p28

DEVELOPER_POLICIES.md

@@ -123,3 +123,18 @@ Use this checklist for every relevant PR:
 - [ ] `mto:agent:regression:test` is OK.
 - [ ] The protected functional flows were manually checked if the touched area can affect them.
 - [ ] README or patch README documents the reason for any intentionally accepted technical fallback.
+## 9. Language cleanup ownership
+
+Generic language cleanup must use `config/retriex/language.yaml` and its cleanup profiles.
+
+Rules:
+
+- add generic German stopwords to `stopword_groups`, not to domain YAML files
+- add user wording such as `ich suche`, `zeige mir` or `habt ihr` to `phrase_groups`
+- add table/list/overview wording to `meta_term_groups`
+- keep commerce intent, product-role, measurement and routing terms in their owning domain YAML
+- never remove protected terms such as `nicht`, `kein`, `testomat`, `indikator`, `ph`, `rx`, `th`, `tc` or `0,02` through generic cleanup
+- prefer `cleanup_profile: ...` references over copied token lists
+
+See `RETRIEX_LANGUAGE_CLEANUP_GUIDE.md` for the detailed ownership rules.
+

RETRIEX_LANGUAGE_CLEANUP_GUIDE.md

@@ -0,0 +1,77 @@
+# RetrieX Language Cleanup Guide
+
+Status: binding for RetrieX 1.5.3+ cleanup-profile work.
+
+This guide defines where language, interaction, commerce and domain tokens belong. Its goal is to keep YAML maintenance simple and avoid duplicated keyword lists.
+
+## 1. Central language cleanup lives in `language.yaml`
+
+Use `config/retriex/language.yaml` for generic language noise only.
+
+Allowed here:
+
+- German function words: `der`, `die`, `das`, `ein`, `eine`, `mit`, `und`, `oder`, `ist`, `sind`, `kann`
+- conversation filler words: `bitte`, `mal`, `gerne`, `noch`, `dazu`, `also`
+- user instruction phrases: `ich suche`, `suche nach`, `zeige mir`, `gib mir`, `habt ihr`, `gibt es`
+- presentation/meta terms: `tabelle`, `liste`, `übersicht`, `tabellarisch`, `auflistung`
+- protected terms that must not be removed generically
+
+Do not add product families, measurement parameters, intent terms or shop semantics here.
+
+## 2. Use cleanup profiles instead of copying lists
+
+Domain configs should reference a cleanup profile whenever they need generic language cleanup.
+
+Current profiles:
+
+- `commerce_query`: cleanup for shop/search query text
+- `rag_evidence`: cleanup for evidence/answer-consistency checks
+- `shop_context_fallback`: cleanup for history-based shop context fallback
+
+Preferred pattern:
+
+```yaml
+cleanup_profile: commerce_query
+```
+
+Avoid adding the same generic words again to `commerce.yaml`, `agent.yaml`, `retrieval.yaml` or `intent.yaml`.
+
+## 3. Keep domain semantics in domain configs
+
+These belong outside `language.yaml`:
+
+- commerce intent terms: `shop`, `produkt`, `artikel`, `preis`, `kosten`, `kaufen`, `bestellen`
+- measurement/domain terms: `wasserhärte`, `chlor`, `redox`, `leitfähigkeit`, `ph`, `rx`, `th`, `tc`
+- product-role terms: device, accessory, reagent, spare part and document-role vocabulary
+- routing and answer behavior rules
+- prompt-specific role or grounding rules
+
+## 4. Protected terms are mandatory guardrails
+
+Never remove these generically unless a later patch explicitly changes the guardrail:
+
+- negations: `nicht`, `kein`, `keine`
+- core product/domain anchors: `testomat`, `indikator`, `indikatortyp`
+- short model/parameter tokens: `ph`, `rx`, `th`, `tc`
+- important numeric anchors: `0,02`
+
+When in doubt, add terms to `protected_terms` rather than removing them through a broad stopword group.
+
+## 5. Change process
+
+Before adding a new token list:
+
+1. Ask whether it is generic language noise.
+2. If yes, add it to `language.yaml` under the correct group/profile.
+3. If no, keep it in the owning domain YAML.
+4. Do not introduce PHP-only token lists.
+5. Run the required checks.
+
+Required checks:
+
+```bash
+bin/console mto:agent:config:validate
+bin/console mto:agent:regression:test
+bin/console mto:agent:config:audit-source --details
+bin/console mto:agent:config:audit-patterns --details
+```

RETRIEX_PATCH_25B_SHOP_CONTEXT_FALLBACK_CLEANUP_PROFILE_HOTFIX_README.md

@@ -0,0 +1,13 @@
+# RetrieX Patch p25b - Shop Context Fallback Cleanup Profile Hotfix
+
+## Goal
+Fix p25 validation failure when `agent.shop_prompt.meta_query_guard.cleanup_profile` is missing in partially updated environments.
+
+## Scope
+- Keeps `cleanup_profile: shop_context_fallback` in `config/retriex/agent.yaml`.
+- Adds a migration-safe fallback in `AgentRunnerConfig::getShopQueryContextFallbackCleanupProfile()`.
+- Keeps the actual value YAML-owned; the fallback only prevents failed intermediate states.
+
+## Safety
+No legacy lists are removed.
+No Commerce, Retrieval scoring, Prompt wording, or shop matching logic is changed.

RETRIEX_PATCH_25C_SHOP_CONTEXT_FALLBACK_CLEANUP_PROFILE_VALIDATION_FIX_README.md

@@ -0,0 +1,13 @@
+# RetrieX Patch p25c - Shop Context Fallback Cleanup Profile Validation Fix
+
+## Goal
+Fix the p25/p25b validation failure for `agent.shop_prompt.meta_query_guard.cleanup_profile`.
+
+## Scope
+- Keeps the YAML-owned value `shop_context_fallback` in `config/retriex/agent.yaml`.
+- Keeps the runtime wiring from p25.
+- Makes validation migration-safe: if the YAML value is missing in an already partially updated environment, validation uses the configured fallback from `AgentRunnerConfig` instead of failing with an empty-string error.
+
+## Safety
+No legacy lists are removed.
+No Commerce, Retrieval scoring, Prompt wording, or shop matching logic is changed.

RETRIEX_PATCH_25_SHOP_CONTEXT_FALLBACK_CLEANUP_PROFILE_README.md

@@ -0,0 +1,14 @@
+# RetrieX Patch p25 - Shop Context Fallback Cleanup Profile
+
+## Goal
+Wire the Shop context fallback/meta guard to the central `shop_context_fallback` cleanup profile from `language.yaml`.
+
+## Scope
+- Adds `shop_prompt.meta_query_guard.cleanup_profile: shop_context_fallback` to `config/retriex/agent.yaml`.
+- Merges central profile stopwords, user-instruction phrases and presentation meta terms with the existing legacy meta/fallback lists.
+- Keeps all legacy lists in place.
+- Adds config validation for the referenced cleanup profile.
+
+## Safety
+No legacy lists are removed in this patch.
+No Commerce, Retrieval scoring, Prompt wording, or shop matching logic is changed.

RETRIEX_PATCH_26_LANGUAGE_CLEANUP_YAML_DEDUP_README.md

@@ -0,0 +1,25 @@
+# RetrieX Patch 26 - Language Cleanup YAML Dedup
+
+Purpose: simplify the maintenance layer after p21-p25.
+
+Changes:
+
+- Removes duplicate generic language noise from legacy Commerce/Agent lists when the terms are already supplied by `language.yaml` cleanup profiles.
+- Keeps all domain-specific, commerce-specific and regression-sensitive tokens in their original domain configs.
+- Adds comments that mark remaining lists as legacy/domain override lists.
+- Does not change profile names, service wiring, routing or retrieval logic.
+
+Guardrails:
+
+- Do not move protected terms into generic removal logic.
+- Keep intent/domain terms in their domain YAML files.
+- Only remove a legacy term when it is covered by an active cleanup profile or clearly duplicated in the same list.
+
+Required checks:
+
+```bash
+bin/console mto:agent:config:validate
+bin/console mto:agent:regression:test
+bin/console mto:agent:config:audit-source --details
+bin/console mto:agent:config:audit-patterns --details
+```

RETRIEX_PATCH_27_LANGUAGE_CLEANUP_REGRESSION_GUARDS_README.md

@@ -0,0 +1,32 @@
+# RetrieX Patch 27 - Language Cleanup Regression Guards
+
+## Ziel
+
+Absicherung der neuen Language-Cleanup-Profile aus p21-p26, ohne weitere produktive Runtime-Logik umzubauen.
+
+## Inhalt
+
+- erweitert `governance.yaml` um prüfbare Language-Cleanup-Guardrails
+- prüft required Cleanup-Profile:
+  - `commerce_query`
+  - `rag_evidence`
+  - `shop_context_fallback`
+- prüft repräsentative Pflichtbegriffe je Profil
+- schützt zentrale Begriffe vor generischem Entfernen, u. a. `nicht`, `kein`, `keine`, `testomat`, `indikator`, `indikatortyp`, `ph`, `rx`, `th`, `tc`, `0,02`
+- erweitert Config-Dump/Validation um Cleanup-Profile
+- erweitert `mto:agent:regression:test` um Language-Cleanup-Checks
+
+## Einspielen
+
+```bash
+unzip -o p27.zip -d /path/to/retriex
+cd /path/to/retriex
+bin/console mto:agent:config:validate
+bin/console mto:agent:regression:test
+bin/console mto:agent:config:audit-source --details
+bin/console mto:agent:config:audit-patterns --details
+```
+
+## Hinweis
+
+Dieser Patch loescht keine Listen und aendert keine fachliche Runtime-Entscheidung. Er macht die neue Cleanup-Schicht regressionssicherer.

RETRIEX_PATCH_28_LANGUAGE_CLEANUP_DOCUMENTATION_README.md

@@ -0,0 +1,22 @@
+# RetrieX Patch 28 - Language Cleanup Documentation
+
+Purpose: document the new Language Cleanup Profile ownership rules for RetrieX 1.5.3+.
+
+Changes:
+
+- adds `RETRIEX_LANGUAGE_CLEANUP_GUIDE.md`
+- extends `DEVELOPER_POLICIES.md` with language cleanup ownership rules
+- no runtime code changes
+- no YAML value changes
+- no list deletion
+
+Install:
+
+```bash
+unzip -o p28.zip -d /path/to/retriex
+cd /path/to/retriex
+bin/console mto:agent:config:validate
+bin/console mto:agent:regression:test
+bin/console mto:agent:config:audit-source --details
+bin/console mto:agent:config:audit-patterns --details
+```

config/retriex/agent.yaml

@@ -208,6 +208,8 @@ parameters:
 
     rag_evidence_guard:
       cleanup_profile: rag_evidence
+      # Legacy/domain override list. Generic German stopwords are provided by
+      # language cleanup profile `rag_evidence`. Keep RAG/product-role terms here.
       stop_terms:
         - suche
         - suchen
@@ -215,24 +217,8 @@ parameters:
         - finden
         - zeige
         - einen
-        - eine
         - einem
-        - einer
-        - der
-        - die
-        - das
-        - den
-        - dem
-        - des
-        - für
-        - fuer
-        - mit
         - ohne
-        - und
-        - oder
-        - kann
-        - können
-        - koennen
         - messen
         - messung
         - tester
@@ -466,19 +452,16 @@ parameters:
           - '/\b(?:indikator(?:typ)?|indicator(?:\s+type)?|reagenz(?:satz|typ)?|reagent(?:\s+set|\s+type)?|typ|type)\s+[A-Za-zÄÖÜäöüß]{0,8}\s*\d{1,5}(?:\s*[A-ZÄÖÜ]{1,4})?(?:\s*%)?\b/iu'
       meta_query_guard:
         enabled: true
+        cleanup_profile: shop_context_fallback
         context_fallback_enabled: true
         context_fallback_question_limit: 12
         context_fallback_history_budget_chars: 20000
         context_fallback_use_full_history: true
         context_fallback_max_terms: 6
+        # Legacy/domain override list. Generic stopwords, user-instruction
+        # phrases and presentation terms are provided by profile
+        # `shop_context_fallback`. Keep shop/price/domain terms here.
         context_fallback_filter_terms:
-          - mit
-          - tabelle
-          - tabellarisch
-          - übersicht
-          - uebersicht
-          - liste
-          - auflistung
           - preis
           - preise
           - preisen
@@ -492,8 +475,6 @@ parameters:
           - welches
           - welchem
           - welchen
-          - ist
-          - sind
           - gut
           - geeignet
           - was
@@ -515,24 +496,15 @@ parameters:
           - nehmen
           - zur
           - zum
-          - für
-          - fuer
           - messen
           - gemessen
         meta_only_terms:
           - shop
-          - tabelle
-          - tabellarisch
-          - übersicht
-          - uebersicht
-          - liste
-          - auflistung
           - preis
           - preise
           - preisen
           - kosten
           - kostet
-          - mit
           - shopsuche
           - shop-suche
           - suche
@@ -542,22 +514,14 @@ parameters:
           - find
           - zeige
           - zeig
-          - bitte
-          - mal
           - im
           - in
           - nach
           - danach
-          - dazu
           - damit
           - dafür
           - dafuer
           - hierzu
-          - den
-          - die
-          - das
-          - der
-          - dem
       language_preservation:
         enabled: true
         language_markers:

config/retriex/commerce.yaml

@@ -21,14 +21,11 @@ parameters:
       - horiba
       - neomeris
 
+    # Legacy/domain override list. General user-instruction phrases are provided by
+    # language cleanup profile `commerce_query`. Keep only commerce-specific or
+    # historically sensitive phrases here.
     phrases_to_remove:
-      - ich suche
       - suche
-      - habt ihr
-      - gibt es
-      - gebe mir
-      - gib mir
-      - zeige mir
       - welches gerät
       - welche gerät
       - welches modell
@@ -42,8 +39,6 @@ parameters:
       - welcher
       - welches
       - welchen
-      - sind
-      - ist
       - geeignet
       - geeigent
       - verfügbarkeit
@@ -53,27 +48,17 @@ parameters:
       - kurze
       - ich
 
+    # Legacy/domain override list. Generic German stopwords and conversation
+    # filler terms are provided by language cleanup profile `commerce_query`.
     filter_search_tokens:
-      - auch
-      - noch
-      - nochmal
       - zusätzlich
-      - dazu
-      - davon
       - stattdessen
-      - bitte
       - preiswerte
-      - gern
       - lösung
-      - eine
       - größer
       - würde
-      - gerne
       - welchem
-      - kann
-      - mit
       - mein
-      - größer
       - zeige
       - zeig
       - such
@@ -81,33 +66,20 @@ parameters:
       - finde
       - find
       - mir
-      - mal
       - von
       - im
       - in
-      - für
-      - fuer
       - welche
       - welcher
       - welches
       - welchen
-      - sind
       - zur
-      - ist
       - geeignet
       - geeigent
       - verfügbarkeit
       - verfuegbarkeit
       - prüfe
       - pruefe
-      - den
-      - die
-      - das
-      - der
-      - dem
-      - des
-      - und
-      - oder
       - sowie
       - seine
       - seinen
@@ -129,8 +101,6 @@ parameters:
       - kostet
       - kosten
       - ua
-      - ein
-      - also
       - gut
       - gute
       - guten

config/retriex/governance.yaml

@@ -68,10 +68,45 @@ parameters:
       protected_stopword_terms:
         - nicht
         - kein
+        - keine
         - welche
         - testomat
         - indikator
+        - indikatortyp
+        - ph
+        - rx
+        - th
+        - tc
         - '0,02'
+      required_cleanup_profiles:
+        - commerce_query
+        - rag_evidence
+        - shop_context_fallback
+      required_profile_terms:
+        commerce_query:
+          stopwords:
+            - der
+            - mit
+            - bitte
+          phrases:
+            - ich suche
+            - suche im shop
+        rag_evidence:
+          stopwords:
+            - der
+            - mit
+            - bitte
+        shop_context_fallback:
+          stopwords:
+            - der
+            - mit
+            - bitte
+          phrases:
+            - zeige mir
+            - suche im shop
+          meta_terms:
+            - tabelle
+            - übersicht
     core_pattern_audit:
       source_roots:
         - src

config/retriex/language.yaml

@@ -53,12 +53,13 @@ parameters:
 
     # Central language cleanup structure for RetrieX 1.5.3+.
     # Legacy key `words` above remains the runtime-compatible default list.
-    # New cleanup profiles are introduced additively and are not yet wired into
+    # Cleanup profiles are the preferred home for generic language noise.
-    # Commerce/Agent runtime logic in this patch.
+    # Domain configs should only keep domain-specific overrides.
     protected_terms:
       - nicht
       - kein
       - keine
+      - welche
       - testomat
       - indikator
       - indikatortyp

src/Agent/AgentRunner.php

@@ -1910,10 +1910,7 @@ final readonly class AgentRunner
 
         $filterTerms = [];
 
-        foreach (array_merge(
+        foreach ($this->getShopQueryContextFallbackFilterTerms() as $term) {
-            $this->agentRunnerConfig->getShopQueryMetaOnlyTerms(),
-            $this->agentRunnerConfig->getShopQueryContextFallbackFilterTerms()
-        ) as $term) {
             foreach ($this->tokenizeShopQueryCandidate($term) as $token) {
                 $filterTerms[$token] = true;
             }
@@ -1941,6 +1938,43 @@ final readonly class AgentRunner
         return implode(' ', $out);
     }
 
+    /** @return string[] */
+    private function getShopQueryContextFallbackFilterTerms(): array
+    {
+        $profileName = $this->agentRunnerConfig->getShopQueryContextFallbackCleanupProfile();
+
+        return $this->mergeUniqueStrings(
+            $this->mergeUniqueStrings(
+                $this->languageCleanupConfig->getStopWordsForProfile($profileName),
+                $this->languageCleanupConfig->getPhrasesForProfile($profileName)
+            ),
+            $this->mergeUniqueStrings(
+                $this->languageCleanupConfig->getMetaTermsForProfile($profileName),
+                $this->mergeUniqueStrings(
+                    $this->agentRunnerConfig->getShopQueryMetaOnlyTerms(),
+                    $this->agentRunnerConfig->getShopQueryContextFallbackFilterTerms()
+                )
+            )
+        );
+    }
+
+    /** @return string[] */
+    private function getShopQueryMetaGuardTerms(): array
+    {
+        $profileName = $this->agentRunnerConfig->getShopQueryContextFallbackCleanupProfile();
+
+        return $this->mergeUniqueStrings(
+            $this->mergeUniqueStrings(
+                $this->languageCleanupConfig->getStopWordsForProfile($profileName),
+                $this->languageCleanupConfig->getPhrasesForProfile($profileName)
+            ),
+            $this->mergeUniqueStrings(
+                $this->languageCleanupConfig->getMetaTermsForProfile($profileName),
+                $this->agentRunnerConfig->getShopQueryMetaOnlyTerms()
+            )
+        );
+    }
+
     /**
      * @return string[]
      */
@@ -1972,7 +2006,7 @@ final readonly class AgentRunner
         }
 
         $metaTerms = [];
-        foreach ($this->agentRunnerConfig->getShopQueryMetaOnlyTerms() as $term) {
+        foreach ($this->getShopQueryMetaGuardTerms() as $term) {
             foreach ($this->tokenizeMetaGuardText($term) as $token) {
                 $metaTerms[$token] = true;
             }

src/Config/AgentRunnerConfig.php

@@ -808,6 +808,17 @@ final class AgentRunnerConfig
         return $this->getRequiredBool('shop_prompt.meta_query_guard.enabled');
     }
 
+    public function getShopQueryContextFallbackCleanupProfile(): string
+    {
+        $value = $this->optionalValue('shop_prompt.meta_query_guard.cleanup_profile');
+
+        if (is_string($value) && trim($value) !== '') {
+            return trim($value);
+        }
+
+        return 'shop_context_fallback';
+    }
+
     /**
      * @return string[]
      */

src/Config/GovernanceConfig.php

@@ -132,6 +132,48 @@ final class GovernanceConfig
         return $this->requiredStringList('language.protected_stopword_terms');
     }
 
+    /** @return string[] */
+    public function getLanguageRequiredCleanupProfiles(): array
+    {
+        return $this->requiredStringList('language.required_cleanup_profiles');
+    }
+
+    /** @return array<string, array{stopwords:string[], phrases:string[], meta_terms:string[]}> */
+    public function getLanguageRequiredProfileTerms(): array
+    {
+        $value = $this->requiredValue('language.required_profile_terms');
+        if (!is_array($value)) {
+            throw $this->invalid('language.required_profile_terms', 'must be a map of cleanup profile term lists');
+        }
+
+        $out = [];
+        foreach ($value as $profileName => $profileTerms) {
+            if (!is_string($profileName) || trim($profileName) === '' || !is_array($profileTerms)) {
+                throw $this->invalid('language.required_profile_terms', 'must be keyed by non-empty cleanup profile names');
+            }
+
+            $normalizedProfileName = trim($profileName);
+            $out[$normalizedProfileName] = [
+                'stopwords' => $this->normalizeStringList($profileTerms['stopwords'] ?? []),
+                'phrases' => $this->normalizeStringList($profileTerms['phrases'] ?? []),
+                'meta_terms' => $this->normalizeStringList($profileTerms['meta_terms'] ?? []),
+            ];
+
+            if ($out[$normalizedProfileName]['stopwords'] === []
+                && $out[$normalizedProfileName]['phrases'] === []
+                && $out[$normalizedProfileName]['meta_terms'] === []
+            ) {
+                throw $this->invalid('language.required_profile_terms.' . $normalizedProfileName, 'must contain at least one required term');
+            }
+        }
+
+        if ($out === []) {
+            throw $this->invalid('language.required_profile_terms', 'must contain at least one cleanup profile');
+        }
+
+        return $out;
+    }
+
     /** @return string[] */
     public function getCorePatternAuditSourceRoots(): array
     {

src/Config/RetriexEffectiveConfigProvider.php

@@ -117,6 +117,77 @@ final readonly class RetriexEffectiveConfigProvider
             $warnings[] = 'Config validation warning: ' . $warning;
         }
 
+        try {
+            $cleanupProfileNames = $this->languageCleanupConfig->getCleanupProfileNames();
+            foreach ($this->governanceConfig->getLanguageRequiredCleanupProfiles() as $profileName) {
+                $key = 'language_cleanup_profile_' . $this->guardrailCheckKey($profileName);
+                $checks[$key] = in_array($profileName, $cleanupProfileNames, true);
+                if (!$checks[$key]) {
+                    $errors[] = 'Missing required language cleanup profile: ' . $profileName . '.';
+                    continue;
+                }
+
+                $this->languageCleanupConfig->getCleanupProfile($profileName);
+            }
+
+            $legacyStopwords = $this->stopWordsConfig->getStopWords();
+            foreach ($this->governanceConfig->getLanguageProtectedStopwordTerms() as $protectedTerm) {
+                $key = 'language_protected_term_' . $this->guardrailCheckKey($protectedTerm);
+                $checks[$key . '_registered'] = $this->languageCleanupConfig->isProtectedTerm($protectedTerm);
+                if (!$checks[$key . '_registered']) {
+                    $errors[] = 'Missing protected language cleanup term: ' . $protectedTerm . '.';
+                }
+
+                $checks[$key . '_not_legacy_stopword'] = !in_array($protectedTerm, $legacyStopwords, true);
+                if (!$checks[$key . '_not_legacy_stopword']) {
+                    $errors[] = 'Protected language cleanup term is still a legacy stopword: ' . $protectedTerm . '.';
+                }
+
+                foreach ($cleanupProfileNames as $profileName) {
+                    $profile = $this->languageCleanupConfig->getCleanupProfile($profileName);
+                    foreach (['stopwords', 'phrases', 'meta_terms'] as $bucket) {
+                        $bucketKey = $key . '_not_in_' . $this->guardrailCheckKey($profileName . '_' . $bucket);
+                        $checks[$bucketKey] = !in_array($protectedTerm, $profile[$bucket] ?? [], true);
+                        if (!$checks[$bucketKey]) {
+                            $errors[] = sprintf('Protected language cleanup term %s is present in %s.%s.', $protectedTerm, $profileName, $bucket);
+                        }
+                    }
+                }
+            }
+
+            foreach ($this->governanceConfig->getLanguageRequiredProfileTerms() as $profileName => $requiredTerms) {
+                $profile = $this->languageCleanupConfig->getCleanupProfile($profileName);
+                foreach ($requiredTerms as $bucket => $terms) {
+                    foreach ($terms as $term) {
+                        $key = 'language_cleanup_profile_' . $this->guardrailCheckKey($profileName . '_' . $bucket . '_' . $term);
+                        $checks[$key] = in_array($term, $profile[$bucket] ?? [], true);
+                        if (!$checks[$key]) {
+                            $errors[] = sprintf('Missing language cleanup profile term: %s.%s must contain %s.', $profileName, $bucket, $term);
+                        }
+                    }
+                }
+            }
+
+            $checks['commerce_query_cleanup_profile_wired'] = $this->commerceQueryParserConfig->getCleanupProfile() === 'commerce_query';
+            if (!$checks['commerce_query_cleanup_profile_wired']) {
+                $errors[] = 'Commerce query parser is not wired to cleanup profile commerce_query.';
+            }
+
+            $checks['rag_evidence_cleanup_profile_wired'] = $this->agentRunnerConfig->getRagEvidenceCleanupProfile() === 'rag_evidence';
+            if (!$checks['rag_evidence_cleanup_profile_wired']) {
+                $errors[] = 'RAG evidence guard is not wired to cleanup profile rag_evidence.';
+            }
+
+            $checks['shop_context_fallback_cleanup_profile_wired'] = $this->agentRunnerConfig->getShopQueryContextFallbackCleanupProfile() === 'shop_context_fallback';
+            if (!$checks['shop_context_fallback_cleanup_profile_wired']) {
+                $errors[] = 'Shop context fallback is not wired to cleanup profile shop_context_fallback.';
+            }
+
+        } catch (\InvalidArgumentException $e) {
+            $checks['language_cleanup_profile_config_valid'] = false;
+            $errors[] = 'Language cleanup profile guardrails failed: ' . $e->getMessage();
+        }
+
         $importantShortModelTokens = $this->retrieverConfig->importantShortModelTokens();
         foreach ($this->governanceConfig->getRegressionProtectedShortModelTokens() as $token) {
             $key = 'important_short_model_token_' . $this->guardrailCheckKey($token);
@@ -543,6 +614,7 @@ final readonly class RetriexEffectiveConfigProvider
                 ],
                 'meta_query_guard' => [
                     'enabled' => $this->agentRunnerConfig->isShopQueryMetaGuardEnabled(),
+                    'cleanup_profile' => $this->agentRunnerConfig->getShopQueryContextFallbackCleanupProfile(),
                     'context_fallback_use_full_history' => $this->agentRunnerConfig->shouldUseFullHistoryForShopQueryContextFallback(),
                     'meta_only_terms' => $this->agentRunnerConfig->getShopQueryMetaOnlyTerms(),
                     'context_fallback_enabled' => $this->agentRunnerConfig->isShopQueryContextFallbackEnabled(),
@@ -799,7 +871,17 @@ final readonly class RetriexEffectiveConfigProvider
 
     private function languageConfig(): array
     {
-        return ['stopwords' => $this->stopWordsConfig->getStopWords()];
+        $profiles = [];
+        foreach ($this->languageCleanupConfig->getCleanupProfileNames() as $profileName) {
+            $profiles[$profileName] = $this->languageCleanupConfig->getCleanupProfile($profileName);
+        }
+
+        return [
+            'stopwords' => $this->stopWordsConfig->getStopWords(),
+            'protected_terms' => $this->languageCleanupConfig->getProtectedTerms(),
+            'cleanup_profile_names' => $this->languageCleanupConfig->getCleanupProfileNames(),
+            'cleanup_profiles' => $profiles,
+        ];
     }
 
     /** @return array<string, mixed> */
@@ -861,6 +943,8 @@ final readonly class RetriexEffectiveConfigProvider
             $this->governanceConfig->getRegressionShopQueryContextFallbackFilterTerms();
             $this->governanceConfig->getVocabularyProtectedShortModelTokens();
             $this->governanceConfig->getLanguageProtectedStopwordTerms();
+            $this->governanceConfig->getLanguageRequiredCleanupProfiles();
+            $this->governanceConfig->getLanguageRequiredProfileTerms();
             $this->governanceConfig->getCorePatternAuditSourceRoots();
             $this->governanceConfig->getCorePatternAuditExcludedPathPrefixes();
             $this->governanceConfig->getCorePatternAuditExcludedPathPatterns();
@@ -1090,6 +1174,18 @@ final readonly class RetriexEffectiveConfigProvider
         $this->validateStringList($this->toList($ragEvidence['aggregate_evidence_terms'] ?? []), 'agent.rag_evidence_guard.aggregate_evidence_terms', $errors, $warnings);
         $this->validateRegexPatternList($ragEvidence['aggregate_answer_evidence_patterns'] ?? [], 'agent.rag_evidence_guard.aggregate_answer_evidence_patterns', $errors);
 
+        $shopPrompt = is_array($agent['shop_prompt'] ?? null) ? $agent['shop_prompt'] : [];
+        $metaQueryGuard = is_array($shopPrompt['meta_query_guard'] ?? null) ? $shopPrompt['meta_query_guard'] : [];
+        $shopContextCleanupProfile = $metaQueryGuard['cleanup_profile'] ?? null;
+        if (!is_string($shopContextCleanupProfile) || trim($shopContextCleanupProfile) === '') {
+            $shopContextCleanupProfile = $this->agentRunnerConfig->getShopQueryContextFallbackCleanupProfile();
+        } else {
+            $shopContextCleanupProfile = trim($shopContextCleanupProfile);
+        }
+        if (!in_array($shopContextCleanupProfile, $this->languageCleanupConfig->getCleanupProfileNames(), true)) {
+            $errors[] = 'agent.shop_prompt.meta_query_guard.cleanup_profile references unknown language cleanup profile: ' . $shopContextCleanupProfile . '.';
+        }
+
         $this->validateStringListMap($agent['shop_query_optimizer'] ?? [], 'agent.shop_query_optimizer', $errors, $warnings);
         $this->validateRegexPattern($agent['optimized_shop_query_prefix_pattern'] ?? null, 'agent.optimized_shop_query_prefix_pattern', $errors);
 
@@ -1319,10 +1415,40 @@ final readonly class RetriexEffectiveConfigProvider
     {
         $this->validateStringListMap($language, 'language', $errors, $warnings);
         $stopwords = is_array($language['stopwords'] ?? null) ? $language['stopwords'] : [];
-        foreach ($this->governanceConfig->getLanguageProtectedStopwordTerms() as $protected) {
+
-            if (in_array($protected, $stopwords, true)) {
+        try {
-                $errors[] = 'language.stopwords must not contain protected term: ' . $protected . '.';
+            $profileNames = $this->languageCleanupConfig->getCleanupProfileNames();
+
+            foreach ($this->governanceConfig->getLanguageRequiredCleanupProfiles() as $profileName) {
+                if (!in_array($profileName, $profileNames, true)) {
+                    $errors[] = 'language.cleanup_profiles must contain required profile: ' . $profileName . '.';
+                    continue;
+                }
+
+                $this->languageCleanupConfig->getCleanupProfile($profileName);
             }
+
+            foreach ($this->governanceConfig->getLanguageProtectedStopwordTerms() as $protected) {
+                if (in_array($protected, $stopwords, true)) {
+                    $errors[] = 'language.stopwords must not contain protected term: ' . $protected . '.';
+                }
+                if (!$this->languageCleanupConfig->isProtectedTerm($protected)) {
+                    $errors[] = 'language.protected_terms must contain protected term: ' . $protected . '.';
+                }
+            }
+
+            foreach ($this->governanceConfig->getLanguageRequiredProfileTerms() as $profileName => $requiredTerms) {
+                $profile = $this->languageCleanupConfig->getCleanupProfile($profileName);
+                foreach ($requiredTerms as $bucket => $terms) {
+                    foreach ($terms as $term) {
+                        if (!in_array($term, $profile[$bucket] ?? [], true)) {
+                            $errors[] = sprintf('language.cleanup_profiles.%s.%s must contain required term: %s.', $profileName, $bucket, $term);
+                        }
+                    }
+                }
+            }
+        } catch (\InvalidArgumentException $e) {
+            $errors[] = $e->getMessage();
         }
     }