libsql-search 0.10.1 → 0.11.1

package/README.md

@@ -9,7 +9,7 @@
 
 Use it when you want:
 
-- one indexing/search API across local and hosted embedding providers
+- one indexing/search API across external embedding providers
 - direct control over vector dimensions, table names, and deployment shape
 - a lightweight library instead of a hosted search product
 
@@ -45,7 +45,7 @@ Note for `0.17.x`: the client no longer exports `./package.json`, so `require("@
 
 ## Quick Start
 
-This example uses the default local provider. Local embeddings run in-process after the initial model download and cache warmup; they are not automatically air-gapped.
+This example uses a separately deployed OpenAI-compatible embedding service. `libsql-search` never loads or hosts an embedding model in-process.
 
 ```ts
 import { createClient } from "@libsql/client";
@@ -56,25 +56,29 @@ const client = createClient({
   authToken: "your-auth-token",
 });
 
-await createTable(client, "articles_local_384", 384);
+const embeddingOptions = {
+  provider: "openai-compatible" as const,
+  baseUrl: process.env.EMBEDDING_BASE_URL!,
+  apiKey: process.env.EMBEDDING_API_KEY,
+  model: "bge-large-en-v1.5",
+  dimensions: 1024,
+};
+
+await createTable(client, "articles_bge_1024", 1024);
 
 await indexContent({
   client,
   contentPath: "./content",
-  tableName: "articles_local_384",
-  embeddingOptions: {
-    provider: "local",
-  },
+  tableName: "articles_bge_1024",
+  embeddingOptions,
 });
 
 const results = await search({
   client,
   query: "how do I deploy my docs site",
-  tableName: "articles_local_384",
+  tableName: "articles_bge_1024",
   limit: 5,
-  embeddingOptions: {
-    provider: "local",
-  },
+  embeddingOptions,
 });
 
 console.log(results.map((result) => ({
@@ -91,7 +95,7 @@ Important behavior:
 - `indexContent()` embeds every document before it touches the database, then replaces the table in one transaction, so a failed rebuild leaves the previous index intact.
 - `indexContent()` throws `IndexingError` when a file fails; pass `failurePolicy: "skip"` to rebuild from the remaining files.
 - `indexContent()` throws `IndexingError` when no source files are found; pass `allowEmptyIndex: true` to intentionally empty the index.
-- Hosted providers send indexed and queried text to external services and may incur provider charges.
+- Every provider sends indexed and queried text to an external service; review that service's privacy, retention, and pricing terms.
 
 ## Search Accuracy And Performance
 
@@ -103,10 +107,10 @@ Two options control the trade-off:
 
 ```ts
 // Widen the index probe to raise recall (default: max(limit * 4, 32))
-await search({ client, query, limit: 10, candidates: 200 });
+await search({ client, query, embeddingOptions, limit: 10, candidates: 200 });
 
 // Bypass the index entirely: exact, but linear in table size
-await search({ client, query, exact: true });
+await search({ client, query, embeddingOptions, exact: true });
 ```
 
 `exact: true` is the only way to guarantee exactness. Use it for small corpora, for correctness checks against the index path, and for tables that have no vector index.
@@ -117,7 +121,6 @@ Requirements: `vector_top_k()` and `libsql_vector_idx()` need a libSQL build wit
 
 Built-in providers:
 
-- `local` with `Xenova/all-MiniLM-L6-v2` at 384 dimensions
 - `cloudflare` with `@cf/baai/bge-m3` at 1024 dimensions
 - `mistral` with `mistral-embed` at 1024 dimensions
 - `gemini` with `gemini-embedding-2` at 128-3072 dimensions, default 3072

package/dist/index.cjs

@@ -6,10 +6,8 @@ var matter = require('gray-matter');
 
 const OPENAI_DEFAULT_DIMENSIONS = 768;
 const OPENAI_COMPATIBLE_DEFAULT_BATCH_SIZE = 32;
-const LOCAL_DIMENSIONS = 384;
 const DEFAULT_MAX_LENGTH = 8e3;
 const DEFAULT_TIMEOUT_MS = 3e4;
-const LOCAL_MODEL = "Xenova/all-MiniLM-L6-v2";
 const GEMINI_MODEL = "gemini-embedding-2";
 const GEMINI_DIMENSIONS = 3072;
 const GEMINI_MIN_DIMENSIONS = 128;
@@ -22,7 +20,6 @@ const MISTRAL_EMBEDDINGS_URL = "https://api.mistral.ai/v1/embeddings";
 const MISTRAL_DIMENSIONS = 1024;
 const CLOUDFLARE_MODEL = "@cf/baai/bge-m3";
 const CLOUDFLARE_DIMENSIONS = 1024;
-const localModelCacheByModel = /* @__PURE__ */ new Map();
 function getEnvironmentVariable(name) {
   const runtime = globalThis;
   const nodeValue = runtime.process?.env?.[name];
@@ -35,67 +32,6 @@ function getEnvironmentVariable(name) {
     return void 0;
   }
 }
-function deletePendingLocalModelCache(modelName, entry) {
-  if (!entry.settled && entry.waiters === 0 && localModelCacheByModel.get(modelName) === entry) {
-    localModelCacheByModel.delete(modelName);
-  }
-}
-async function getLocalEmbeddingModel(modelName, signal) {
-  const cached = localModelCacheByModel.get(modelName);
-  if (cached) {
-    cached.waiters++;
-    try {
-      return await waitForLocalEmbeddingModel(modelName, cached, signal);
-    } finally {
-      cached.waiters--;
-      deletePendingLocalModelCache(modelName, cached);
-    }
-  }
-  const modelPromise = (async () => {
-    console.log(`Loading local embedding model (${modelName})...`);
-    const { pipeline } = await import('@huggingface/transformers');
-    const model = await pipeline("feature-extraction", modelName);
-    console.log("Local model loaded successfully");
-    return model;
-  })();
-  const entry = {
-    promise: modelPromise,
-    settled: false,
-    waiters: 1
-  };
-  localModelCacheByModel.set(modelName, entry);
-  modelPromise.then(() => {
-    entry.settled = true;
-  }).catch(() => {
-    localModelCacheByModel.delete(modelName);
-  });
-  try {
-    return await waitForLocalEmbeddingModel(modelName, entry, signal);
-  } finally {
-    entry.waiters--;
-    deletePendingLocalModelCache(modelName, entry);
-  }
-}
-async function waitForLocalEmbeddingModel(modelName, entry, signal) {
-  if (signal.aborted) {
-    deletePendingLocalModelCache(modelName, entry);
-    throw providerError("local", "model inference was aborted");
-  }
-  let rejectAbort = () => {
-  };
-  const abortPromise = new Promise((_resolve, reject) => {
-    rejectAbort = reject;
-  });
-  const onAbort = () => {
-    rejectAbort(providerError("local", "model inference was aborted"));
-  };
-  signal.addEventListener("abort", onAbort, { once: true });
-  try {
-    return await Promise.race([entry.promise, abortPromise]);
-  } finally {
-    signal.removeEventListener("abort", onAbort);
-  }
-}
 function getPositiveInteger(value, optionName) {
   if (!Number.isInteger(value) || value <= 0) {
     throw new Error(`Invalid ${optionName}: expected a positive integer`);
@@ -106,15 +42,17 @@ function getTimeoutMs(value) {
   return getPositiveInteger(value ?? DEFAULT_TIMEOUT_MS, "timeoutMs");
 }
 function resolveProviderName(provider) {
-  switch (provider ?? "local") {
-    case "local":
+  switch (provider) {
     case "gemini":
     case "openai":
     case "mistral":
     case "cloudflare":
     case "openai-compatible":
-      return provider ?? "local";
+      return provider;
     default:
+      if (provider === void 0) {
+        throw new Error("Embedding provider is required");
+      }
       throw new Error(`Unknown embedding provider: ${String(provider)}`);
   }
 }
@@ -326,13 +264,6 @@ function normalizeOpenAICompatibleEmbeddingsUrl(baseUrl) {
 }
 function createProviderMetadata(provider, dimensions, model) {
   switch (provider) {
-    case "local":
-      return Object.freeze({
-        name: "local",
-        model: LOCAL_MODEL,
-        dimensions,
-        batch: Object.freeze({ mode: "sequential" })
-      });
     case "gemini":
       return Object.freeze({
         name: "gemini",
@@ -371,15 +302,6 @@ function createProviderMetadata(provider, dimensions, model) {
   }
 }
 function getEffectiveDimensions(provider, dimensions) {
-  if (provider === "local") {
-    if (dimensions !== void 0 && dimensions !== LOCAL_DIMENSIONS) {
-      throw providerError(
-        "local",
-        `${LOCAL_MODEL} returns ${LOCAL_DIMENSIONS} dimensions; received dimensions ${String(dimensions)}`
-      );
-    }
-    return LOCAL_DIMENSIONS;
-  }
   if (provider === "gemini") {
     const effectiveDimensions = dimensions ?? GEMINI_DIMENSIONS;
     if (!Number.isInteger(effectiveDimensions) || effectiveDimensions < GEMINI_MIN_DIMENSIONS || effectiveDimensions > GEMINI_MAX_DIMENSIONS) {
@@ -440,42 +362,6 @@ function createEmbeddingBatchResult(metadata, intent, embeddings) {
     intent
   });
 }
-class LocalEmbeddingProvider {
-  metadata;
-  #timeoutMs;
-  constructor(metadata, timeoutMs) {
-    this.metadata = metadata;
-    this.#timeoutMs = timeoutMs;
-  }
-  async embed(texts, options = {}) {
-    const intent = resolveIntent(options.intent);
-    if (texts.length === 0) {
-      return createEmbeddingBatchResult(this.metadata, intent, []);
-    }
-    assertBatchSize(this.metadata, texts.length);
-    const vectors = await withTimeout(
-      "local",
-      "model inference",
-      this.#timeoutMs,
-      async (signal) => {
-        const model = await getLocalEmbeddingModel(this.metadata.model, signal);
-        return embedSequentially("local", "model inference", texts, signal, async (text) => {
-          const output = await model(text, {
-            pooling: "mean",
-            normalize: true
-          });
-          return Array.from(output.data);
-        });
-      },
-      options.signal
-    );
-    return createEmbeddingBatchResult(
-      this.metadata,
-      intent,
-      validateEmbeddingBatch(vectors, texts.length, this.metadata.dimensions, "local")
-    );
-  }
-}
 class GeminiEmbeddingProvider {
   metadata;
   #apiKey;
@@ -628,13 +514,11 @@ class OpenAICompatibleEmbeddingProvider {
     );
   }
 }
-function createEmbeddingProvider(options = {}) {
+function createEmbeddingProvider(options) {
   const provider = resolveProviderName(options.provider);
   const metadata = getEmbeddingProviderMetadata(options);
   const timeoutMs = getTimeoutMs(options.timeoutMs);
   switch (provider) {
-    case "local":
-      return new LocalEmbeddingProvider(metadata, timeoutMs);
     case "gemini": {
       const key = getOptionalTrimmedCredential(
         options.apiKey ?? getEnvironmentVariable("GEMINI_API_KEY")
@@ -733,7 +617,7 @@ function createEmbeddingProvider(options = {}) {
     }
   }
 }
-function getEmbeddingProviderMetadata(options = {}) {
+function getEmbeddingProviderMetadata(options) {
   const provider = resolveProviderName(options.provider);
   if (provider === "openai-compatible") {
     normalizeOpenAICompatibleEmbeddingsUrl(getRequiredTrimmedString(options.baseUrl, "baseUrl"));
@@ -743,7 +627,7 @@ function getEmbeddingProviderMetadata(options = {}) {
   }
   return createProviderMetadata(provider, getEffectiveDimensions(provider, options.dimensions), options.model);
 }
-async function generateEmbeddings(texts, options = {}) {
+async function generateEmbeddings(texts, options) {
   const maxLength = getPositiveInteger(options.maxLength ?? DEFAULT_MAX_LENGTH, "maxLength");
   const intent = resolveIntent(options.intent);
   if (texts.length === 0) {
@@ -756,7 +640,7 @@ async function generateEmbeddings(texts, options = {}) {
   });
   return result.embeddings;
 }
-async function generateEmbedding(text, options = {}) {
+async function generateEmbedding(text, options) {
   const [embedding] = await generateEmbeddings([text], options);
   if (!embedding) {
     throw new Error("Embedding provider returned no embedding");
@@ -889,7 +773,7 @@ async function indexContent(options) {
   const {
     client,
     contentPath,
-    embeddingOptions = {},
+    embeddingOptions,
     fileExtensions = [".md", ".markdown"],
     exclude = ["node_modules", ".git", "dist", "build"],
     tableName = "articles",
@@ -899,6 +783,9 @@ async function indexContent(options) {
   } = options;
   const database = resolveDatabase(client);
   const quotedTableName = quoteSqlIdentifier(tableName, "tableName");
+  if (embeddingOptions === void 0) {
+    throw new TypeError("embeddingOptions is required");
+  }
   let files;
   try {
     files = await findFiles(contentPath, contentPath, fileExtensions, exclude);
@@ -1178,7 +1065,7 @@ async function search(options) {
     query,
     limit = 10,
     tableName = "articles",
-    embeddingOptions = {},
+    embeddingOptions,
     candidates,
     exact = false
   } = options;
@@ -1190,6 +1077,9 @@ async function search(options) {
     "embedding index name"
   );
   const candidateCount = normalizeSearchCandidates(candidates, resultLimit);
+  if (embeddingOptions === void 0) {
+    throw new TypeError("embeddingOptions is required");
+  }
   const queryEmbedding = await generateEmbedding(query, {
     ...embeddingOptions,
     intent: embeddingOptions.intent ?? "query"

package/dist/index.d.ts

@@ -2,10 +2,10 @@ import { Client } from '@libsql/client';
 
 /**
  * Multi-provider embedding generation
- * Supports local Hugging Face Transformers, Gemini, OpenAI, Mistral,
- * Cloudflare Workers AI, and custom OpenAI-compatible endpoints
+ * Supports Gemini, OpenAI, Mistral, Cloudflare Workers AI, and custom
+ * OpenAI-compatible endpoints.
  */
-type EmbeddingProvider = 'local' | 'gemini' | 'openai' | 'mistral' | 'cloudflare' | 'openai-compatible';
+type EmbeddingProvider = 'gemini' | 'openai' | 'mistral' | 'cloudflare' | 'openai-compatible';
 type EmbeddingIntent = 'document' | 'query';
 type EmbeddingBatchMode = 'native' | 'sequential';
 interface EmbeddingBatchBehavior {
@@ -34,7 +34,7 @@ interface EmbeddingBatchResult {
     intent: EmbeddingIntent;
 }
 interface EmbeddingOptions {
-    provider?: EmbeddingProvider;
+    provider: EmbeddingProvider;
     apiKey?: string;
     accountId?: string;
     apiToken?: string;
@@ -53,16 +53,16 @@ interface EmbeddingBatchItemResult {
 }
 type EmbeddingBatchItem = number[] | EmbeddingBatchItemResult;
 declare function validateEmbeddingBatch(items: EmbeddingBatchItem[], expectedCount: number, expectedDimensions: number, provider: EmbeddingProvider): number[][];
-declare function createEmbeddingProvider(options?: EmbeddingOptions): EmbeddingProviderClient;
-declare function getEmbeddingProviderMetadata(options?: EmbeddingOptions): EmbeddingProviderMetadata;
+declare function createEmbeddingProvider(options: EmbeddingOptions): EmbeddingProviderClient;
+declare function getEmbeddingProviderMetadata(options: EmbeddingOptions): EmbeddingProviderMetadata;
 /**
  * Generate embeddings using the specified provider.
  */
-declare function generateEmbeddings(texts: string[], options?: EmbeddingOptions): Promise<number[][]>;
+declare function generateEmbeddings(texts: string[], options: EmbeddingOptions): Promise<number[][]>;
 /**
  * Generate an embedding using the specified provider.
  */
-declare function generateEmbedding(text: string, options?: EmbeddingOptions): Promise<number[]>;
+declare function generateEmbedding(text: string, options: EmbeddingOptions): Promise<number[]>;
 /**
  * Pad or truncate embedding to target dimensions
  */
@@ -189,7 +189,7 @@ interface IndexerOptions {
      */
     client: DatabaseClient;
     contentPath: string;
-    embeddingOptions?: EmbeddingOptions;
+    embeddingOptions: EmbeddingOptions;
     fileExtensions?: string[];
     exclude?: string[];
     tableName?: string;
@@ -282,7 +282,7 @@ interface SearchOptions {
     query: string;
     limit?: number;
     tableName?: string;
-    embeddingOptions?: EmbeddingOptions;
+    embeddingOptions: EmbeddingOptions;
     /**
      * How many candidates to pull from the vector index before the exact
      * re-rank. Must be an integer from `limit` through {@link MAX_SEARCH_CANDIDATES}.

package/dist/index.esm.js

@@ -4,10 +4,8 @@ import matter from 'gray-matter';
 
 const OPENAI_DEFAULT_DIMENSIONS = 768;
 const OPENAI_COMPATIBLE_DEFAULT_BATCH_SIZE = 32;
-const LOCAL_DIMENSIONS = 384;
 const DEFAULT_MAX_LENGTH = 8e3;
 const DEFAULT_TIMEOUT_MS = 3e4;
-const LOCAL_MODEL = "Xenova/all-MiniLM-L6-v2";
 const GEMINI_MODEL = "gemini-embedding-2";
 const GEMINI_DIMENSIONS = 3072;
 const GEMINI_MIN_DIMENSIONS = 128;
@@ -20,7 +18,6 @@ const MISTRAL_EMBEDDINGS_URL = "https://api.mistral.ai/v1/embeddings";
 const MISTRAL_DIMENSIONS = 1024;
 const CLOUDFLARE_MODEL = "@cf/baai/bge-m3";
 const CLOUDFLARE_DIMENSIONS = 1024;
-const localModelCacheByModel = /* @__PURE__ */ new Map();
 function getEnvironmentVariable(name) {
   const runtime = globalThis;
   const nodeValue = runtime.process?.env?.[name];
@@ -33,67 +30,6 @@ function getEnvironmentVariable(name) {
     return void 0;
   }
 }
-function deletePendingLocalModelCache(modelName, entry) {
-  if (!entry.settled && entry.waiters === 0 && localModelCacheByModel.get(modelName) === entry) {
-    localModelCacheByModel.delete(modelName);
-  }
-}
-async function getLocalEmbeddingModel(modelName, signal) {
-  const cached = localModelCacheByModel.get(modelName);
-  if (cached) {
-    cached.waiters++;
-    try {
-      return await waitForLocalEmbeddingModel(modelName, cached, signal);
-    } finally {
-      cached.waiters--;
-      deletePendingLocalModelCache(modelName, cached);
-    }
-  }
-  const modelPromise = (async () => {
-    console.log(`Loading local embedding model (${modelName})...`);
-    const { pipeline } = await import('@huggingface/transformers');
-    const model = await pipeline("feature-extraction", modelName);
-    console.log("Local model loaded successfully");
-    return model;
-  })();
-  const entry = {
-    promise: modelPromise,
-    settled: false,
-    waiters: 1
-  };
-  localModelCacheByModel.set(modelName, entry);
-  modelPromise.then(() => {
-    entry.settled = true;
-  }).catch(() => {
-    localModelCacheByModel.delete(modelName);
-  });
-  try {
-    return await waitForLocalEmbeddingModel(modelName, entry, signal);
-  } finally {
-    entry.waiters--;
-    deletePendingLocalModelCache(modelName, entry);
-  }
-}
-async function waitForLocalEmbeddingModel(modelName, entry, signal) {
-  if (signal.aborted) {
-    deletePendingLocalModelCache(modelName, entry);
-    throw providerError("local", "model inference was aborted");
-  }
-  let rejectAbort = () => {
-  };
-  const abortPromise = new Promise((_resolve, reject) => {
-    rejectAbort = reject;
-  });
-  const onAbort = () => {
-    rejectAbort(providerError("local", "model inference was aborted"));
-  };
-  signal.addEventListener("abort", onAbort, { once: true });
-  try {
-    return await Promise.race([entry.promise, abortPromise]);
-  } finally {
-    signal.removeEventListener("abort", onAbort);
-  }
-}
 function getPositiveInteger(value, optionName) {
   if (!Number.isInteger(value) || value <= 0) {
     throw new Error(`Invalid ${optionName}: expected a positive integer`);
@@ -104,15 +40,17 @@ function getTimeoutMs(value) {
   return getPositiveInteger(value ?? DEFAULT_TIMEOUT_MS, "timeoutMs");
 }
 function resolveProviderName(provider) {
-  switch (provider ?? "local") {
-    case "local":
+  switch (provider) {
     case "gemini":
     case "openai":
     case "mistral":
     case "cloudflare":
     case "openai-compatible":
-      return provider ?? "local";
+      return provider;
     default:
+      if (provider === void 0) {
+        throw new Error("Embedding provider is required");
+      }
       throw new Error(`Unknown embedding provider: ${String(provider)}`);
   }
 }
@@ -324,13 +262,6 @@ function normalizeOpenAICompatibleEmbeddingsUrl(baseUrl) {
 }
 function createProviderMetadata(provider, dimensions, model) {
   switch (provider) {
-    case "local":
-      return Object.freeze({
-        name: "local",
-        model: LOCAL_MODEL,
-        dimensions,
-        batch: Object.freeze({ mode: "sequential" })
-      });
     case "gemini":
       return Object.freeze({
         name: "gemini",
@@ -369,15 +300,6 @@ function createProviderMetadata(provider, dimensions, model) {
   }
 }
 function getEffectiveDimensions(provider, dimensions) {
-  if (provider === "local") {
-    if (dimensions !== void 0 && dimensions !== LOCAL_DIMENSIONS) {
-      throw providerError(
-        "local",
-        `${LOCAL_MODEL} returns ${LOCAL_DIMENSIONS} dimensions; received dimensions ${String(dimensions)}`
-      );
-    }
-    return LOCAL_DIMENSIONS;
-  }
   if (provider === "gemini") {
     const effectiveDimensions = dimensions ?? GEMINI_DIMENSIONS;
     if (!Number.isInteger(effectiveDimensions) || effectiveDimensions < GEMINI_MIN_DIMENSIONS || effectiveDimensions > GEMINI_MAX_DIMENSIONS) {
@@ -438,42 +360,6 @@ function createEmbeddingBatchResult(metadata, intent, embeddings) {
     intent
   });
 }
-class LocalEmbeddingProvider {
-  metadata;
-  #timeoutMs;
-  constructor(metadata, timeoutMs) {
-    this.metadata = metadata;
-    this.#timeoutMs = timeoutMs;
-  }
-  async embed(texts, options = {}) {
-    const intent = resolveIntent(options.intent);
-    if (texts.length === 0) {
-      return createEmbeddingBatchResult(this.metadata, intent, []);
-    }
-    assertBatchSize(this.metadata, texts.length);
-    const vectors = await withTimeout(
-      "local",
-      "model inference",
-      this.#timeoutMs,
-      async (signal) => {
-        const model = await getLocalEmbeddingModel(this.metadata.model, signal);
-        return embedSequentially("local", "model inference", texts, signal, async (text) => {
-          const output = await model(text, {
-            pooling: "mean",
-            normalize: true
-          });
-          return Array.from(output.data);
-        });
-      },
-      options.signal
-    );
-    return createEmbeddingBatchResult(
-      this.metadata,
-      intent,
-      validateEmbeddingBatch(vectors, texts.length, this.metadata.dimensions, "local")
-    );
-  }
-}
 class GeminiEmbeddingProvider {
   metadata;
   #apiKey;
@@ -626,13 +512,11 @@ class OpenAICompatibleEmbeddingProvider {
     );
   }
 }
-function createEmbeddingProvider(options = {}) {
+function createEmbeddingProvider(options) {
   const provider = resolveProviderName(options.provider);
   const metadata = getEmbeddingProviderMetadata(options);
   const timeoutMs = getTimeoutMs(options.timeoutMs);
   switch (provider) {
-    case "local":
-      return new LocalEmbeddingProvider(metadata, timeoutMs);
     case "gemini": {
       const key = getOptionalTrimmedCredential(
         options.apiKey ?? getEnvironmentVariable("GEMINI_API_KEY")
@@ -731,7 +615,7 @@ function createEmbeddingProvider(options = {}) {
     }
   }
 }
-function getEmbeddingProviderMetadata(options = {}) {
+function getEmbeddingProviderMetadata(options) {
   const provider = resolveProviderName(options.provider);
   if (provider === "openai-compatible") {
     normalizeOpenAICompatibleEmbeddingsUrl(getRequiredTrimmedString(options.baseUrl, "baseUrl"));
@@ -741,7 +625,7 @@ function getEmbeddingProviderMetadata(options = {}) {
   }
   return createProviderMetadata(provider, getEffectiveDimensions(provider, options.dimensions), options.model);
 }
-async function generateEmbeddings(texts, options = {}) {
+async function generateEmbeddings(texts, options) {
   const maxLength = getPositiveInteger(options.maxLength ?? DEFAULT_MAX_LENGTH, "maxLength");
   const intent = resolveIntent(options.intent);
   if (texts.length === 0) {
@@ -754,7 +638,7 @@ async function generateEmbeddings(texts, options = {}) {
   });
   return result.embeddings;
 }
-async function generateEmbedding(text, options = {}) {
+async function generateEmbedding(text, options) {
   const [embedding] = await generateEmbeddings([text], options);
   if (!embedding) {
     throw new Error("Embedding provider returned no embedding");
@@ -887,7 +771,7 @@ async function indexContent(options) {
   const {
     client,
     contentPath,
-    embeddingOptions = {},
+    embeddingOptions,
     fileExtensions = [".md", ".markdown"],
     exclude = ["node_modules", ".git", "dist", "build"],
     tableName = "articles",
@@ -897,6 +781,9 @@ async function indexContent(options) {
   } = options;
   const database = resolveDatabase(client);
   const quotedTableName = quoteSqlIdentifier(tableName, "tableName");
+  if (embeddingOptions === void 0) {
+    throw new TypeError("embeddingOptions is required");
+  }
   let files;
   try {
     files = await findFiles(contentPath, contentPath, fileExtensions, exclude);
@@ -1176,7 +1063,7 @@ async function search(options) {
     query,
     limit = 10,
     tableName = "articles",
-    embeddingOptions = {},
+    embeddingOptions,
     candidates,
     exact = false
   } = options;
@@ -1188,6 +1075,9 @@ async function search(options) {
     "embedding index name"
   );
   const candidateCount = normalizeSearchCandidates(candidates, resultLimit);
+  if (embeddingOptions === void 0) {
+    throw new TypeError("embeddingOptions is required");
+  }
   const queryEmbedding = await generateEmbedding(query, {
     ...embeddingOptions,
     intent: embeddingOptions.intent ?? "query"