optimize retrieval

src/Agent/AgentRunner.php

@@ -4,6 +4,9 @@ declare(strict_types=1);
 
 namespace App\Agent;
 
+use App\Commerce\CommerceReferenceResolver;
+use App\Commerce\CommerceReferenceStore;
+use App\Commerce\Dto\CommerceReferenceContext;
 use App\Commerce\SearchRepairService;
 use App\Commerce\ShopSearchService;
 use App\Config\AgentRunnerConfig;
@@ -30,6 +33,8 @@ final readonly class AgentRunner
         private RetrieverInterface $retriever,
         private ShopSearchService $shopSearchService,
         private SearchRepairService $searchRepairService,
+        private CommerceReferenceStore $commerceReferenceStore,
+        private CommerceReferenceResolver $commerceReferenceResolver,
         private CommerceIntentLite $commerceIntentLite,
         private OllamaClient $ollamaClient,
         private LoggerInterface $agentLogger,
@@ -52,7 +57,8 @@ final readonly class AgentRunner
 
         $shopResults = [];
         $primaryShopResults = [];
-        $sources = [];
+        $factSources = [];
+        $contextSignals = [];
         $optimizedShopQuery = '';
         $shopSearchQuery = '';
         $commerceIntent = CommerceIntentLite::NONE;
@@ -60,6 +66,8 @@ final readonly class AgentRunner
         $attemptedShopRepair = false;
         $usedShopRepair = false;
         $shopRepairQueries = [];
+        $activeCommerceReference = null;
+        $shopChecked = false;
 
         $this->agentLogger->info('Agent run started', [
             'userId' => $userId,
@@ -72,19 +80,18 @@ final readonly class AgentRunner
             }
 
             yield $this->systemMsg('Ich analysiere deine Anfrage...', 'think');
-
             yield $this->systemMsg('Ich prüfe auf Internetquellen...', 'think');
 
             $urlContent = $this->urlAnalyzer->extractContentFromPrompt($prompt);
             if ($urlContent !== '') {
-                $this->addSource($sources, 'Externe URL');
+                $this->addBadge($factSources, 'Externe URL');
             }
 
             yield $this->systemMsg('Ich hole relevante Daten aus meinem RAG-Wissen...', 'think');
 
             $knowledgeChunks = $this->retriever->retrieve($prompt);
             if ($knowledgeChunks !== []) {
-                $this->addSource($sources, 'RAG Wissen');
+                $this->addBadge($factSources, 'RAG Wissen');
             }
 
             $commerceIntent = $this->detectCommerceIntent($prompt);
@@ -93,18 +100,53 @@ final readonly class AgentRunner
                 yield $this->systemMsg('Ich optimiere die Recherche...', 'think');
 
                 $commerceHistoryContext = $this->buildCommerceHistoryContext($userId);
+                $activeCommerceReference = $this->loadCommerceReference($userId);
 
                 if ($commerceHistoryContext !== '') {
-                    $this->addSource($sources, 'Chatverlauf');
+                    $this->addBadge($contextSignals, 'Gesprächskontext');
                 }
 
-                $optimizedShopQuery = $this->buildOptimizedShopQuery(
+                if ($activeCommerceReference !== null) {
+                    $this->addBadge($contextSignals, 'Commerce-Referenz');
+                }
+
+                $isReferenceOnlyFollowUp = $this->isReferenceOnlyCommerceFollowUp(
                     $prompt,
-                    $userId,
-                    $commerceHistoryContext
+                    $activeCommerceReference
                 );
 
-                $shopSearchQuery = $optimizedShopQuery !== '' ? $optimizedShopQuery : $prompt;
+                if ($isReferenceOnlyFollowUp) {
+                    $shopSearchQuery = $this->buildDeterministicReferenceShopQuery($activeCommerceReference);
+
+                    if ($shopSearchQuery !== '') {
+                        $this->addBadge($contextSignals, 'Deterministische Referenzsuche');
+                    }
+
+                    $this->agentLogger->info('Using deterministic reference shop query', [
+                        'userId' => $userId,
+                        'commerceIntent' => $commerceIntent,
+                        'prompt' => $prompt,
+                        'shopSearchQuery' => $shopSearchQuery,
+                        'referenceProductName' => $activeCommerceReference?->productName,
+                        'referenceFocusTerms' => $activeCommerceReference?->focusTerms,
+                    ]);
+                } else {
+                    $optimizedShopQuery = $this->buildOptimizedShopQuery(
+                        $prompt,
+                        $userId,
+                        $commerceHistoryContext
+                    );
+
+                    if ($optimizedShopQuery !== '' && $optimizedShopQuery !== $prompt) {
+                        $this->addBadge($contextSignals, 'Query-Optimierung');
+                    }
+
+                    $shopSearchQuery = $optimizedShopQuery !== '' ? $optimizedShopQuery : $prompt;
+                }
+
+                if ($shopSearchQuery === '') {
+                    $shopSearchQuery = $prompt;
+                }
 
                 $this->agentLogger->info('Commerce search prepared', [
                     'userId' => $userId,
@@ -112,8 +154,11 @@ final readonly class AgentRunner
                     'usedOptimizedShopQuery' => $optimizedShopQuery !== '',
                     'optimizedShopQuery' => $optimizedShopQuery,
                     'shopSearchQuery' => $shopSearchQuery,
+                    'usedDeterministicReferenceQuery' => $isReferenceOnlyFollowUp,
                     'hasCommerceHistoryContext' => $commerceHistoryContext !== '',
                     'commerceHistoryContextLength' => mb_strlen($commerceHistoryContext),
+                    'hasActiveCommerceReference' => $activeCommerceReference !== null,
+                    'activeCommerceReferenceProduct' => $activeCommerceReference?->productName,
                 ]);
 
                 yield $this->systemMsg(
@@ -121,11 +166,14 @@ final readonly class AgentRunner
                     'think'
                 );
 
+                $shopChecked = true;
+
                 $primaryShopResults = $this->searchShop(
                     $shopSearchQuery,
                     $commerceIntent,
                     $userId,
-                    $commerceHistoryContext
+                    $commerceHistoryContext,
+                    $activeCommerceReference
                 );
 
                 $repairPayload = $this->repairShopResults(
@@ -144,11 +192,13 @@ final readonly class AgentRunner
                 $shopRepairQueries = $repairPayload['repairQueries'];
 
                 if ($shopResults !== []) {
-                    $this->addSource($sources, 'Shopsystem');
+                    $this->addBadge($factSources, 'Shopsystem');
+                } elseif ($shopChecked) {
+                    $this->addBadge($factSources, 'Shopsystem geprüft');
                 }
 
                 if ($attemptedShopRepair) {
-                    $this->addSource($sources, 'Erweiterte Shopsuche');
+                    $this->addBadge($contextSignals, 'Erweiterte Shopsuche');
                 }
             }
 
@@ -176,6 +226,7 @@ final readonly class AgentRunner
                     'shopSearchQuery' => $shopSearchQuery,
                     'primaryShopResultsCount' => count($primaryShopResults),
                     'shopResultsCount' => count($shopResults),
+                    'shopChecked' => $shopChecked,
                     'attemptedShopRepair' => $attemptedShopRepair,
                     'usedShopRepair' => $usedShopRepair,
                     'shopRepairQueries' => $shopRepairQueries,
@@ -192,14 +243,22 @@ final readonly class AgentRunner
                 ]);
             }
 
-            if ($sources !== []) {
-                yield $this->emitSources($sources, 'Genutzte Quellen: ');
+            if ($factSources !== [] || $contextSignals !== []) {
+                yield $this->emitSourceSummary(
+                    $factSources,
+                    $contextSignals,
+                    'Genutzte Datenpfade'
+                );
             }
 
             $fullOutput = yield from $this->streamFinalAnswer($finalPrompt);
 
-            if ($sources !== []) {
-                yield $this->emitSources($sources, 'Quellen: ');
+            if ($factSources !== [] || $contextSignals !== []) {
+                yield $this->emitSourceSummary(
+                    $factSources,
+                    $contextSignals,
+                    'Quellen und Signale'
+                );
             }
 
             if ($this->debug) {
@@ -207,10 +266,11 @@ final readonly class AgentRunner
             }
 
             if ($fullOutput !== '') {
-                $this->contextService->appendHistory(
-                    $userId,
-                    $prompt,
-                    $fullOutput
+                $this->persistConversationState(
+                    userId: $userId,
+                    prompt: $prompt,
+                    fullOutput: $fullOutput,
+                    shopResults: $shopResults
                 );
             }
 
@@ -221,6 +281,7 @@ final readonly class AgentRunner
                 'commerceIntent' => $commerceIntent,
                 'primaryShopResultsCount' => count($primaryShopResults),
                 'shopResultsCount' => count($shopResults),
+                'shopChecked' => $shopChecked,
                 'attemptedShopRepair' => $attemptedShopRepair,
                 'usedShopRepair' => $usedShopRepair,
                 'shopRepairQueries' => $shopRepairQueries,
@@ -231,6 +292,8 @@ final readonly class AgentRunner
                 'shopSearchQuery' => $shopSearchQuery,
                 'hasCommerceHistoryContext' => $commerceHistoryContext !== '',
                 'commerceHistoryContextLength' => mb_strlen($commerceHistoryContext),
+                'hasActiveCommerceReference' => $activeCommerceReference !== null,
+                'activeCommerceReferenceProduct' => $activeCommerceReference?->productName,
             ]);
         } catch (Throwable $e) {
             $this->agentLogger->error('Agent run failed', [
@@ -298,6 +361,42 @@ final readonly class AgentRunner
         return $this->sanitizeOptimizedShopQuery($optimizedQuery);
     }
 
+    private function isReferenceOnlyCommerceFollowUp(
+        string $prompt,
+        ?CommerceReferenceContext $referenceContext
+    ): bool {
+        if ($referenceContext === null) {
+            return false;
+        }
+
+        $normalizedPrompt = mb_strtolower(trim($prompt), 'UTF-8');
+        $normalizedPrompt = preg_replace('/[^\p{L}\p{N}\s]+/u', ' ', $normalizedPrompt) ?? $normalizedPrompt;
+        $normalizedPrompt = preg_replace('/\s+/u', ' ', $normalizedPrompt) ?? $normalizedPrompt;
+        $normalizedPrompt = trim($normalizedPrompt);
+
+        if ($normalizedPrompt === '') {
+            return false;
+        }
+
+        if (preg_match('/\b(testomat|lab|evo|eco|calc|thcl|808|2000)\b/u', $normalizedPrompt) === 1) {
+            return false;
+        }
+
+        return preg_match(
+                '/\b(preis|preise|kosten|kostet|dazu|dafuer|dafür|davon|was kostet das|verfuegbarkeit|verfügbarkeit|shop|link)\b/u',
+                $normalizedPrompt
+            ) === 1;
+    }
+
+    private function buildDeterministicReferenceShopQuery(?CommerceReferenceContext $referenceContext): string
+    {
+        if ($referenceContext === null) {
+            return '';
+        }
+
+        return trim($referenceContext->buildReferenceSearchText());
+    }
+
     /**
      * @return array{
      *     results: array,
@@ -346,13 +445,15 @@ final readonly class AgentRunner
         string $query,
         string $commerceIntent,
         string $userId,
-        string $commerceHistoryContext = ''
+        string $commerceHistoryContext = '',
+        ?CommerceReferenceContext $referenceContext = null
     ): array {
         try {
             return $this->shopSearchService->search(
                 $query,
                 $commerceIntent,
-                $commerceHistoryContext
+                $commerceHistoryContext,
+                $referenceContext
             );
         } catch (Throwable $e) {
             $this->agentLogger->warning('Shop search failed, continuing without shop results', [
@@ -361,6 +462,8 @@ final readonly class AgentRunner
                 'query' => $query,
                 'hasCommerceHistoryContext' => $commerceHistoryContext !== '',
                 'commerceHistoryContextLength' => mb_strlen($commerceHistoryContext),
+                'hasReferenceContext' => $referenceContext !== null,
+                'referenceProductName' => $referenceContext?->productName,
                 'exception' => $e,
             ]);
 
@@ -376,6 +479,64 @@ final readonly class AgentRunner
         );
     }
 
+    private function loadCommerceReference(string $userId): ?CommerceReferenceContext
+    {
+        try {
+            return $this->commerceReferenceStore->load($userId);
+        } catch (Throwable $e) {
+            $this->agentLogger->warning('Failed to load commerce reference context', [
+                'userId' => $userId,
+                'exception' => $e,
+            ]);
+
+            return null;
+        }
+    }
+
+    /**
+     * @param array<int, mixed> $shopResults
+     */
+    private function storeCommerceReference(string $userId, string $prompt, string $answer, array $shopResults): void
+    {
+        try {
+            $referenceContext = $this->commerceReferenceResolver->resolveFromCommerceTurn(
+                $prompt,
+                $answer,
+                $shopResults
+            );
+
+            if ($referenceContext === null) {
+                return;
+            }
+
+            $this->commerceReferenceStore->save($userId, $referenceContext);
+        } catch (Throwable $e) {
+            $this->agentLogger->warning('Failed to persist commerce reference context', [
+                'userId' => $userId,
+                'exception' => $e,
+            ]);
+        }
+    }
+
+    /**
+     * @param array<int, mixed> $shopResults
+     */
+    private function persistConversationState(
+        string $userId,
+        string $prompt,
+        string $fullOutput,
+        array $shopResults
+    ): void {
+        $this->contextService->appendHistory($userId, $prompt, $fullOutput);
+
+        $this->storeCommerceReference(
+            userId: $userId,
+            prompt: $prompt,
+            answer: $fullOutput,
+            shopResults: $shopResults
+        );
+    }
+
     private function limitKnowledgeChunks(array $knowledgeChunks, string $commerceIntent): array
     {
         return match ($commerceIntent) {
@@ -447,22 +608,36 @@ final readonly class AgentRunner
     }
 
     /**
-     * @param string[] $sources
+     * @param string[] $factSources
+     * @param string[] $contextSignals
      */
-    private function emitSources(array $sources, string $prefix): string
+    private function emitSourceSummary(array $factSources, array $contextSignals, string $label): string
     {
-        return $this->systemMsg($prefix . implode(' ', $sources), 'info');
+        $parts = [];
+
+        if ($factSources !== []) {
+            $parts[] = 'Fakten: ' . implode(' ', $factSources);
+        }
+
+        if ($contextSignals !== []) {
+            $parts[] = 'Kontext: ' . implode(' ', $contextSignals);
+        }
+
+        return $this->systemMsg(
+            $label . ': ' . implode(' &nbsp;&nbsp; ', $parts),
+            'info'
+        );
     }
 
     /**
-     * @param string[] $sources
+     * @param string[] $target
      */
-    private function addSource(array &$sources, string $label): void
+    private function addBadge(array &$target, string $label): void
     {
         $badge = $this->badge($label);
 
-        if (!in_array($badge, $sources, true)) {
-            $sources[] = $badge;
+        if (!in_array($badge, $target, true)) {
+            $target[] = $badge;
         }
     }
 

src/Agent/PromptBuilder.php

@@ -14,7 +14,6 @@ use RuntimeException;
 
 final readonly class PromptBuilder
 {
-
     public function __construct(
         private ContextService $contextService,
         private SystemPromptRepository $systemPromptRepository,
@@ -49,12 +48,24 @@ final readonly class PromptBuilder
 
         $hasShopResults = $shopResults !== [];
         $isTechnicalProductQuestion = $this->isLikelyTechnicalProductQuestion($prompt);
+        $isPriceDrivenQuestion = $this->isLikelyPriceDrivenQuestion($prompt);
 
         $systemBlock = $this->buildSystemBlock();
         $shopBlock = $this->buildShopBlock($shopResults, $swagFullOutPut);
-        $outputPriorityBlock = $this->buildOutputPriorityBlock($hasShopResults);
-        $responseFormatBlock = $this->buildResponseFormatBlock($prompt, $hasShopResults, $isTechnicalProductQuestion);
-        $knowledgeBlock = $this->buildKnowledgeBlock($knowledgeChunks, $urlContent, $prompt, $hasShopResults);
+        $outputPriorityBlock = $this->buildOutputPriorityBlock($hasShopResults, $isPriceDrivenQuestion);
+        $responseFormatBlock = $this->buildResponseFormatBlock(
+            $prompt,
+            $hasShopResults,
+            $isTechnicalProductQuestion,
+            $isPriceDrivenQuestion
+        );
+        $knowledgeBlock = $this->buildKnowledgeBlock(
+            $knowledgeChunks,
+            $urlContent,
+            $prompt,
+            $hasShopResults,
+            $isPriceDrivenQuestion
+        );
         $userBlock = $this->buildUserBlock($prompt);
 
         $fixedPrompt = $this->implodeBlocks([
@@ -231,7 +242,9 @@ final readonly class PromptBuilder
                 "Output real URL values exactly as provided in the shop results. Do not replace them with placeholders, link labels, or product names.\n" .
                 "Do not infer undocumented technical specifications from shop data.\n" .
                 "Commercial fields from shop data may only be assigned to a product if the shop item clearly matches the same product identity.\n" .
-                "Do not merge a device identified in retrieved knowledge with price, URL, product number, or availability from a different shop item such as a reagent, accessory, kit, consumable, or service item.";
+                "Do not merge a device identified in retrieved knowledge with price, URL, product number, or availability from a different shop item such as a reagent, accessory, kit, consumable, or service item.\n" .
+                "If shop results only contain accessories, reagents, indicators, or consumables, do not conclude that no matching main device exists unless the sources explicitly support that conclusion.\n" .
+                "If the user asks for price filtering, use the numeric prices in these live shop results as the decisive source for filtering.";
 
             if ($totalCount > count($limitedShopResults)) {
                 $header .= "\n" .
@@ -247,12 +260,20 @@ final readonly class PromptBuilder
     /**
      * Build a small priority block that tells the model what to surface first.
      */
-    private function buildOutputPriorityBlock(bool $hasShopResults): string
+    private function buildOutputPriorityBlock(bool $hasShopResults, bool $isPriceDrivenQuestion): string
     {
         if (!$hasShopResults) {
             return '';
         }
 
+        if ($isPriceDrivenQuestion) {
+            return
+                "OUTPUT PRIORITY:\n" .
+                "For price-driven questions, evaluate shop results first for numeric price filtering.\n" .
+                "Use retrieved knowledge afterwards only to add technical context or explain missing commercial coverage.\n" .
+                "Do not let accessory-only shop results prove that no matching device exists unless the sources explicitly support that conclusion.\n";
+        }
+
         return
             "OUTPUT PRIORITY:\n" .
             "Use retrieved knowledge first to determine the technically matching product or answer.\n" .
@@ -263,7 +284,8 @@ final readonly class PromptBuilder
     private function buildResponseFormatBlock(
         string $prompt,
         bool $hasShopResults,
-        bool $isTechnicalProductQuestion
+        bool $isTechnicalProductQuestion,
+        bool $isPriceDrivenQuestion
     ): string {
         $rules = [
             "RESPONSE FORMAT RULES:",
@@ -284,6 +306,8 @@ final readonly class PromptBuilder
             $rules[] = "- Only use shop price, URL, product number, or availability for the main product when the shop result clearly matches that same main product.";
             $rules[] = "- If the matching shop item appears to be an accessory, reagent, consumable, set, or kit, keep it separate and do not present its commercial fields as the main device.";
             $rules[] = "- If the commercial match is uncertain, say that commercial details for the main product are not clearly available in the provided shop results.";
+            $rules[] = "- If the question includes a price threshold, filter using only explicit numeric shop prices.";
+            $rules[] = "- Do not say that no device exists above a threshold merely because only cheaper accessories were found in the shop results.";
         } else {
             $rules[] = "- If no shop results are present, do not compensate by inventing external products or external manufacturers.";
         }
@@ -294,6 +318,12 @@ final readonly class PromptBuilder
             $rules[] = "- If the sources only support a negative finding, output only that negative finding and do not add speculative alternatives.";
         }
 
+        if ($isPriceDrivenQuestion) {
+            $rules[] = "- For price-driven questions, answer the threshold result first.";
+            $rules[] = "- If no grounded shop product fulfills the threshold, say that clearly.";
+            $rules[] = "- Then optionally explain whether retrieved knowledge mentions relevant devices that are not commercially listed in the current shop results.";
+        }
+
         if ($this->asksForAccessoryOrBundle($prompt)) {
             $rules[] = "- If the user asks for a matching accessory, separate the answer into: main device and matching accessory.";
             $rules[] = "- The main device must come first. The accessory must not replace the main device.";
@@ -304,14 +334,13 @@ final readonly class PromptBuilder
         return implode("\n", $rules);
     }
 
-    /**
-     * Build the knowledge block.
-     *
-     * Retrieved knowledge remains the main source for technical matching and explanation.
-     * Shop data is preferred for current commercial fields.
-     */
-    private function buildKnowledgeBlock(array $knowledgeChunks, string $urlContent, string $prompt, bool $hasShopResults): string
-    {
+    private function buildKnowledgeBlock(
+        array $knowledgeChunks,
+        string $urlContent,
+        string $prompt,
+        bool $hasShopResults,
+        bool $isPriceDrivenQuestion
+    ): string {
         $knowledgeParts = [];
         $isTechnicalProductQuestion = $this->isLikelyTechnicalProductQuestion($prompt);
 
@@ -334,7 +363,7 @@ final readonly class PromptBuilder
                     "LANGUAGE RULES:\n" .
                     implode("\n", $this->buildLanguageRules()),
                     "FACT GROUNDING RULES:\n" .
-                    implode("\n", $this->buildFactGroundingRules($isTechnicalProductQuestion, $hasShopResults)),
+                    implode("\n", $this->buildFactGroundingRules($isTechnicalProductQuestion, $hasShopResults, $isPriceDrivenQuestion)),
                     "RETRIEVED KNOWLEDGE (primary for technical matching and factual explanation):\n" .
                     "Source: Documents\n" .
                     implode("\n\n", $lines),
@@ -354,13 +383,6 @@ final readonly class PromptBuilder
         return $this->implodeBlocks($knowledgeParts);
     }
 
-    /**
-     * Resolve how many characters may still be used by history.
-     *
-     * The active model num_ctx is converted into a conservative prompt budget.
-     * Shop, knowledge and user question are fixed priority blocks.
-     * History only receives the remaining space.
-     */
     private function resolveHistoryBudgetChars(string $fixedPrompt): int
     {
         $numCtx = $this->modelGenerationConfigProvider->getActiveNumCtx();
@@ -407,8 +429,11 @@ final readonly class PromptBuilder
     /**
      * @return string[]
      */
-    private function buildFactGroundingRules(bool $isTechnicalProductQuestion, bool $hasShopResults): array
-    {
+    private function buildFactGroundingRules(
+        bool $isTechnicalProductQuestion,
+        bool $hasShopResults,
+        bool $isPriceDrivenQuestion
+    ): array {
         $rules = [
             "- State only facts that are explicitly present in the provided sources.",
             "- Extract concrete values exactly when they are present, including units, ranges, model names, indicator names, IP classes, temperatures, pressures, dimensions, counts, relay outputs, current outputs, and error codes.",
@@ -437,6 +462,11 @@ final readonly class PromptBuilder
                 "- Only use commercial fields for the main product when the shop item and the technically identified product clearly refer to the same product identity.",
                 "- If the shop match is ambiguous, keep the technical identification and commercial details separate.",
             ]);
+
+            if ($isPriceDrivenQuestion) {
+                $rules[] = "- For price-threshold questions, shop prices are authoritative for the threshold check.";
+                $rules[] = "- Accessory-only shop hits do not prove that no qualifying device exists.";
+            }
         } else {
             $rules[] = "- Use retrieved knowledge as authoritative for factual answers.";
             $rules[] = "- If no shop results are present, do not compensate with external recommendations or external product suggestions.";
@@ -523,6 +553,20 @@ final readonly class PromptBuilder
         return preg_match('/\b[\p{L}]{2,}\s?\d{2,5}\b/u', $prompt) === 1;
     }
 
+    private function isLikelyPriceDrivenQuestion(string $prompt): bool
+    {
+        $normalized = mb_strtolower($prompt, 'UTF-8');
+
+        if (preg_match('/\b(mehr\s+als|über|ueber|größer\s+als|groesser\s+als|unter|bis|ab|mindestens|min)\s+\d+(?:[.,]\d+)?\s*(?:euro|eur|€)\b/u', $normalized) === 1) {
+            return true;
+        }
+
+        return str_contains($normalized, 'preis')
+            || str_contains($normalized, 'preise')
+            || str_contains($normalized, 'kosten')
+            || str_contains($normalized, 'kostet');
+    }
+
     private function asksForAccessoryOrBundle(string $prompt): bool
     {
         $normalized = mb_strtolower($prompt, 'UTF-8');