@retrivora-ai/rag-engine 0.3.7 → 0.3.9

package/dist/{MongoDBProvider-6X5F5YZS.mjs → MongoDBProvider-AJYTFXQA.mjs}

@@ -1,6 +1,6 @@
 import {
   MongoDBProvider
-} from "./chunk-DESUAKF3.mjs";
+} from "./chunk-DMXS64U2.mjs";
 import "./chunk-IMP6FUCY.mjs";
 import "./chunk-FWCSY2DS.mjs";
 export {

package/dist/{chunk-TYZC74CO.mjs → chunk-3DU2ASYU.mjs}

@@ -1079,7 +1079,6 @@ ${context}`;
   }
   async batchEmbed(texts, options) {
     var _a, _b, _c, _d, _e, _f, _g, _h, _i, _j;
-    const vectors = [];
     const model = this.sanitizeModel(
       (_c = (_b = options == null ? void 0 : options.model) != null ? _b : (_a = this.embeddingConfig) == null ? void 0 : _a.model) != null ? _c : "text-embedding-004"
     );
@@ -1379,7 +1378,7 @@ var ProviderRegistry = class {
         return new PostgreSQLProvider(config);
       }
       case "mongodb": {
-        const { MongoDBProvider } = await import("./MongoDBProvider-6X5F5YZS.mjs");
+        const { MongoDBProvider } = await import("./MongoDBProvider-AJYTFXQA.mjs");
         return new MongoDBProvider(config);
       }
       case "milvus": {

package/dist/{chunk-DESUAKF3.mjs → chunk-DMXS64U2.mjs}

@@ -56,7 +56,7 @@ var MongoDBProvider = class extends BaseVectorProvider {
     const pipeline = [
       {
         $vectorSearch: __spreadValues({
-          index: this.config.options.indexName || "vector_index",
+          index: this.config.indexName || "vector_index",
           path: this.embeddingKey,
           queryVector: vector,
           numCandidates: Math.max(topK * 10, 100),
@@ -94,6 +94,7 @@ var MongoDBProvider = class extends BaseVectorProvider {
     const sanitized = super.sanitizeFilter(filter);
     const mongoFilter = {};
     for (const [key, value] of Object.entries(sanitized)) {
+      if (key.startsWith("__")) continue;
       if (key === "namespace") {
         mongoFilter[key] = value;
       } else {

package/dist/handlers/index.js

@@ -424,7 +424,7 @@ var init_MongoDBProvider = __esm({
         const pipeline = [
           {
             $vectorSearch: __spreadValues({
-              index: this.config.options.indexName || "vector_index",
+              index: this.config.indexName || "vector_index",
               path: this.embeddingKey,
               queryVector: vector,
               numCandidates: Math.max(topK * 10, 100),
@@ -462,6 +462,7 @@ var init_MongoDBProvider = __esm({
         const sanitized = super.sanitizeFilter(filter);
         const mongoFilter = {};
         for (const [key, value] of Object.entries(sanitized)) {
+          if (key.startsWith("__")) continue;
           if (key === "namespace") {
             mongoFilter[key] = value;
           } else {
@@ -2273,7 +2274,6 @@ ${context}`;
   }
   async batchEmbed(texts, options) {
     var _a, _b, _c, _d, _e, _f, _g, _h, _i, _j;
-    const vectors = [];
     const model = this.sanitizeModel(
       (_c = (_b = options == null ? void 0 : options.model) != null ? _b : (_a = this.embeddingConfig) == null ? void 0 : _a.model) != null ? _c : "text-embedding-004"
     );

package/dist/handlers/index.mjs

@@ -3,7 +3,7 @@ import {
   createHealthHandler,
   createIngestHandler,
   createUploadHandler
-} from "../chunk-TYZC74CO.mjs";
+} from "../chunk-3DU2ASYU.mjs";
 import "../chunk-EDLTMSNY.mjs";
 import "../chunk-FWCSY2DS.mjs";
 export {

package/dist/server.js

@@ -436,7 +436,7 @@ var init_MongoDBProvider = __esm({
         const pipeline = [
           {
             $vectorSearch: __spreadValues({
-              index: this.config.options.indexName || "vector_index",
+              index: this.config.indexName || "vector_index",
               path: this.embeddingKey,
               queryVector: vector,
               numCandidates: Math.max(topK * 10, 100),
@@ -474,6 +474,7 @@ var init_MongoDBProvider = __esm({
         const sanitized = super.sanitizeFilter(filter);
         const mongoFilter = {};
         for (const [key, value] of Object.entries(sanitized)) {
+          if (key.startsWith("__")) continue;
           if (key === "namespace") {
             mongoFilter[key] = value;
           } else {
@@ -2317,7 +2318,6 @@ ${context}`;
   }
   async batchEmbed(texts, options) {
     var _a, _b, _c, _d, _e, _f, _g, _h, _i, _j;
-    const vectors = [];
     const model = this.sanitizeModel(
       (_c = (_b = options == null ? void 0 : options.model) != null ? _b : (_a = this.embeddingConfig) == null ? void 0 : _a.model) != null ? _c : "text-embedding-004"
     );

package/dist/server.mjs

@@ -34,7 +34,7 @@ import {
   createIngestHandler,
   createUploadHandler,
   getRagConfig
-} from "./chunk-TYZC74CO.mjs";
+} from "./chunk-3DU2ASYU.mjs";
 import "./chunk-EDLTMSNY.mjs";
 import {
   PineconeProvider
@@ -44,7 +44,7 @@ import {
 } from "./chunk-LJWWPTWE.mjs";
 import {
   MongoDBProvider
-} from "./chunk-DESUAKF3.mjs";
+} from "./chunk-DMXS64U2.mjs";
 import {
   MilvusProvider
 } from "./chunk-3QWAK3RZ.mjs";

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@retrivora-ai/rag-engine",
-  "version": "0.3.7",
+  "version": "0.3.9",
   "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/core/ConfigValidator.ts

@@ -512,7 +512,7 @@ export class ConfigValidator {
     }
 
     if (config.borderRadius) {
-      if (!UI_BORDER_RADIUS_OPTIONS.includes(config.borderRadius as any)) {
+      if (!UI_BORDER_RADIUS_OPTIONS.includes(config.borderRadius as (typeof UI_BORDER_RADIUS_OPTIONS)[number])) {
         errors.push({
           field: 'ui.borderRadius',
           message: `borderRadius must be one of: ${UI_BORDER_RADIUS_OPTIONS.join(', ')}`,
@@ -522,7 +522,7 @@ export class ConfigValidator {
     }
 
     if (config.visualStyle) {
-      if (!UI_VISUAL_STYLES.includes(config.visualStyle as any)) {
+      if (!UI_VISUAL_STYLES.includes(config.visualStyle as (typeof UI_VISUAL_STYLES)[number])) {
         errors.push({
           field: 'ui.visualStyle',
           message: `visualStyle must be one of: ${UI_VISUAL_STYLES.join(', ')}`,

package/src/llm/providers/GeminiProvider.ts

@@ -120,7 +120,6 @@ export class GeminiProvider implements ILLMProvider {
   }
 
   async batchEmbed(texts: string[], options?: EmbedOptions): Promise<number[][]> {
-    const vectors: number[][] = [];
     // Sequential fallback for Gemini (API doesn't have a direct batch endpoint in the same way, though you can pass multiple contents.
     // For simplicity and to match the other providers' structure, we'll iterate or pass array.
     // The new SDK supports an array of contents.

package/src/providers/vectordb/MongoDBProvider.ts

@@ -71,7 +71,7 @@ export class MongoDBProvider extends BaseVectorProvider {
     const pipeline: Record<string, unknown>[] = [
       {
         $vectorSearch: {
-          index: this.config.options.indexName || 'vector_index',
+          index: this.config.indexName || 'vector_index',
           path: this.embeddingKey,
           queryVector: vector,
           numCandidates: Math.max(topK * 10, 100),
@@ -116,7 +116,10 @@ export class MongoDBProvider extends BaseVectorProvider {
     const mongoFilter: Record<string, unknown> = {};
 
     for (const [key, value] of Object.entries(sanitized)) {
-      // If the key is already a metadata field or special key, keep it (e.g. namespace)
+      // Skip internal engine fields
+      if (key.startsWith('__')) continue;
+
+      // If the key is already a special key, keep it (e.g. namespace)
       if (key === 'namespace') {
         mongoFilter[key] = value;
       } else {

package/src/providers/vectordb/PineconeProvider.ts

@@ -71,7 +71,7 @@ export class PineconeProvider extends BaseVectorProvider {
       vector,
       topK,
       includeMetadata: true,
-      ...(Object.keys(pineconeFilter).length > 0 ? { filter: pineconeFilter as any } : {}),
+      ...(Object.keys(pineconeFilter).length > 0 ? { filter: pineconeFilter as Record<string, string | number | boolean | string[] | number[]> } : {}),
     });
     return (result.matches ?? []).map((m) => ({
       id: m.id,