@google/adk 0.1.0

package/dist/web/index_web.js

@@ -0,0 +1,6 @@
+/**
+ * @license
+ * Copyright 2025 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+export * from "./common.js";

package/dist/web/memory/base_memory_service.js

@@ -0,0 +1,5 @@
+/**
+ * @license
+ * Copyright 2025 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */

package/dist/web/memory/in_memory_memory_service.js

@@ -0,0 +1,67 @@
+/**
+ * @license
+ * Copyright 2025 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+class InMemoryMemoryService {
+  constructor() {
+    this.memories = [];
+    this.sessionEvents = {};
+  }
+  async addSessionToMemory(session) {
+    const userKey = getUserKey(session.appName, session.userId);
+    if (!this.sessionEvents[userKey]) {
+      this.sessionEvents[userKey] = {};
+    }
+    this.sessionEvents[userKey][session.id] = session.events.filter(
+      (event) => {
+        var _a, _b, _c;
+        return ((_c = (_b = (_a = event.content) == null ? void 0 : _a.parts) == null ? void 0 : _b.length) != null ? _c : 0) > 0;
+      }
+    );
+  }
+  async searchMemory(req) {
+    var _a, _b;
+    const userKey = getUserKey(req.appName, req.userId);
+    if (!this.sessionEvents[userKey]) {
+      return Promise.resolve({ memories: [] });
+    }
+    const wordsInQuery = req.query.toLowerCase().split(/\s+/);
+    const response = { memories: [] };
+    for (const sessionEvents of Object.values(this.sessionEvents[userKey])) {
+      for (const event of sessionEvents) {
+        if (!((_b = (_a = event.content) == null ? void 0 : _a.parts) == null ? void 0 : _b.length)) {
+          continue;
+        }
+        const joinedText = event.content.parts.map((part) => part.text).filter((text) => !!text).join(" ");
+        const wordsInEvent = extractWordsLower(joinedText);
+        if (!wordsInEvent.size) {
+          continue;
+        }
+        const matchQuery = wordsInQuery.some((queryWord) => wordsInEvent.has(queryWord));
+        if (matchQuery) {
+          response.memories.push({
+            content: event.content,
+            author: event.author,
+            timestamp: formatTimestamp(event.timestamp)
+          });
+        }
+      }
+    }
+    return response;
+  }
+}
+function getUserKey(appName, userId) {
+  return "".concat(appName, "/").concat(userId);
+}
+function extractWordsLower(text) {
+  return new Set(
+    [...text.matchAll(/[A-Za-z]+/)].map((match) => match[0].toLowerCase())
+  );
+}
+function formatTimestamp(timestamp) {
+  return new Date(timestamp).toISOString();
+}
+export {
+  InMemoryMemoryService
+};

package/dist/web/memory/memory_entry.js

@@ -0,0 +1,5 @@
+/**
+ * @license
+ * Copyright 2025 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */

package/dist/web/models/base_llm.js

@@ -0,0 +1,47 @@
+/**
+ * @license
+ * Copyright 2025 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+class BaseLlm {
+  /**
+   * Creates an instance of BaseLLM.
+   *
+   * @param model The name of the LLM, e.g. gemini-1.5-flash or
+   *     gemini-1.5-flash-001.
+   */
+  constructor(model) {
+    this.model = model;
+  }
+  /**
+   * Appends a user content, so that model can continue to output.
+   *
+   * @param llmRequest LlmRequest, the request to send to the LLM.
+   */
+  maybeAppendUserContent(llmRequest) {
+    var _a;
+    if (llmRequest.contents.length === 0) {
+      llmRequest.contents.push({
+        role: "user",
+        parts: [
+          { text: "Handle the requests as specified in the System Instruction." }
+        ]
+      });
+    }
+    if (((_a = llmRequest.contents[llmRequest.contents.length - 1]) == null ? void 0 : _a.role) !== "user") {
+      llmRequest.contents.push({
+        role: "user",
+        parts: [{
+          text: "Continue processing previous requests as instructed. Exit or provide a summary if no more outputs are needed."
+        }]
+      });
+    }
+  }
+}
+/**
+ * List of supported models in regex for LlmRegistry.
+ */
+BaseLlm.supportedModels = [];
+export {
+  BaseLlm
+};

package/dist/web/models/base_llm_connection.js

@@ -0,0 +1,5 @@
+/**
+ * @license
+ * Copyright 2025 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */

package/dist/web/models/gemini_llm_connection.js

@@ -0,0 +1,120 @@
+var __knownSymbol = (name, symbol) => (symbol = Symbol[name]) ? symbol : Symbol.for("Symbol." + name);
+var __await = function(promise, isYieldStar) {
+  this[0] = promise;
+  this[1] = isYieldStar;
+};
+var __asyncGenerator = (__this, __arguments, generator) => {
+  var resume = (k, v, yes, no) => {
+    try {
+      var x = generator[k](v), isAwait = (v = x.value) instanceof __await, done = x.done;
+      Promise.resolve(isAwait ? v[0] : v).then((y) => isAwait ? resume(k === "return" ? k : "next", v[1] ? { done: y.done, value: y.value } : y, yes, no) : yes({ value: y, done })).catch((e) => resume("throw", e, yes, no));
+    } catch (e) {
+      no(e);
+    }
+  }, method = (k) => it[k] = (x) => new Promise((yes, no) => resume(k, x, yes, no)), it = {};
+  return generator = generator.apply(__this, __arguments), it[__knownSymbol("asyncIterator")] = () => it, method("next"), method("throw"), method("return"), it;
+};
+/**
+ * @license
+ * Copyright 2025 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+import { logger } from "../utils/logger.js";
+class GeminiLlmConnection {
+  constructor(geminiSession) {
+    this.geminiSession = geminiSession;
+  }
+  /**
+   * Sends the conversation history to the gemini model.
+   *
+   * You call this method right after setting up the model connection.
+   * The model will respond if the last content is from user, otherwise it will
+   * wait for new user input before responding.
+   *
+   * @param history The conversation history to send to the model.
+   */
+  async sendHistory(history) {
+    const contents = history.filter(
+      (content) => {
+        var _a;
+        return content.parts && ((_a = content.parts[0]) == null ? void 0 : _a.text);
+      }
+    );
+    if (contents.length > 0) {
+      this.geminiSession.sendClientContent({
+        turns: contents,
+        turnComplete: contents[contents.length - 1].role === "user"
+      });
+    } else {
+      logger.info("no content is sent");
+    }
+  }
+  /**
+   * Sends a user content to the gemini model.
+   *
+   * The model will respond immediately upon receiving the content.
+   * If you send function responses, all parts in the content should be function
+   * responses.
+   *
+   * @param content The content to send to the model.
+   */
+  async sendContent(content) {
+    if (!content.parts) {
+      throw new Error("Content must have parts.");
+    }
+    if (content.parts[0].functionResponse) {
+      const functionResponses = content.parts.map((part) => part.functionResponse).filter((fr) => !!fr);
+      logger.debug("Sending LLM function response:", functionResponses);
+      this.geminiSession.sendToolResponse({
+        functionResponses
+      });
+    } else {
+      logger.debug("Sending LLM new content", content);
+      this.geminiSession.sendClientContent({
+        turns: [content],
+        turnComplete: true
+      });
+    }
+  }
+  /**
+   * Sends a chunk of audio or a frame of video to the model in realtime.
+   *
+   * @param blob The blob to send to the model.
+   */
+  async sendRealtime(blob) {
+    logger.debug("Sending LLM Blob:", blob);
+    this.geminiSession.sendRealtimeInput({ media: blob });
+  }
+  /**
+   * Builds a full text response.
+   *
+   * The text should not be partial and the returned LlmResponse is not be
+   * partial.
+   *
+   * @param text The text to be included in the response.
+   * @returns An LlmResponse containing the full text.
+   */
+  buildFullTextResponse(text) {
+    return {
+      content: {
+        role: "model",
+        parts: [{ text }]
+      }
+    };
+  }
+  // TODO(b/425992518): GenAI SDK inconsistent API, missing methods.
+  receive() {
+    return __asyncGenerator(this, null, function* () {
+      throw new Error("Not Implemented.");
+    });
+  }
+  /**
+   * Closes the llm server connection.
+   */
+  async close() {
+    this.geminiSession.close();
+  }
+}
+export {
+  GeminiLlmConnection
+};

package/dist/web/models/google_llm.js

@@ -0,0 +1,332 @@
+var __defProp = Object.defineProperty;
+var __getOwnPropSymbols = Object.getOwnPropertySymbols;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __propIsEnum = Object.prototype.propertyIsEnumerable;
+var __knownSymbol = (name, symbol) => (symbol = Symbol[name]) ? symbol : Symbol.for("Symbol." + name);
+var __defNormalProp = (obj, key, value) => key in obj ? __defProp(obj, key, { enumerable: true, configurable: true, writable: true, value }) : obj[key] = value;
+var __spreadValues = (a, b) => {
+  for (var prop in b || (b = {}))
+    if (__hasOwnProp.call(b, prop))
+      __defNormalProp(a, prop, b[prop]);
+  if (__getOwnPropSymbols)
+    for (var prop of __getOwnPropSymbols(b)) {
+      if (__propIsEnum.call(b, prop))
+        __defNormalProp(a, prop, b[prop]);
+    }
+  return a;
+};
+var __await = function(promise, isYieldStar) {
+  this[0] = promise;
+  this[1] = isYieldStar;
+};
+var __asyncGenerator = (__this, __arguments, generator) => {
+  var resume = (k, v, yes, no) => {
+    try {
+      var x = generator[k](v), isAwait = (v = x.value) instanceof __await, done = x.done;
+      Promise.resolve(isAwait ? v[0] : v).then((y) => isAwait ? resume(k === "return" ? k : "next", v[1] ? { done: y.done, value: y.value } : y, yes, no) : yes({ value: y, done })).catch((e) => resume("throw", e, yes, no));
+    } catch (e) {
+      no(e);
+    }
+  }, method = (k) => it[k] = (x) => new Promise((yes, no) => resume(k, x, yes, no)), it = {};
+  return generator = generator.apply(__this, __arguments), it[__knownSymbol("asyncIterator")] = () => it, method("next"), method("throw"), method("return"), it;
+};
+var __forAwait = (obj, it, method) => (it = obj[__knownSymbol("asyncIterator")]) ? it.call(obj) : (obj = obj[__knownSymbol("iterator")](), it = {}, method = (key, fn) => (fn = obj[key]) && (it[key] = (arg) => new Promise((yes, no, done) => (arg = fn.call(obj, arg), done = arg.done, Promise.resolve(arg.value).then((value) => yes({ value, done }), no)))), method("next"), method("return"), it);
+/**
+ * @license
+ * Copyright 2025 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+import { createPartFromText, FinishReason, GoogleGenAI } from "@google/genai";
+import { isBrowser } from "../utils/env_aware_utils.js";
+import { logger } from "../utils/logger.js";
+import { GoogleLLMVariant } from "../utils/variant_utils.js";
+import { version } from "../version.js";
+import { BaseLlm } from "./base_llm.js";
+import { GeminiLlmConnection } from "./gemini_llm_connection.js";
+import { createLlmResponse } from "./llm_response.js";
+const AGENT_ENGINE_TELEMETRY_TAG = "remote_reasoning_engine";
+const AGENT_ENGINE_TELEMETRY_ENV_VARIABLE_NAME = "GOOGLE_CLOUD_AGENT_ENGINE_ID";
+class Gemini extends BaseLlm {
+  /**
+   * @param params The parameters for creating a Gemini instance.
+   */
+  constructor({
+    model = "gemini-2.5-flash",
+    apiKey,
+    vertexai,
+    project,
+    location,
+    headers
+  } = {}) {
+    super(model);
+    this.project = project;
+    this.location = location;
+    this.apiKey = apiKey;
+    this.headers = headers;
+    const canReadEnv = typeof process === "object";
+    this.vertexai = !!vertexai;
+    if (!this.vertexai && canReadEnv) {
+      const vertexAIfromEnv = process.env["GOOGLE_GENAI_USE_VERTEXAI"];
+      if (vertexAIfromEnv) {
+        this.vertexai = vertexAIfromEnv.toLowerCase() === "true" || vertexAIfromEnv === "1";
+      }
+    }
+    if (this.vertexai) {
+      if (canReadEnv && !this.project) {
+        this.project = process.env["GOOGLE_CLOUD_PROJECT"];
+      }
+      if (canReadEnv && !this.location) {
+        this.location = process.env["GOOGLE_CLOUD_LOCATION"];
+      }
+      if (!this.project) {
+        throw new Error(
+          "VertexAI project must be provided via constructor or GOOGLE_CLOUD_PROJECT environment variable."
+        );
+      }
+      if (!this.location) {
+        throw new Error(
+          "VertexAI location must be provided via constructor or GOOGLE_CLOUD_LOCATION environment variable."
+        );
+      }
+    } else {
+      if (!this.apiKey && canReadEnv) {
+        this.apiKey = process.env["GOOGLE_GENAI_API_KEY"] || process.env["GEMINI_API_KEY"];
+      }
+      if (!this.apiKey) {
+        throw new Error(
+          "API key must be provided via constructor or GOOGLE_GENAI_API_KEY or GEMINI_API_KEY environment variable."
+        );
+      }
+    }
+  }
+  /**
+   * Sends a request to the Gemini model.
+   *
+   * @param llmRequest LlmRequest, the request to send to the Gemini model.
+   * @param stream bool = false, whether to do streaming call.
+   * @yields LlmResponse: The model response.
+   */
+  generateContentAsync(llmRequest, stream = false) {
+    return __asyncGenerator(this, null, function* () {
+      var _a, _b, _c, _d, _e, _f, _g;
+      this.preprocessRequest(llmRequest);
+      this.maybeAppendUserContent(llmRequest);
+      logger.info(
+        "Sending out request, model: ".concat(llmRequest.model, ", backend: ").concat(this.apiBackend, ", stream: ").concat(stream)
+      );
+      if ((_a = llmRequest.config) == null ? void 0 : _a.httpOptions) {
+        llmRequest.config.httpOptions.headers = __spreadValues(__spreadValues({}, llmRequest.config.httpOptions.headers), this.trackingHeaders);
+      }
+      if (stream) {
+        const streamResult = yield new __await(this.apiClient.models.generateContentStream({
+          model: (_b = llmRequest.model) != null ? _b : this.model,
+          contents: llmRequest.contents,
+          config: llmRequest.config
+        }));
+        let thoughtText = "";
+        let text = "";
+        let usageMetadata;
+        let lastResponse;
+        try {
+          for (var iter = __forAwait(streamResult), more, temp, error; more = !(temp = yield new __await(iter.next())).done; more = false) {
+            const response = temp.value;
+            lastResponse = response;
+            const llmResponse = createLlmResponse(response);
+            usageMetadata = llmResponse.usageMetadata;
+            const firstPart = (_d = (_c = llmResponse.content) == null ? void 0 : _c.parts) == null ? void 0 : _d[0];
+            if (firstPart == null ? void 0 : firstPart.text) {
+              if ("thought" in firstPart && firstPart.thought) {
+                thoughtText += firstPart.text;
+              } else {
+                text += firstPart.text;
+              }
+              llmResponse.partial = true;
+            } else if ((thoughtText || text) && (!firstPart || !firstPart.inlineData)) {
+              const parts = [];
+              if (thoughtText) {
+                parts.push({ text: thoughtText, thought: true });
+              }
+              if (text) {
+                parts.push(createPartFromText(text));
+              }
+              yield {
+                content: {
+                  role: "model",
+                  parts
+                },
+                usageMetadata: llmResponse.usageMetadata
+              };
+              thoughtText = "";
+              text = "";
+            }
+            yield llmResponse;
+          }
+        } catch (temp) {
+          error = [temp];
+        } finally {
+          try {
+            more && (temp = iter.return) && (yield new __await(temp.call(iter)));
+          } finally {
+            if (error)
+              throw error[0];
+          }
+        }
+        if ((text || thoughtText) && ((_f = (_e = lastResponse == null ? void 0 : lastResponse.candidates) == null ? void 0 : _e[0]) == null ? void 0 : _f.finishReason) === FinishReason.STOP) {
+          const parts = [];
+          if (thoughtText) {
+            parts.push({ text: thoughtText, thought: true });
+          }
+          if (text) {
+            parts.push({ text });
+          }
+          yield {
+            content: {
+              role: "model",
+              parts
+            },
+            usageMetadata
+          };
+        }
+      } else {
+        const response = yield new __await(this.apiClient.models.generateContent({
+          model: (_g = llmRequest.model) != null ? _g : this.model,
+          contents: llmRequest.contents,
+          config: llmRequest.config
+        }));
+        yield createLlmResponse(response);
+      }
+    });
+  }
+  get apiClient() {
+    if (this._apiClient) {
+      return this._apiClient;
+    }
+    const combinedHeaders = __spreadValues(__spreadValues({}, this.trackingHeaders), this.headers);
+    if (this.vertexai) {
+      this._apiClient = new GoogleGenAI({
+        vertexai: this.vertexai,
+        project: this.project,
+        location: this.location,
+        httpOptions: { headers: combinedHeaders }
+      });
+    } else {
+      this._apiClient = new GoogleGenAI({
+        apiKey: this.apiKey,
+        httpOptions: { headers: combinedHeaders }
+      });
+    }
+    return this._apiClient;
+  }
+  get apiBackend() {
+    if (!this._apiBackend) {
+      this._apiBackend = this.apiClient.vertexai ? GoogleLLMVariant.VERTEX_AI : GoogleLLMVariant.GEMINI_API;
+    }
+    return this._apiBackend;
+  }
+  get trackingHeaders() {
+    if (!this._trackingHeaders) {
+      let frameworkLabel = "google-adk/".concat(version);
+      if (!isBrowser() && process.env[AGENT_ENGINE_TELEMETRY_ENV_VARIABLE_NAME]) {
+        frameworkLabel = "".concat(frameworkLabel, "+").concat(AGENT_ENGINE_TELEMETRY_TAG);
+      }
+      const languageLabel = "gl-typescript/".concat(isBrowser() ? window.navigator.userAgent : process.version);
+      const versionHeaderValue = "".concat(frameworkLabel, " ").concat(languageLabel);
+      this._trackingHeaders = {
+        "x-goog-api-client": versionHeaderValue,
+        "user-agent": versionHeaderValue
+      };
+    }
+    return this._trackingHeaders;
+  }
+  get liveApiVersion() {
+    if (!this._liveApiVersion) {
+      this._liveApiVersion = this.apiBackend === GoogleLLMVariant.VERTEX_AI ? "v1beta1" : "v1alpha";
+    }
+    return this._liveApiVersion;
+  }
+  get liveApiClient() {
+    if (!this._liveApiClient) {
+      this._liveApiClient = new GoogleGenAI({
+        apiKey: this.apiKey,
+        httpOptions: {
+          headers: this.trackingHeaders,
+          apiVersion: this.liveApiVersion
+        }
+      });
+    }
+    return this._liveApiClient;
+  }
+  /**
+   * Connects to the Gemini model and returns an llm connection.
+   *
+   * @param llmRequest LlmRequest, the request to send to the Gemini model.
+   * @returns BaseLlmConnection, the connection to the Gemini model.
+   */
+  async connect(llmRequest) {
+    var _a, _b, _c, _d;
+    if ((_a = llmRequest.liveConnectConfig) == null ? void 0 : _a.httpOptions) {
+      if (!llmRequest.liveConnectConfig.httpOptions.headers) {
+        llmRequest.liveConnectConfig.httpOptions.headers = {};
+      }
+      Object.assign(
+        llmRequest.liveConnectConfig.httpOptions.headers,
+        this.trackingHeaders
+      );
+      llmRequest.liveConnectConfig.httpOptions.apiVersion = this.liveApiVersion;
+    }
+    if ((_b = llmRequest.config) == null ? void 0 : _b.systemInstruction) {
+      llmRequest.liveConnectConfig.systemInstruction = {
+        role: "system",
+        // TODO - b/425992518: validate type casting works well.
+        parts: [createPartFromText(llmRequest.config.systemInstruction)]
+      };
+    }
+    llmRequest.liveConnectConfig.tools = (_c = llmRequest.config) == null ? void 0 : _c.tools;
+    const liveSession = await this.liveApiClient.live.connect({
+      model: (_d = llmRequest.model) != null ? _d : this.model,
+      config: llmRequest.liveConnectConfig,
+      callbacks: {
+        // TODO - b/425992518: GenAI SDK inconsistent API, missing methods.
+        onmessage: () => {
+        }
+      }
+    });
+    return new GeminiLlmConnection(liveSession);
+  }
+  preprocessRequest(llmRequest) {
+    if (this.apiBackend === GoogleLLMVariant.GEMINI_API) {
+      if (llmRequest.config) {
+        llmRequest.config.labels = void 0;
+      }
+      if (llmRequest.contents) {
+        for (const content of llmRequest.contents) {
+          if (!content.parts) continue;
+          for (const part of content.parts) {
+            removeDisplayNameIfPresent(part.inlineData);
+            removeDisplayNameIfPresent(part.fileData);
+          }
+        }
+      }
+    }
+  }
+}
+/**
+ * A list of model name patterns that are supported by this LLM.
+ *
+ * @returns A list of supported models.
+ */
+Gemini.supportedModels = [
+  /gemini-.*/,
+  // fine-tuned vertex endpoint pattern
+  /projects\/.+\/locations\/.+\/endpoints\/.+/,
+  // vertex gemini long name
+  /projects\/.+\/locations\/.+\/publishers\/google\/models\/gemini.+/
+];
+function removeDisplayNameIfPresent(dataObj) {
+  if (dataObj && dataObj.displayName) {
+    dataObj.displayName = void 0;
+  }
+}
+export {
+  Gemini
+};

package/dist/web/models/llm_request.js

@@ -0,0 +1,50 @@
+/**
+ * @license
+ * Copyright 2025 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+function appendInstructions(llmRequest, instructions) {
+  if (!llmRequest.config) {
+    llmRequest.config = {};
+  }
+  const newInstructions = instructions.join("\n\n");
+  if (llmRequest.config.systemInstruction) {
+    llmRequest.config.systemInstruction += "\n\n" + newInstructions;
+  } else {
+    llmRequest.config.systemInstruction = newInstructions;
+  }
+}
+function appendTools(llmRequest, tools) {
+  if (!(tools == null ? void 0 : tools.length)) {
+    return;
+  }
+  const functionDeclarations = [];
+  for (const tool of tools) {
+    const declaration = tool._getDeclaration();
+    if (declaration) {
+      functionDeclarations.push(declaration);
+      llmRequest.toolsDict[tool.name] = tool;
+    }
+  }
+  if (functionDeclarations.length) {
+    if (!llmRequest.config) {
+      llmRequest.config = {};
+    }
+    if (!llmRequest.config.tools) {
+      llmRequest.config.tools = [];
+    }
+    llmRequest.config.tools.push({ functionDeclarations });
+  }
+}
+function setOutputSchema(llmRequest, schema) {
+  if (!llmRequest.config) {
+    llmRequest.config = {};
+  }
+  llmRequest.config.responseSchema = schema;
+  llmRequest.config.responseMimeType = "application/json";
+}
+export {
+  appendInstructions,
+  appendTools,
+  setOutputSchema
+};

package/dist/web/models/llm_response.js

@@ -0,0 +1,41 @@
+/**
+ * @license
+ * Copyright 2025 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+function createLlmResponse(response) {
+  var _a;
+  const usageMetadata = response.usageMetadata;
+  if (response.candidates && response.candidates.length > 0) {
+    const candidate = response.candidates[0];
+    if (((_a = candidate.content) == null ? void 0 : _a.parts) && candidate.content.parts.length > 0) {
+      return {
+        content: candidate.content,
+        groundingMetadata: candidate.groundingMetadata,
+        usageMetadata,
+        finishReason: candidate.finishReason
+      };
+    }
+    return {
+      errorCode: candidate.finishReason,
+      errorMessage: candidate.finishMessage,
+      usageMetadata,
+      finishReason: candidate.finishReason
+    };
+  }
+  if (response.promptFeedback) {
+    return {
+      errorCode: response.promptFeedback.blockReason,
+      errorMessage: response.promptFeedback.blockReasonMessage,
+      usageMetadata
+    };
+  }
+  return {
+    errorCode: "UNKNOWN_ERROR",
+    errorMessage: "Unknown error.",
+    usageMetadata
+  };
+}
+export {
+  createLlmResponse
+};