0411\u0429\u0418\u0419");
+}
 function repairAddressMojibake(value) {
-    const source = String(value ?? "");
-    if (!looksLikeMojibakeForAddress(source)) {
+    const source = repairKnownReplacementDamagedAddressText(String(value ?? ""));
+    if (!looksLikeMojibakeForAddress(source) && !/[\u0080-\u009f\uFFFD]/.test(source)) {
        return source;
    }
    let candidate = source;
    for (let pass = 0; pass < 3; pass += 1) {
        let improved = false;
        try {
-            const fromWin1251 = iconv.encode(candidate, "win1251").toString("utf8");
+            const fromWin1251 = decodeUtf8FromWin1251MojibakeForAddress(candidate);
            if (textMojibakeScoreForAddress(fromWin1251) > textMojibakeScoreForAddress(candidate)) {
                candidate = fromWin1251;
                improved = true;
@ -1942,11 +1964,16 @@ function repairAddressMojibake(value) {
            }
        }
        catch (_error) { }
+        const repairedKnownText = repairKnownReplacementDamagedAddressText(candidate);
+        if (repairedKnownText !== candidate) {
+            candidate = repairedKnownText;
+            improved = true;
+        }
        if (!improved) {
            break;
        }
    }
-    return candidate;
+    return repairKnownReplacementDamagedAddressText(candidate);
 }
 function sanitizeOutgoingAssistantText(value, fallback = "Не смог сформировать читаемый ответ. Уточните запрос.") {
    const repaired = repairAddressMojibake(String(value ?? ""));
@ -4268,18 +4295,25 @@ function scoreOrganizationMentionInMessage(message, organization) {
                variantScore = Math.max(variantScore, variant.length * 5);
                continue;
            }
+            let fuzzyTokenScore = 0;
            const fuzzyMatched = messageTokens.some((messageToken) => {
                if (messageToken === variant) {
+                    fuzzyTokenScore = Math.max(fuzzyTokenScore, variant.length * 5);
                    return true;
                }
                if (messageToken.length >= 5 && variant.length >= 5) {
-                    return messageToken.startsWith(variant) || variant.startsWith(messageToken);
+                    const prefixMatched = messageToken.startsWith(variant) || variant.startsWith(messageToken);
+                    if (prefixMatched) {
+                        const prefixLength = Math.min(messageToken.length, variant.length);
+                        fuzzyTokenScore = Math.max(fuzzyTokenScore, variant.length * (prefixLength >= 7 ? 5 : 3));
+                    }
+                    return prefixMatched;
                }
                return false;
            });
            if (fuzzyMatched) {
                matched = true;
-                variantScore = Math.max(variantScore, Math.max(20, variant.length * 3));
+                variantScore = Math.max(variantScore, Math.max(20, fuzzyTokenScore || variant.length * 3));
            }
        }
        if (matched) {
--- a/llm_normalizer/backend/tests/addressTextRepair.test.ts
+++ b/llm_normalizer/backend/tests/addressTextRepair.test.ts
@ -0,0 +1,35 @@
+import { describe, expect, it } from "vitest";
+import { repairAddressMojibakeText } from "../src/services/addressTextRepair";
+
+describe("address text mojibake repair", () => {
+  it("preserves C1 bytes when repairing old autorun Russian text", () => {
+    const damagedTitle = String.fromCharCode(
+      0x420, 0x2018, 0x420, 0x45b, 0x420, 0x203a, 0x420, 0xac, 0x420, 0x401, 0x420, 0x45b, 0x420,
+      0x2122, 0x20, 0x420, 0x45b, 0x420, 0x2018, 0x420, 0xa9, 0x420, 0x98, 0x420, 0x2122, 0x20,
+      0x420, 0xa0, 0x421, 0x453, 0x421, 0x2021, 0x420, 0x405, 0x420, 0xb0, 0x421, 0x40f, 0x20,
+      0x421, 0x403, 0x420, 0xb5, 0x421, 0x403, 0x421, 0x403, 0x420, 0x451, 0x421, 0x40f, 0x20,
+      0x31, 0x36, 0x2e, 0x30, 0x34, 0x2e, 0x32, 0x30, 0x32, 0x36, 0x2c, 0x20, 0x32, 0x31,
+      0x3a, 0x32, 0x36, 0x3a, 0x30, 0x36
+    );
+    const damagedAlternative = String.fromCharCode(
+      0x420, 0x452, 0x420, 0x203a, 0x420, 0xac, 0x420, 0x45e, 0x420, 0x2022, 0x420, 0xa0, 0x420,
+      0x45c, 0x420, 0x452, 0x420, 0x45e, 0x420, 0x98, 0x420, 0x2019, 0x420, 0x452
+    );
+
+    expect(repairAddressMojibakeText(damagedTitle)).toBe(
+      "\u0411\u041e\u041b\u042c\u0428\u041e\u0419 \u041e\u0411\u0429\u0418\u0419 \u0420\u0443\u0447\u043d\u0430\u044f \u0441\u0435\u0441\u0441\u0438\u044f 16.04.2026, 21:26:06"
+    );
+    expect(repairAddressMojibakeText(damagedAlternative)).toBe(
+      "\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422\u0418\u0412\u0410"
+    );
+  });
+
+  it("repairs already lossy known replacement fragments", () => {
+    expect(repairAddressMojibakeText("\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422\uFFFD?\u0412\u0410")).toBe(
+      "\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422\u0418\u0412\u0410"
+    );
+    expect(repairAddressMojibakeText("\u0411\u041e\u041b\u042c\u0428\u041e\u0419 \u041e\u0411\u0429\uFFFD\u0419")).toBe(
+      "\u0411\u041e\u041b\u042c\u0428\u041e\u0419 \u041e\u0411\u0429\u0418\u0419"
+    );
+  });
+});
--- a/llm_normalizer/backend/tests/assistantAddressFollowupContext.test.ts
+++ b/llm_normalizer/backend/tests/assistantAddressFollowupContext.test.ts
@ -2276,6 +2276,101 @@ describe("assistant address follow-up carryover", () => {
    expect(normalizerService.normalize).not.toHaveBeenCalled();
  });

+  it("continues the original inventory query after replacement-damaged organization clarification", async () => {
+    const calls: Array<{ message: string; options?: any }> = [];
+    const firstMessage = "кайф - что там на складе по остаткам?";
+    const secondMessage = "АЛЬТЕРНАТ\uFFFD?ВА";
+    const repairedSecondMessage = "\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422\u0418\u0412\u0410";
+    const addressQueryService = {
+      tryHandle: vi.fn(async (message: string, options?: any) => {
+        calls.push({ message, options });
+        if (message === firstMessage) {
+          return buildAddressLimitedLaneResult("missing_anchor", {
+            reply_text: [
+              "Нужно уточнить организацию, чтобы не смешивать компании в одном ответе.",
+              "Сейчас в доступном контуре вижу такие организации:",
+              "- ООО Альтернатива Плюс",
+              "- ООО Лайсвуд",
+              "- РАЙМ"
+            ].join("\n"),
+            debug: {
+              ...buildAddressLimitedLaneResult("missing_anchor").debug,
+              detected_intent: "inventory_on_hand_as_of_date",
+              extracted_filters: {
+                as_of_date: "2026-04-15"
+              },
+              selected_recipe: null,
+              organization_candidates: ["ООО Альтернатива Плюс", "ООО Лайсвуд", "РАЙМ"],
+              reasons: ["organization_clarification_required", "multiple_known_organizations_detected"]
+            }
+          });
+        }
+        if (
+          (message === secondMessage || message === repairedSecondMessage) &&
+          options?.followupContext &&
+          options?.activeOrganization === "ООО Альтернатива Плюс"
+        ) {
+          return buildAddressLaneResult({
+            reply_text: "На 15.04.2026 по ООО Альтернатива Плюс подтвержден складской остаток по всем складам.",
+            debug: {
+              ...buildAddressLaneResult().debug,
+              detected_intent: "inventory_on_hand_as_of_date",
+              extracted_filters: {
+                as_of_date: "2026-04-15",
+                organization: "ООО Альтернатива Плюс"
+              },
+              reasons: ["address_followup_context_applied", "organization_grounded_from_scope_candidates"]
+            }
+          });
+        }
+        return null;
+      })
+    } as any;
+
+    const normalizerService = {
+      normalize: vi.fn(async () => ({
+        assistant_reply: "normalizer_fallback_should_not_be_used",
+        reply_type: "partial_coverage",
+        debug: {}
+      }))
+    } as any;
+
+    const sessions = new AssistantSessionStore();
+    const service = new AssistantService(
+      normalizerService,
+      sessions as any,
+      {} as any,
+      { persistSession: vi.fn() } as any,
+      addressQueryService
+    );
+
+    const sessionId = `asst-address-org-clarification-damaged-${Date.now()}`;
+    const first = await service.handleMessage({
+      session_id: sessionId,
+      user_message: firstMessage,
+      useMock: true
+    } as any);
+    expect(first.ok).toBe(true);
+    expect(first.reply_type).toBe("partial_coverage");
+
+    const second = await service.handleMessage({
+      session_id: sessionId,
+      user_message: secondMessage,
+      useMock: true
+    } as any);
+
+    expect(second.ok).toBe(true);
+    expect(second.reply_type).toBe("factual");
+    expect(calls).toHaveLength(2);
+    expect(calls[1].message).toBe(repairedSecondMessage);
+    expect(calls[1].options?.activeOrganization).toBe("ООО Альтернатива Плюс");
+    expect(calls[1].options?.knownOrganizations).toEqual(["ООО Альтернатива Плюс", "ООО Лайсвуд", "РАЙМ"]);
+    expect(calls[1].options?.followupContext?.previous_intent).toBe("inventory_on_hand_as_of_date");
+    expect(calls[1].options?.followupContext?.previous_filters?.organization).toBe("ООО Альтернатива Плюс");
+    expect(calls[1].options?.followupContext?.root_filters?.organization).toBe("ООО Альтернатива Плюс");
+    expect(normalizerService.normalize).not.toHaveBeenCalled();
+  });
+
  it("keeps historical inventory date follow-up alive after company clarification and a capability answer", async () => {
    const calls: Array<{ message: string; options?: any }> = [];
    const firstMessage = "покажи остатки по складу";
--- a/llm_normalizer/backend/tests/assistantOrganizationMatcher.test.ts
+++ b/llm_normalizer/backend/tests/assistantOrganizationMatcher.test.ts
@ -42,6 +42,26 @@ describe("assistant organization matcher", () => {
    expect(score).toBeGreaterThanOrEqual(90);
  });

+  it("matches replacement-damaged organization clarification when the live candidate is unique", () => {
+    const resolved = resolveOrganizationSelectionFromMessage("АЛЬТЕРНАТ\uFFFD?ВА", [
+      "ООО Альтернатива Плюс",
+      "ООО Лайсвуд",
+      "РАЙМ"
+    ]);
+
+    expect(resolved).toBe("ООО Альтернатива Плюс");
+  });
+
+  it("keeps replacement-damaged organization clarification ambiguous when candidates share the token", () => {
+    const resolved = resolveOrganizationSelectionFromMessage("АЛЬТЕРНАТ\uFFFD?ВА", [
+      "ООО Альтернатива Плюс",
+      "ООО Альтернатива Минус",
+      "ООО Лайсвуд"
+    ]);
+
+    expect(resolved).toBeNull();
+  });
+
  it("treats minor live label corruption as the same organization entity", () => {
    expect(organizationsLikelySameEntity("Альтернатива Плюс", 'ООО "Альтернати"а Плюс"')).toBe(true);
  });
--- a/llm_normalizer/backend/tests/autoRunsQuestionSplit.test.ts
+++ b/llm_normalizer/backend/tests/autoRunsQuestionSplit.test.ts
@ -31,4 +31,26 @@ describe("autoruns question extraction", () => {
    expect(parsed[0]).toMatch(/поставщик/i);
    expect(parsed[0]).toMatch(/коротко/i);
  });
+
+  it("repairs old autorun C1-control mojibake before exposing cards and questions", () => {
+    const damagedTitle = String.fromCharCode(
+      0x420, 0x2018, 0x420, 0x45b, 0x420, 0x203a, 0x420, 0xac, 0x420, 0x401, 0x420, 0x45b, 0x420,
+      0x2122, 0x20, 0x420, 0x45b, 0x420, 0x2018, 0x420, 0xa9, 0x420, 0x98, 0x420, 0x2122, 0x20,
+      0x420, 0xa0, 0x421, 0x453, 0x421, 0x2021, 0x420, 0x405, 0x420, 0xb0, 0x421, 0x40f, 0x20,
+      0x421, 0x403, 0x420, 0xb5, 0x421, 0x403, 0x421, 0x403, 0x420, 0x451, 0x421, 0x40f, 0x20,
+      0x31, 0x36, 0x2e, 0x30, 0x34, 0x2e, 0x32, 0x30, 0x32, 0x36, 0x2c, 0x20, 0x32, 0x31,
+      0x3a, 0x32, 0x36, 0x3a, 0x30, 0x36
+    );
+    const damagedAlternative = String.fromCharCode(
+      0x420, 0x452, 0x420, 0x203a, 0x420, 0xac, 0x420, 0x45e, 0x420, 0x2022, 0x420, 0xa0, 0x420,
+      0x45c, 0x420, 0x452, 0x420, 0x45e, 0x420, 0x98, 0x420, 0x2019, 0x420, 0x452
+    );
+
+    expect(__autoRunsQuestionTestUtils.repairAutogenMojibake(damagedTitle)).toBe(
+      "\u0411\u041e\u041b\u042c\u0428\u041e\u0419 \u041e\u0411\u0429\u0418\u0419 \u0420\u0443\u0447\u043d\u0430\u044f \u0441\u0435\u0441\u0441\u0438\u044f 16.04.2026, 21:26:06"
+    );
+    expect(__autoRunsQuestionTestUtils.repairAutogenMojibake(damagedAlternative)).toBe(
+      "\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422\u0418\u0412\u0410"
+    );
+  });
 });
--- a/llm_normalizer/backend/tests/evalRuntimeQuestionSplit.test.ts
+++ b/llm_normalizer/backend/tests/evalRuntimeQuestionSplit.test.ts
@ -33,4 +33,37 @@ describe("eval runtime question splitting", () => {
    expect(parsed[0]).toMatch(/поставщик/i);
    expect(parsed[0]).toMatch(/коротко/i);
  });
+
+  it("repairs mojibake questions before runtime job materialization", () => {
+    const parsed = __evalRouteTestUtils.normalizeRuntimeQuestions([
+      "РєР°Р№С„ - С‡С‚Рѕ С‚Р°Рј РЅР° СЃРєР»Р°РґРµ РїРѕ РѕСЃС‚Р°С‚РєР°Рј?"
+    ]);
+
+    expect(parsed).toEqual(["кайф - что там на складе по остаткам?"]);
+  });
+
+  it("repairs damaged clarification as one scenario turn when splitting is disabled", () => {
+    const parsed = __evalRouteTestUtils.normalizeRuntimeQuestions(
+      ["\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422\uFFFD?\u0412\u0410"],
+      {
+        dedupe: false,
+        splitCandidates: false
+      }
+    );
+
+    expect(parsed).toEqual(["\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422\u0418\u0412\u0410"]);
+  });
+
+  it("repairs C1-control autorun clarification before runtime job materialization", () => {
+    const damagedAlternative = String.fromCharCode(
+      0x420, 0x452, 0x420, 0x203a, 0x420, 0xac, 0x420, 0x45e, 0x420, 0x2022, 0x420, 0xa0, 0x420,
+      0x45c, 0x420, 0x452, 0x420, 0x45e, 0x420, 0x98, 0x420, 0x2019, 0x420, 0x452
+    );
+    const parsed = __evalRouteTestUtils.normalizeRuntimeQuestions([damagedAlternative], {
+      dedupe: false,
+      splitCandidates: false
+    });
+
+    expect(parsed).toEqual(["\u0410\u041b\u042c\u0422\u0415\u0420\u041d\u0410\u0422\u0418\u0412\u0410"]);
+  });
 });
--- a/llm_normalizer/data/eval_cases/assistant_saved_session_runtime_job-elT8zDcc9I.json
+++ b/llm_normalizer/data/eval_cases/assistant_saved_session_runtime_job-elT8zDcc9I.json
@ -0,0 +1,123 @@
+{
+  "suite_id": "assistant_saved_session_runtime_job-elT8zDcc9I",
+  "suite_version": "0.1.0",
+  "schema_version": "assistant_saved_session_runtime_v0_1",
+  "title": "Р‘РћР›Р¬РЁРћР™ РћР‘Р©РР™ Р СѓС‡РЅР°СЏ СЃРµСЃСЃРёСЏ 16.04.2026, 21:26:06",
+  "scenario_count": 1,
+  "case_ids": [
+    "SAVED-001"
+  ],
+  "cases": [
+    {
+      "case_id": "SAVED-001",
+      "scenario_tag": "saved_user_sessions_runtime",
+      "title": "Р‘РћР›Р¬РЁРћР™ РћР‘Р©РР™ Р СѓС‡РЅР°СЏ СЃРµСЃСЃРёСЏ 16.04.2026, 21:26:06",
+      "question_type": "followup",
+      "broadness_level": "medium",
+      "turns": [
+        {
+          "user_message": "приветик - че как там дела"
+        },
+        {
+          "user_message": "расскажи что можешь интересного"
+        },
+        {
+          "user_message": "кайф - что там на складе по остаткам?"
+        },
+        {
+          "user_message": "АЛЬТЕРНАТ<D090>?ВА"
+        },
+        {
+          "user_message": "а исторические остатки на другие даты умеешь?"
+        },
+        {
+          "user_message": "давай на июль 2017"
+        },
+        {
+          "user_message": "март 2016"
+        },
+        {
+          "user_message": "По выбранному объекту \"Рабочая станция универсального специалиста (индивидуальное изготовление)\": где взяли это?"
+        },
+        {
+          "user_message": "а кому продали?"
+        },
+        {
+          "user_message": "у тебя написано кто контрагент: рабочая станция - это ошибка?"
+        },
+        {
+          "user_message": "ндс можешь прикинуть на дату покупки рабочей станции?"
+        },
+        {
+          "user_message": "а какой ндс мы должны сгрузить на март 2020?"
+        },
+        {
+          "user_message": "прикинь какой ндс нам надо заплатить на февраль 2017"
+        },
+        {
+          "user_message": "кто у нас самый доходный клиент за все время"
+        },
+        {
+          "user_message": "кто нам должен денег на май 2017"
+        },
+        {
+          "user_message": "а какой ндс мы должны примерно заплатить за этот период?"
+        },
+        {
+          "user_message": "мы должны комуто денег на сегодня?"
+        },
+        {
+          "user_message": "а нам?"
+        },
+        {
+          "user_message": "какой у нас самый доходный год"
+        },
+        {
+          "user_message": "а за 2017 мы скок заработали?"
+        },
+        {
+          "user_message": "сколько вообще денег мы заработали за все время?"
+        },
+        {
+          "user_message": "ты умеешь считать дельту по договорам?"
+        },
+        {
+          "user_message": "по чепурнову покажи все доки"
+        },
+        {
+          "user_message": "а по свк"
+        },
+        {
+          "user_message": "а сейчас у нас есть что на складе?"
+        },
+        {
+          "user_message": "что нам отгружал чепурнов? какой товар или услугу?"
+        },
+        {
+          "user_message": "какие остатки на складе на сегодня"
+        },
+        {
+          "user_message": "остатки на март 2016"
+        },
+        {
+          "user_message": "хвосты покажи по счету 60 на август 2022"
+        },
+        {
+          "user_message": "Есть ли остатки товара, которые закупались очень давно"
+        },
+        {
+          "user_message": "Какие конкретно номенклатуры формируют остаток по складу на май 2020"
+        },
+        {
+          "user_message": "а по Альтернативе Плюс сколько лет активности в базе 1С?"
+        },
+        {
+          "user_message": "Как ты оценишь деятельность компании?"
+        },
+        {
+          "user_message": "какое нетто по деньгам с Группа СВК за 2020 год: сколько получили и сколько заплатили?"
+        }
+      ]
+    }
+  ]
+}