@retrivora-ai/rag-engine 1.0.5 → 1.0.6

package/dist/{RagConfig-BEmz4hVP.d.mts → RagConfig-BjC6zSTV.d.mts}

@@ -244,6 +244,10 @@ interface RAGConfig {
     useGraphRetrieval?: boolean;
     /** Whether to perform reranking on retrieved results */
     useReranking?: boolean;
+    /** List of metadata fields that are valid for filtering.
+     * Used to dynamically extract hints from natural language queries.
+     */
+    filterableFields?: string[];
 }
 interface RagConfig {
     /**

package/dist/{RagConfig-BEmz4hVP.d.ts → RagConfig-BjC6zSTV.d.ts}

@@ -244,6 +244,10 @@ interface RAGConfig {
     useGraphRetrieval?: boolean;
     /** Whether to perform reranking on retrieved results */
     useReranking?: boolean;
+    /** List of metadata fields that are valid for filtering.
+     * Used to dynamically extract hints from natural language queries.
+     */
+    filterableFields?: string[];
 }
 interface RagConfig {
     /**

package/dist/{chunk-MWL4AGQI.mjs → chunk-ZCDJSGUW.mjs}

@@ -1904,15 +1904,18 @@ var QueryProcessor = class {
    * Checks if a string is likely a question word or common prompt phrase.
    */
   static isLikelyPromptPhrase(value) {
-    return /^(what|which|who|where|when|why|how)\b/i.test(value.trim());
+    return /^(what|which|who|where|when|why|how|the|is|are|was|were|in|on|at|by|for|with|about|a|an|to|of)\b/i.test(value.trim());
   }
   /**
    * Scans a natural language question for potential metadata hints and keywords.
+   * @param question The user's query
+   * @param validFields Optional list of known filterable fields to look for
    */
-  static extractQueryFieldHints(question) {
+  static extractQueryFieldHints(question, validFields = []) {
     var _a, _b, _c, _d;
     if (!question.trim()) return [];
     const hints = /* @__PURE__ */ new Map();
+    const fieldsToSearch = [.../* @__PURE__ */ new Set([...this.COMMON_METADATA_FIELDS, ...validFields])];
     const addHint = (value, field) => {
       const normalizedValue = this.normalizeHintValue(value);
       if (!normalizedValue) return;
@@ -1943,11 +1946,28 @@ var QueryProcessor = class {
         if (name) addHint(name, "name");
       }
     }
+    if (fieldsToSearch.length > 0) {
+      for (const field of fieldsToSearch) {
+        const forwardPattern = new RegExp(`\\b${field}\\s*(?:is|are|:|of)?\\s+["']?([^"\\n?.!,]{2,60})["']?(?=[?.!,]|$)`, "gi");
+        for (const match of question.matchAll(forwardPattern)) {
+          const value = match[1];
+          if (value && !this.isLikelyPromptPhrase(value)) {
+            addHint(value, field);
+          }
+        }
+        const reversePattern = new RegExp(`\\b([^"\\n?.!,\\s]{2,60})\\s+${field}\\b`, "gi");
+        for (const match of question.matchAll(reversePattern)) {
+          const value = match[1];
+          if (value && !this.isLikelyPromptPhrase(value)) {
+            addHint(value, field);
+          }
+        }
+      }
+    }
     const universalPatterns = [
       { regex: /([a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,})/gi, field: "email", group: 1 },
       { regex: /(\+?\d[\d\-\.\s\(\)]{6,}\d)/g, field: "phone", group: 1 },
       { regex: /(\b\d{4}-\d{2}-\d{2}\b|\b\d{1,2}\/\d{1,2}\/\d{2,4}\b)/g, field: "date", group: 1 },
-      { regex: /(\$\s?\d{1,3}(?:,\d{3})*(?:\.\d+)?)/g, field: "amount", group: 1 },
       { regex: /\b(ID|id|identifier)[: ]\s*([A-Za-z0-9\-]{3,})\b/gi, field: "id", group: 2 },
       { regex: /"([^"]{2,120})"/g, group: 1 },
       { regex: /'([^']{2,120})'/g, group: 1 }
@@ -2012,6 +2032,28 @@ var QueryProcessor = class {
     return filter;
   }
 };
+QueryProcessor.COMMON_METADATA_FIELDS = [
+  "category",
+  "type",
+  "brand",
+  "status",
+  "priority",
+  "id",
+  "name",
+  "email",
+  "phone",
+  "price",
+  "rating",
+  "color",
+  "size",
+  "material",
+  "sku",
+  "role",
+  "department",
+  "location",
+  "tag",
+  "label"
+];
 
 // src/core/Pipeline.ts
 var LRUEmbeddingCache = class {
@@ -2241,7 +2283,7 @@ var Pipeline = class {
    */
   askStream(_0) {
     return __asyncGenerator(this, arguments, function* (question, history = [], namespace) {
-      var _a, _b, _c, _d, _e, _f;
+      var _a, _b, _c, _d, _e, _f, _g;
       yield new __await(this.initialize());
       const ns = namespace != null ? namespace : this.config.projectId;
       const topK = (_b = (_a = this.config.rag) == null ? void 0 : _a.topK) != null ? _b : 5;
@@ -2251,7 +2293,7 @@ var Pipeline = class {
         if ((_e = this.config.rag) == null ? void 0 : _e.useQueryTransformation) {
           searchQuery = yield new __await(this.rewriteQuery(question, history));
         }
-        const hints = QueryProcessor.extractQueryFieldHints(question);
+        const hints = QueryProcessor.extractQueryFieldHints(question, (_f = this.config.rag) == null ? void 0 : _f.filterableFields);
         const filter = QueryProcessor.buildQueryFilter(question, hints);
         const { sources: rawSources, graphData } = yield new __await(this.retrieve(searchQuery, {
           namespace: ns,
@@ -2259,7 +2301,7 @@ var Pipeline = class {
           filter
         }));
         let sources = rawSources.filter((m) => m.score >= scoreThreshold);
-        if ((_f = this.config.rag) == null ? void 0 : _f.useReranking) {
+        if ((_g = this.config.rag) == null ? void 0 : _g.useReranking) {
           sources = yield new __await(this.reranker.rerank(sources, question, topK));
         } else {
           sources = sources.slice(0, topK);

package/dist/handlers/index.d.mts

@@ -1,3 +1,3 @@
-import '../RagConfig-BEmz4hVP.mjs';
-export { c as createChatHandler, d as createHealthHandler, e as createIngestHandler, f as createStreamHandler, g as createUploadHandler, s as sseErrorFrame, h as sseFrame, i as sseMetaFrame, j as sseTextFrame } from '../index-DFSy0CG6.mjs';
+import '../RagConfig-BjC6zSTV.mjs';
+export { c as createChatHandler, d as createHealthHandler, e as createIngestHandler, f as createStreamHandler, g as createUploadHandler, s as sseErrorFrame, h as sseFrame, i as sseMetaFrame, j as sseTextFrame } from '../index-CU_fQq__.mjs';
 import 'next/server';

package/dist/handlers/index.d.ts

@@ -1,3 +1,3 @@
-import '../RagConfig-BEmz4hVP.js';
-export { c as createChatHandler, d as createHealthHandler, e as createIngestHandler, f as createStreamHandler, g as createUploadHandler, s as sseErrorFrame, h as sseFrame, i as sseMetaFrame, j as sseTextFrame } from '../index-CSfaCeux.js';
+import '../RagConfig-BjC6zSTV.js';
+export { c as createChatHandler, d as createHealthHandler, e as createIngestHandler, f as createStreamHandler, g as createUploadHandler, s as sseErrorFrame, h as sseFrame, i as sseMetaFrame, j as sseTextFrame } from '../index-C3bLmWcR.js';
 import 'next/server';

package/dist/handlers/index.js

@@ -3362,15 +3362,18 @@ var QueryProcessor = class {
    * Checks if a string is likely a question word or common prompt phrase.
    */
   static isLikelyPromptPhrase(value) {
-    return /^(what|which|who|where|when|why|how)\b/i.test(value.trim());
+    return /^(what|which|who|where|when|why|how|the|is|are|was|were|in|on|at|by|for|with|about|a|an|to|of)\b/i.test(value.trim());
   }
   /**
    * Scans a natural language question for potential metadata hints and keywords.
+   * @param question The user's query
+   * @param validFields Optional list of known filterable fields to look for
    */
-  static extractQueryFieldHints(question) {
+  static extractQueryFieldHints(question, validFields = []) {
     var _a, _b, _c, _d;
     if (!question.trim()) return [];
     const hints = /* @__PURE__ */ new Map();
+    const fieldsToSearch = [.../* @__PURE__ */ new Set([...this.COMMON_METADATA_FIELDS, ...validFields])];
     const addHint = (value, field) => {
       const normalizedValue = this.normalizeHintValue(value);
       if (!normalizedValue) return;
@@ -3401,11 +3404,28 @@ var QueryProcessor = class {
         if (name) addHint(name, "name");
       }
     }
+    if (fieldsToSearch.length > 0) {
+      for (const field of fieldsToSearch) {
+        const forwardPattern = new RegExp(`\\b${field}\\s*(?:is|are|:|of)?\\s+["']?([^"\\n?.!,]{2,60})["']?(?=[?.!,]|$)`, "gi");
+        for (const match of question.matchAll(forwardPattern)) {
+          const value = match[1];
+          if (value && !this.isLikelyPromptPhrase(value)) {
+            addHint(value, field);
+          }
+        }
+        const reversePattern = new RegExp(`\\b([^"\\n?.!,\\s]{2,60})\\s+${field}\\b`, "gi");
+        for (const match of question.matchAll(reversePattern)) {
+          const value = match[1];
+          if (value && !this.isLikelyPromptPhrase(value)) {
+            addHint(value, field);
+          }
+        }
+      }
+    }
     const universalPatterns = [
       { regex: /([a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,})/gi, field: "email", group: 1 },
       { regex: /(\+?\d[\d\-\.\s\(\)]{6,}\d)/g, field: "phone", group: 1 },
       { regex: /(\b\d{4}-\d{2}-\d{2}\b|\b\d{1,2}\/\d{1,2}\/\d{2,4}\b)/g, field: "date", group: 1 },
-      { regex: /(\$\s?\d{1,3}(?:,\d{3})*(?:\.\d+)?)/g, field: "amount", group: 1 },
       { regex: /\b(ID|id|identifier)[: ]\s*([A-Za-z0-9\-]{3,})\b/gi, field: "id", group: 2 },
       { regex: /"([^"]{2,120})"/g, group: 1 },
       { regex: /'([^']{2,120})'/g, group: 1 }
@@ -3470,6 +3490,28 @@ var QueryProcessor = class {
     return filter;
   }
 };
+QueryProcessor.COMMON_METADATA_FIELDS = [
+  "category",
+  "type",
+  "brand",
+  "status",
+  "priority",
+  "id",
+  "name",
+  "email",
+  "phone",
+  "price",
+  "rating",
+  "color",
+  "size",
+  "material",
+  "sku",
+  "role",
+  "department",
+  "location",
+  "tag",
+  "label"
+];
 
 // src/core/Pipeline.ts
 var LRUEmbeddingCache = class {
@@ -3699,7 +3741,7 @@ var Pipeline = class {
    */
   askStream(_0) {
     return __asyncGenerator(this, arguments, function* (question, history = [], namespace) {
-      var _a, _b, _c, _d, _e, _f;
+      var _a, _b, _c, _d, _e, _f, _g;
       yield new __await(this.initialize());
       const ns = namespace != null ? namespace : this.config.projectId;
       const topK = (_b = (_a = this.config.rag) == null ? void 0 : _a.topK) != null ? _b : 5;
@@ -3709,7 +3751,7 @@ var Pipeline = class {
         if ((_e = this.config.rag) == null ? void 0 : _e.useQueryTransformation) {
           searchQuery = yield new __await(this.rewriteQuery(question, history));
         }
-        const hints = QueryProcessor.extractQueryFieldHints(question);
+        const hints = QueryProcessor.extractQueryFieldHints(question, (_f = this.config.rag) == null ? void 0 : _f.filterableFields);
         const filter = QueryProcessor.buildQueryFilter(question, hints);
         const { sources: rawSources, graphData } = yield new __await(this.retrieve(searchQuery, {
           namespace: ns,
@@ -3717,7 +3759,7 @@ var Pipeline = class {
           filter
         }));
         let sources = rawSources.filter((m) => m.score >= scoreThreshold);
-        if ((_f = this.config.rag) == null ? void 0 : _f.useReranking) {
+        if ((_g = this.config.rag) == null ? void 0 : _g.useReranking) {
           sources = yield new __await(this.reranker.rerank(sources, question, topK));
         } else {
           sources = sources.slice(0, topK);

package/dist/handlers/index.mjs

@@ -8,7 +8,7 @@ import {
   sseFrame,
   sseMetaFrame,
   sseTextFrame
-} from "../chunk-MWL4AGQI.mjs";
+} from "../chunk-ZCDJSGUW.mjs";
 import "../chunk-YLTMFW4M.mjs";
 import "../chunk-X4TOT24V.mjs";
 export {

package/dist/{index-CSfaCeux.d.ts → index-C3bLmWcR.d.ts}

@@ -1,4 +1,4 @@
-import { g as RagConfig, C as ChatMessage, b as ChatResponse, e as IngestDocument } from './RagConfig-BEmz4hVP.js';
+import { g as RagConfig, C as ChatMessage, b as ChatResponse, e as IngestDocument } from './RagConfig-BjC6zSTV.js';
 import { NextRequest, NextResponse } from 'next/server';
 
 interface ValidationError {

package/dist/{index-DFSy0CG6.d.mts → index-CU_fQq__.d.mts}

@@ -1,4 +1,4 @@
-import { g as RagConfig, C as ChatMessage, b as ChatResponse, e as IngestDocument } from './RagConfig-BEmz4hVP.mjs';
+import { g as RagConfig, C as ChatMessage, b as ChatResponse, e as IngestDocument } from './RagConfig-BjC6zSTV.mjs';
 import { NextRequest, NextResponse } from 'next/server';
 
 interface ValidationError {

package/dist/index.d.mts

@@ -1,6 +1,6 @@
 import React$1, { ReactNode } from 'react';
-import { C as ChatMessage, V as VectorMatch, U as UIConfig } from './RagConfig-BEmz4hVP.mjs';
-export { a as ChatOptions, b as ChatResponse, E as EmbedOptions, c as EmbeddingConfig, d as EmbeddingProvider, I as ILLMProvider, e as IngestDocument, L as LLMConfig, f as LLMProvider, R as RAGConfig, g as RagConfig, h as UpsertDocument, i as VectorDBConfig, j as VectorDBProvider } from './RagConfig-BEmz4hVP.mjs';
+import { C as ChatMessage, V as VectorMatch, U as UIConfig } from './RagConfig-BjC6zSTV.mjs';
+export { a as ChatOptions, b as ChatResponse, E as EmbedOptions, c as EmbeddingConfig, d as EmbeddingProvider, I as ILLMProvider, e as IngestDocument, L as LLMConfig, f as LLMProvider, R as RAGConfig, g as RagConfig, h as UpsertDocument, i as VectorDBConfig, j as VectorDBProvider } from './RagConfig-BjC6zSTV.mjs';
 export { C as Chunk, a as ChunkOptions } from './DocumentChunker-Dh9TvmGG.mjs';
 
 interface ChatWidgetProps {

package/dist/index.d.ts

@@ -1,6 +1,6 @@
 import React$1, { ReactNode } from 'react';
-import { C as ChatMessage, V as VectorMatch, U as UIConfig } from './RagConfig-BEmz4hVP.js';
-export { a as ChatOptions, b as ChatResponse, E as EmbedOptions, c as EmbeddingConfig, d as EmbeddingProvider, I as ILLMProvider, e as IngestDocument, L as LLMConfig, f as LLMProvider, R as RAGConfig, g as RagConfig, h as UpsertDocument, i as VectorDBConfig, j as VectorDBProvider } from './RagConfig-BEmz4hVP.js';
+import { C as ChatMessage, V as VectorMatch, U as UIConfig } from './RagConfig-BjC6zSTV.js';
+export { a as ChatOptions, b as ChatResponse, E as EmbedOptions, c as EmbeddingConfig, d as EmbeddingProvider, I as ILLMProvider, e as IngestDocument, L as LLMConfig, f as LLMProvider, R as RAGConfig, g as RagConfig, h as UpsertDocument, i as VectorDBConfig, j as VectorDBProvider } from './RagConfig-BjC6zSTV.js';
 export { C as Chunk, a as ChunkOptions } from './DocumentChunker-Dh9TvmGG.js';
 
 interface ChatWidgetProps {

package/dist/server.d.mts

@@ -1,7 +1,7 @@
-import { i as VectorDBConfig, L as LLMConfig, c as EmbeddingConfig, g as RagConfig, e as IngestDocument, C as ChatMessage, b as ChatResponse, k as RetrievalResult, h as UpsertDocument, V as VectorMatch, G as GraphDBConfig, l as GraphNode, m as Edge, n as GraphSearchResult, I as ILLMProvider, j as VectorDBProvider, f as LLMProvider, d as EmbeddingProvider, R as RAGConfig, U as UIConfig, a as ChatOptions, E as EmbedOptions } from './RagConfig-BEmz4hVP.mjs';
+import { i as VectorDBConfig, L as LLMConfig, c as EmbeddingConfig, g as RagConfig, e as IngestDocument, C as ChatMessage, b as ChatResponse, k as RetrievalResult, h as UpsertDocument, V as VectorMatch, G as GraphDBConfig, l as GraphNode, m as Edge, n as GraphSearchResult, I as ILLMProvider, j as VectorDBProvider, f as LLMProvider, d as EmbeddingProvider, R as RAGConfig, U as UIConfig, a as ChatOptions, E as EmbedOptions } from './RagConfig-BjC6zSTV.mjs';
 export { C as Chunk, a as ChunkOptions, D as DocumentChunker } from './DocumentChunker-Dh9TvmGG.mjs';
-import { H as HealthCheckResult, I as IProviderValidator, a as IProviderHealthChecker } from './index-DFSy0CG6.mjs';
-export { C as ConfigValidator, V as ValidationError, b as VectorPlugin, c as createChatHandler, d as createHealthHandler, e as createIngestHandler, f as createStreamHandler, g as createUploadHandler, s as sseErrorFrame, h as sseFrame, i as sseMetaFrame, j as sseTextFrame } from './index-DFSy0CG6.mjs';
+import { H as HealthCheckResult, I as IProviderValidator, a as IProviderHealthChecker } from './index-CU_fQq__.mjs';
+export { C as ConfigValidator, V as ValidationError, b as VectorPlugin, c as createChatHandler, d as createHealthHandler, e as createIngestHandler, f as createStreamHandler, g as createUploadHandler, s as sseErrorFrame, h as sseFrame, i as sseMetaFrame, j as sseTextFrame } from './index-CU_fQq__.mjs';
 import 'next/server';
 
 /**

package/dist/server.d.ts

@@ -1,7 +1,7 @@
-import { i as VectorDBConfig, L as LLMConfig, c as EmbeddingConfig, g as RagConfig, e as IngestDocument, C as ChatMessage, b as ChatResponse, k as RetrievalResult, h as UpsertDocument, V as VectorMatch, G as GraphDBConfig, l as GraphNode, m as Edge, n as GraphSearchResult, I as ILLMProvider, j as VectorDBProvider, f as LLMProvider, d as EmbeddingProvider, R as RAGConfig, U as UIConfig, a as ChatOptions, E as EmbedOptions } from './RagConfig-BEmz4hVP.js';
+import { i as VectorDBConfig, L as LLMConfig, c as EmbeddingConfig, g as RagConfig, e as IngestDocument, C as ChatMessage, b as ChatResponse, k as RetrievalResult, h as UpsertDocument, V as VectorMatch, G as GraphDBConfig, l as GraphNode, m as Edge, n as GraphSearchResult, I as ILLMProvider, j as VectorDBProvider, f as LLMProvider, d as EmbeddingProvider, R as RAGConfig, U as UIConfig, a as ChatOptions, E as EmbedOptions } from './RagConfig-BjC6zSTV.js';
 export { C as Chunk, a as ChunkOptions, D as DocumentChunker } from './DocumentChunker-Dh9TvmGG.js';
-import { H as HealthCheckResult, I as IProviderValidator, a as IProviderHealthChecker } from './index-CSfaCeux.js';
-export { C as ConfigValidator, V as ValidationError, b as VectorPlugin, c as createChatHandler, d as createHealthHandler, e as createIngestHandler, f as createStreamHandler, g as createUploadHandler, s as sseErrorFrame, h as sseFrame, i as sseMetaFrame, j as sseTextFrame } from './index-CSfaCeux.js';
+import { H as HealthCheckResult, I as IProviderValidator, a as IProviderHealthChecker } from './index-C3bLmWcR.js';
+export { C as ConfigValidator, V as ValidationError, b as VectorPlugin, c as createChatHandler, d as createHealthHandler, e as createIngestHandler, f as createStreamHandler, g as createUploadHandler, s as sseErrorFrame, h as sseFrame, i as sseMetaFrame, j as sseTextFrame } from './index-C3bLmWcR.js';
 import 'next/server';
 
 /**

package/dist/server.js

@@ -3455,15 +3455,18 @@ var QueryProcessor = class {
    * Checks if a string is likely a question word or common prompt phrase.
    */
   static isLikelyPromptPhrase(value) {
-    return /^(what|which|who|where|when|why|how)\b/i.test(value.trim());
+    return /^(what|which|who|where|when|why|how|the|is|are|was|were|in|on|at|by|for|with|about|a|an|to|of)\b/i.test(value.trim());
   }
   /**
    * Scans a natural language question for potential metadata hints and keywords.
+   * @param question The user's query
+   * @param validFields Optional list of known filterable fields to look for
    */
-  static extractQueryFieldHints(question) {
+  static extractQueryFieldHints(question, validFields = []) {
     var _a, _b, _c, _d;
     if (!question.trim()) return [];
     const hints = /* @__PURE__ */ new Map();
+    const fieldsToSearch = [.../* @__PURE__ */ new Set([...this.COMMON_METADATA_FIELDS, ...validFields])];
     const addHint = (value, field) => {
       const normalizedValue = this.normalizeHintValue(value);
       if (!normalizedValue) return;
@@ -3494,11 +3497,28 @@ var QueryProcessor = class {
         if (name) addHint(name, "name");
       }
     }
+    if (fieldsToSearch.length > 0) {
+      for (const field of fieldsToSearch) {
+        const forwardPattern = new RegExp(`\\b${field}\\s*(?:is|are|:|of)?\\s+["']?([^"\\n?.!,]{2,60})["']?(?=[?.!,]|$)`, "gi");
+        for (const match of question.matchAll(forwardPattern)) {
+          const value = match[1];
+          if (value && !this.isLikelyPromptPhrase(value)) {
+            addHint(value, field);
+          }
+        }
+        const reversePattern = new RegExp(`\\b([^"\\n?.!,\\s]{2,60})\\s+${field}\\b`, "gi");
+        for (const match of question.matchAll(reversePattern)) {
+          const value = match[1];
+          if (value && !this.isLikelyPromptPhrase(value)) {
+            addHint(value, field);
+          }
+        }
+      }
+    }
     const universalPatterns = [
       { regex: /([a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,})/gi, field: "email", group: 1 },
       { regex: /(\+?\d[\d\-\.\s\(\)]{6,}\d)/g, field: "phone", group: 1 },
       { regex: /(\b\d{4}-\d{2}-\d{2}\b|\b\d{1,2}\/\d{1,2}\/\d{2,4}\b)/g, field: "date", group: 1 },
-      { regex: /(\$\s?\d{1,3}(?:,\d{3})*(?:\.\d+)?)/g, field: "amount", group: 1 },
       { regex: /\b(ID|id|identifier)[: ]\s*([A-Za-z0-9\-]{3,})\b/gi, field: "id", group: 2 },
       { regex: /"([^"]{2,120})"/g, group: 1 },
       { regex: /'([^']{2,120})'/g, group: 1 }
@@ -3563,6 +3583,28 @@ var QueryProcessor = class {
     return filter;
   }
 };
+QueryProcessor.COMMON_METADATA_FIELDS = [
+  "category",
+  "type",
+  "brand",
+  "status",
+  "priority",
+  "id",
+  "name",
+  "email",
+  "phone",
+  "price",
+  "rating",
+  "color",
+  "size",
+  "material",
+  "sku",
+  "role",
+  "department",
+  "location",
+  "tag",
+  "label"
+];
 
 // src/core/Pipeline.ts
 var LRUEmbeddingCache = class {
@@ -3792,7 +3834,7 @@ var Pipeline = class {
    */
   askStream(_0) {
     return __asyncGenerator(this, arguments, function* (question, history = [], namespace) {
-      var _a, _b, _c, _d, _e, _f;
+      var _a, _b, _c, _d, _e, _f, _g;
       yield new __await(this.initialize());
       const ns = namespace != null ? namespace : this.config.projectId;
       const topK = (_b = (_a = this.config.rag) == null ? void 0 : _a.topK) != null ? _b : 5;
@@ -3802,7 +3844,7 @@ var Pipeline = class {
         if ((_e = this.config.rag) == null ? void 0 : _e.useQueryTransformation) {
           searchQuery = yield new __await(this.rewriteQuery(question, history));
         }
-        const hints = QueryProcessor.extractQueryFieldHints(question);
+        const hints = QueryProcessor.extractQueryFieldHints(question, (_f = this.config.rag) == null ? void 0 : _f.filterableFields);
         const filter = QueryProcessor.buildQueryFilter(question, hints);
         const { sources: rawSources, graphData } = yield new __await(this.retrieve(searchQuery, {
           namespace: ns,
@@ -3810,7 +3852,7 @@ var Pipeline = class {
           filter
         }));
         let sources = rawSources.filter((m) => m.score >= scoreThreshold);
-        if ((_f = this.config.rag) == null ? void 0 : _f.useReranking) {
+        if ((_g = this.config.rag) == null ? void 0 : _g.useReranking) {
           sources = yield new __await(this.reranker.rerank(sources, question, topK));
         } else {
           sources = sources.slice(0, topK);

package/dist/server.mjs

@@ -40,7 +40,7 @@ import {
   sseFrame,
   sseMetaFrame,
   sseTextFrame
-} from "./chunk-MWL4AGQI.mjs";
+} from "./chunk-ZCDJSGUW.mjs";
 import "./chunk-YLTMFW4M.mjs";
 import {
   PineconeProvider

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@retrivora-ai/rag-engine",
-  "version": "1.0.5",
+  "version": "1.0.6",
   "description": "Retrivora AI is a plug-and-play AI engine for RAG chat experiences — generic vector DB + LLM provider, embeddable or standalone.",
   "author": "Abhinav Alkuchi",
   "license": "MIT",

package/src/config/RagConfig.ts

@@ -174,6 +174,10 @@ export interface RAGConfig {
   useGraphRetrieval?: boolean;
   /** Whether to perform reranking on retrieved results */
   useReranking?: boolean;
+  /** List of metadata fields that are valid for filtering. 
+   * Used to dynamically extract hints from natural language queries. 
+   */
+  filterableFields?: string[];
 }
 
 // ---------------------------------------------------------------------------

package/src/core/Pipeline.ts

@@ -315,7 +315,7 @@ export class Pipeline {
       }
 
       // 2. Parallel Retrieval
-      const hints = QueryProcessor.extractQueryFieldHints(question);
+      const hints = QueryProcessor.extractQueryFieldHints(question, this.config.rag?.filterableFields);
       const filter = QueryProcessor.buildQueryFilter(question, hints);
 
       const { sources: rawSources, graphData } = await this.retrieve(searchQuery, { 

package/src/core/QueryProcessor.ts

@@ -31,16 +31,25 @@ export class QueryProcessor {
    * Checks if a string is likely a question word or common prompt phrase.
    */
   static isLikelyPromptPhrase(value: string): boolean {
-    return /^(what|which|who|where|when|why|how)\b/i.test(value.trim());
+    return /^(what|which|who|where|when|why|how|the|is|are|was|were|in|on|at|by|for|with|about|a|an|to|of)\b/i.test(value.trim());
   }
 
+  private static readonly COMMON_METADATA_FIELDS = [
+    'category', 'type', 'brand', 'status', 'priority', 'id', 'name', 
+    'email', 'phone', 'price', 'rating', 'color', 'size', 'material',
+    'sku', 'role', 'department', 'location', 'tag', 'label'
+  ];
+
   /**
    * Scans a natural language question for potential metadata hints and keywords.
+   * @param question The user's query
+   * @param validFields Optional list of known filterable fields to look for
    */
-  static extractQueryFieldHints(question: string): QueryFieldHint[] {
+  static extractQueryFieldHints(question: string, validFields: string[] = []): QueryFieldHint[] {
     if (!question.trim()) return [];
 
     const hints = new Map<string, QueryFieldHint>();
+    const fieldsToSearch = [...new Set([...this.COMMON_METADATA_FIELDS, ...validFields])];
 
     const addHint = (value: string, field?: string) => {
       const normalizedValue = this.normalizeHintValue(value);
@@ -88,12 +97,35 @@ export class QueryProcessor {
       }
     }
 
-    // 4. Universal patterns (email, phone, etc.)
+    // 4. Dynamic Schema-driven patterns
+    // We look for common metadata fields (and any user-defined ones) in natural language.
+    if (fieldsToSearch.length > 0) {
+      for (const field of fieldsToSearch) {
+        // Pattern 1: "[field] is [value]", "[field]: [value]", "[field] of [value]"
+        const forwardPattern = new RegExp(`\\b${field}\\s*(?:is|are|:|of)?\\s+["']?([^"\\n?.!,]{2,60})["']?(?=[?.!,]|$)`, 'gi');
+        for (const match of question.matchAll(forwardPattern)) {
+          const value = match[1];
+          if (value && !this.isLikelyPromptPhrase(value)) {
+            addHint(value, field);
+          }
+        }
+
+        // Pattern 2: "[value] [field]" (e.g. "blue color", "electronics category")
+        const reversePattern = new RegExp(`\\b([^"\\n?.!,\\s]{2,60})\\s+${field}\\b`, 'gi');
+        for (const match of question.matchAll(reversePattern)) {
+          const value = match[1];
+          if (value && !this.isLikelyPromptPhrase(value)) {
+            addHint(value, field);
+          }
+        }
+      }
+    }
+
+    // 5. Universal patterns (email, phone, etc.)
     const universalPatterns: Array<{ regex: RegExp; field?: string; group?: number }> = [
       { regex: /([a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,})/gi, field: 'email', group: 1 },
       { regex: /(\+?\d[\d\-\.\s\(\)]{6,}\d)/g, field: 'phone', group: 1 },
       { regex: /(\b\d{4}-\d{2}-\d{2}\b|\b\d{1,2}\/\d{1,2}\/\d{2,4}\b)/g, field: 'date', group: 1 },
-      { regex: /(\$\s?\d{1,3}(?:,\d{3})*(?:\.\d+)?)/g, field: 'amount', group: 1 },
       { regex: /\b(ID|id|identifier)[: ]\s*([A-Za-z0-9\-]{3,})\b/gi, field: 'id', group: 2 },
       { regex: /"([^"]{2,120})"/g, group: 1 },
       { regex: /'([^']{2,120})'/g, group: 1 },