@elizaos/plugin-google-genai 2.0.0-beta.1 → 2.0.3-beta.2

package/dist/cjs/index.node.cjs

@@ -1,840 +0,0 @@
-var __create = Object.create;
-var __getProtoOf = Object.getPrototypeOf;
-var __defProp = Object.defineProperty;
-var __getOwnPropNames = Object.getOwnPropertyNames;
-var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
-var __hasOwnProp = Object.prototype.hasOwnProperty;
-function __accessProp(key) {
-  return this[key];
-}
-var __toESMCache_node;
-var __toESMCache_esm;
-var __toESM = (mod, isNodeMode, target) => {
-  var canCache = mod != null && typeof mod === "object";
-  if (canCache) {
-    var cache = isNodeMode ? __toESMCache_node ??= new WeakMap : __toESMCache_esm ??= new WeakMap;
-    var cached = cache.get(mod);
-    if (cached)
-      return cached;
-  }
-  target = mod != null ? __create(__getProtoOf(mod)) : {};
-  const to = isNodeMode || !mod || !mod.__esModule ? __defProp(target, "default", { value: mod, enumerable: true }) : target;
-  for (let key of __getOwnPropNames(mod))
-    if (!__hasOwnProp.call(to, key))
-      __defProp(to, key, {
-        get: __accessProp.bind(mod, key),
-        enumerable: true
-      });
-  if (canCache)
-    cache.set(mod, to);
-  return to;
-};
-var __toCommonJS = (from) => {
-  var entry = (__moduleCache ??= new WeakMap).get(from), desc;
-  if (entry)
-    return entry;
-  entry = __defProp({}, "__esModule", { value: true });
-  if (from && typeof from === "object" || typeof from === "function") {
-    for (var key of __getOwnPropNames(from))
-      if (!__hasOwnProp.call(entry, key))
-        __defProp(entry, key, {
-          get: __accessProp.bind(from, key),
-          enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable
-        });
-  }
-  __moduleCache.set(from, entry);
-  return entry;
-};
-var __moduleCache;
-var __returnValue = (v) => v;
-function __exportSetter(name, newValue) {
-  this[name] = __returnValue.bind(null, newValue);
-}
-var __export = (target, all) => {
-  for (var name in all)
-    __defProp(target, name, {
-      get: all[name],
-      enumerable: true,
-      configurable: true,
-      set: __exportSetter.bind(all, name)
-    });
-};
-
-// index.node.ts
-var exports_index_node = {};
-__export(exports_index_node, {
-  googleGenAIPlugin: () => googleGenAIPlugin,
-  default: () => index_node_default
-});
-module.exports = __toCommonJS(exports_index_node);
-
-// index.ts
-var ElizaCore3 = __toESM(require("@elizaos/core"));
-var import_core6 = require("@elizaos/core");
-var import_genai3 = require("@google/genai");
-
-// init.ts
-var import_core2 = require("@elizaos/core");
-var import_genai2 = require("@google/genai");
-
-// utils/config.ts
-var import_core = require("@elizaos/core");
-var import_genai = require("@google/genai");
-function getEnvValue(key) {
-  if (typeof process === "undefined") {
-    return;
-  }
-  const value = process.env[key];
-  return value === undefined ? undefined : String(value);
-}
-function getSetting(runtime, key, defaultValue) {
-  const runtimeValue = runtime.getSetting(key);
-  if (runtimeValue !== undefined) {
-    return String(runtimeValue);
-  }
-  return getEnvValue(key) ?? defaultValue;
-}
-function getApiKey(runtime) {
-  return getSetting(runtime, "GOOGLE_GENERATIVE_AI_API_KEY");
-}
-function getSmallModel(runtime) {
-  return getSetting(runtime, "GOOGLE_SMALL_MODEL") ?? getSetting(runtime, "SMALL_MODEL", "gemini-2.0-flash-001") ?? "gemini-2.0-flash-001";
-}
-function getNanoModel(runtime) {
-  return getSetting(runtime, "GOOGLE_NANO_MODEL") ?? getSetting(runtime, "NANO_MODEL") ?? getSmallModel(runtime);
-}
-function getMediumModel(runtime) {
-  return getSetting(runtime, "GOOGLE_MEDIUM_MODEL") ?? getSetting(runtime, "MEDIUM_MODEL") ?? getSmallModel(runtime);
-}
-function getLargeModel(runtime) {
-  return getSetting(runtime, "GOOGLE_LARGE_MODEL") ?? getSetting(runtime, "LARGE_MODEL", "gemini-2.5-pro-preview-03-25") ?? "gemini-2.5-pro-preview-03-25";
-}
-function getMegaModel(runtime) {
-  return getSetting(runtime, "GOOGLE_MEGA_MODEL") ?? getSetting(runtime, "MEGA_MODEL") ?? getLargeModel(runtime);
-}
-function getResponseHandlerModel(runtime) {
-  return getSetting(runtime, "GOOGLE_RESPONSE_HANDLER_MODEL") ?? getSetting(runtime, "GOOGLE_SHOULD_RESPOND_MODEL") ?? getSetting(runtime, "RESPONSE_HANDLER_MODEL") ?? getSetting(runtime, "SHOULD_RESPOND_MODEL") ?? getNanoModel(runtime);
-}
-function getActionPlannerModel(runtime) {
-  return getSetting(runtime, "GOOGLE_ACTION_PLANNER_MODEL") ?? getSetting(runtime, "GOOGLE_PLANNER_MODEL") ?? getSetting(runtime, "ACTION_PLANNER_MODEL") ?? getSetting(runtime, "PLANNER_MODEL") ?? getMediumModel(runtime);
-}
-function getImageModel(runtime) {
-  return getSetting(runtime, "GOOGLE_IMAGE_MODEL") ?? getSetting(runtime, "IMAGE_MODEL", "gemini-2.5-pro-preview-03-25") ?? "gemini-2.5-pro-preview-03-25";
-}
-function getEmbeddingModel(runtime) {
-  return getSetting(runtime, "GOOGLE_EMBEDDING_MODEL", "text-embedding-004") ?? "text-embedding-004";
-}
-function createGoogleGenAI(runtime) {
-  const apiKey = getApiKey(runtime);
-  if (!apiKey) {
-    import_core.logger.error("Google Generative AI API Key is missing");
-    return null;
-  }
-  return new import_genai.GoogleGenAI({ apiKey });
-}
-function getSafetySettings() {
-  return [
-    {
-      category: import_genai.HarmCategory.HARM_CATEGORY_HARASSMENT,
-      threshold: import_genai.HarmBlockThreshold.BLOCK_MEDIUM_AND_ABOVE
-    },
-    {
-      category: import_genai.HarmCategory.HARM_CATEGORY_HATE_SPEECH,
-      threshold: import_genai.HarmBlockThreshold.BLOCK_MEDIUM_AND_ABOVE
-    },
-    {
-      category: import_genai.HarmCategory.HARM_CATEGORY_SEXUALLY_EXPLICIT,
-      threshold: import_genai.HarmBlockThreshold.BLOCK_MEDIUM_AND_ABOVE
-    },
-    {
-      category: import_genai.HarmCategory.HARM_CATEGORY_DANGEROUS_CONTENT,
-      threshold: import_genai.HarmBlockThreshold.BLOCK_MEDIUM_AND_ABOVE
-    }
-  ];
-}
-
-// init.ts
-function initializeGoogleGenAI(_config, runtime) {
-  (async () => {
-    try {
-      const apiKey = getApiKey(runtime);
-      if (!apiKey) {
-        import_core2.logger.warn("GOOGLE_GENERATIVE_AI_API_KEY is not set");
-        return;
-      }
-      const genAI = new import_genai2.GoogleGenAI({ apiKey });
-      const modelList = await genAI.models.list();
-      const models = [];
-      for await (const model of modelList) {
-        models.push(model);
-      }
-      import_core2.logger.log(`Google AI API key validated. Available models: ${models.length}`);
-    } catch (error) {
-      import_core2.logger.warn(`Google AI configuration error: ${error instanceof Error ? error.message : String(error)}`);
-    }
-  })();
-}
-
-// models/embedding.ts
-var ElizaCore = __toESM(require("@elizaos/core"));
-var import_core3 = require("@elizaos/core");
-
-// utils/events.ts
-var MODEL_USED_EVENT = "MODEL_USED";
-function emitModelUsageEvent(runtime, type, _prompt, usage) {
-  runtime.emitEvent(MODEL_USED_EVENT, {
-    runtime,
-    source: "plugin-google-genai",
-    type,
-    tokens: {
-      prompt: usage.promptTokens,
-      completion: usage.completionTokens,
-      total: usage.totalTokens
-    }
-  });
-}
-
-// utils/tokenization.ts
-async function countTokens(text) {
-  return Math.ceil(text.length / 4);
-}
-
-// models/embedding.ts
-var TEXT_EMBEDDING_MODEL_TYPE = ElizaCore.ModelType?.TEXT_EMBEDDING ?? "TEXT_EMBEDDING";
-async function handleTextEmbedding(runtime, params) {
-  const genAI = createGoogleGenAI(runtime);
-  if (!genAI) {
-    throw new Error("Google Generative AI client not initialized");
-  }
-  const embeddingModelName = getEmbeddingModel(runtime);
-  import_core3.logger.debug(`[TEXT_EMBEDDING] Using model: ${embeddingModelName}`);
-  if (params === null) {
-    return Array(768).fill(0);
-  }
-  let text = typeof params === "string" ? params : typeof params === "object" && params.text ? params.text : "";
-  if (!text.trim()) {
-    import_core3.logger.warn("Empty text for embedding");
-    return Array(768).fill(0);
-  }
-  const maxChars = 8192 * 4;
-  if (text.length > maxChars) {
-    import_core3.logger.warn(`[Google GenAI] Embedding input too long (~${Math.ceil(text.length / 4)} tokens), truncating to ~8192 tokens`);
-    text = text.slice(0, maxChars);
-  }
-  try {
-    const response = await genAI.models.embedContent({
-      model: embeddingModelName,
-      contents: text
-    });
-    const embedding = response.embeddings?.[0]?.values || [];
-    const promptTokens = await countTokens(text);
-    emitModelUsageEvent(runtime, TEXT_EMBEDDING_MODEL_TYPE, text, {
-      promptTokens,
-      completionTokens: 0,
-      totalTokens: promptTokens
-    });
-    import_core3.logger.log(`Got embedding with length ${embedding.length}`);
-    return embedding;
-  } catch (error) {
-    import_core3.logger.error(`Error generating embedding: ${error instanceof Error ? error.message : String(error)}`);
-    return Array(768).fill(0);
-  }
-}
-// models/image.ts
-var import_core4 = require("@elizaos/core");
-async function handleImageDescription(runtime, params) {
-  const genAI = createGoogleGenAI(runtime);
-  if (!genAI) {
-    throw new Error("Google Generative AI client not initialized");
-  }
-  let imageUrl;
-  let promptText;
-  const modelName = getImageModel(runtime);
-  import_core4.logger.log(`[IMAGE_DESCRIPTION] Using model: ${modelName}`);
-  if (typeof params === "string") {
-    imageUrl = params;
-    promptText = "Please analyze this image and provide a title and detailed description.";
-  } else {
-    imageUrl = params.imageUrl;
-    promptText = params.prompt || "Please analyze this image and provide a title and detailed description.";
-  }
-  try {
-    const imageResponse = await fetch(imageUrl);
-    if (!imageResponse.ok) {
-      throw new Error(`Failed to fetch image: ${imageResponse.statusText}`);
-    }
-    const imageData = await imageResponse.arrayBuffer();
-    const base64Image = Buffer.from(imageData).toString("base64");
-    const contentType = imageResponse.headers.get("content-type") || "image/jpeg";
-    const details = {
-      model: modelName,
-      systemPrompt: "",
-      userPrompt: promptText,
-      temperature: 0.7,
-      maxTokens: 8192,
-      purpose: "external_llm",
-      actionType: "google-genai.IMAGE_DESCRIPTION.generateContent"
-    };
-    const response = await import_core4.recordLlmCall(runtime, details, async () => {
-      const result = await genAI.models.generateContent({
-        model: modelName,
-        contents: [
-          {
-            role: "user",
-            parts: [
-              { text: promptText },
-              {
-                inlineData: {
-                  mimeType: contentType,
-                  data: base64Image
-                }
-              }
-            ]
-          }
-        ],
-        config: {
-          temperature: 0.7,
-          topK: 40,
-          topP: 0.95,
-          maxOutputTokens: 8192,
-          safetySettings: getSafetySettings()
-        }
-      });
-      const responseText2 = result.text || "";
-      details.response = responseText2;
-      details.promptTokens = await countTokens(promptText);
-      details.completionTokens = await countTokens(responseText2);
-      return result;
-    });
-    const responseText = response.text || "";
-    try {
-      const jsonResponse = JSON.parse(responseText);
-      if (typeof jsonResponse.title === "string" && typeof jsonResponse.description === "string") {
-        return {
-          title: jsonResponse.title,
-          description: jsonResponse.description
-        };
-      }
-    } catch {}
-    const titleMatch = responseText.match(/title[:\s]+(.+?)(?:\n|$)/i);
-    const title = titleMatch?.[1]?.trim() || "Image Analysis";
-    const description = titleMatch ? responseText.replace(/title[:\s]+(.+?)(?:\n|$)/i, "").trim() : responseText.trim();
-    return { title, description };
-  } catch (error) {
-    const message = error instanceof Error ? error.message : String(error);
-    import_core4.logger.error(`Error analyzing image: ${message}`);
-    return {
-      title: "Failed to analyze image",
-      description: `Error: ${message}`
-    };
-  }
-}
-// models/text.ts
-var ElizaCore2 = __toESM(require("@elizaos/core"));
-var import_core5 = require("@elizaos/core");
-var CORE_MODEL_TYPES = ElizaCore2.ModelType ?? {};
-var TEXT_NANO_MODEL_TYPE = CORE_MODEL_TYPES.TEXT_NANO ?? "TEXT_NANO";
-var TEXT_MEDIUM_MODEL_TYPE = CORE_MODEL_TYPES.TEXT_MEDIUM ?? "TEXT_MEDIUM";
-var TEXT_SMALL_MODEL_TYPE = CORE_MODEL_TYPES.TEXT_SMALL ?? "TEXT_SMALL";
-var TEXT_LARGE_MODEL_TYPE = CORE_MODEL_TYPES.TEXT_LARGE ?? "TEXT_LARGE";
-var TEXT_MEGA_MODEL_TYPE = CORE_MODEL_TYPES.TEXT_MEGA ?? "TEXT_MEGA";
-var RESPONSE_HANDLER_MODEL_TYPE = CORE_MODEL_TYPES.RESPONSE_HANDLER ?? "RESPONSE_HANDLER";
-var ACTION_PLANNER_MODEL_TYPE = CORE_MODEL_TYPES.ACTION_PLANNER ?? "ACTION_PLANNER";
-function normalizeToolsForGoogle(tools) {
-  if (!tools)
-    return;
-  if (Array.isArray(tools) && tools.length > 0 && typeof tools[0] === "object" && tools[0] !== null && "functionDeclarations" in tools[0]) {
-    return tools;
-  }
-  const flat = Array.isArray(tools) ? tools : Object.entries(tools).map(([name, value]) => ({ name, ...value }));
-  const declarations = [];
-  for (const tool of flat) {
-    const name = tool.name ?? tool.function?.name;
-    if (!name) {
-      throw new Error("[GoogleGenAI] Tool definition is missing a name.");
-    }
-    const description = tool.description ?? tool.function?.description;
-    const parameters = tool.parameters ?? tool.inputSchema ?? tool.function?.parameters ?? {
-      type: "object",
-      properties: {}
-    };
-    declarations.push({
-      name,
-      ...description ? { description } : {},
-      parameters
-    });
-  }
-  return declarations.length > 0 ? [{ functionDeclarations: declarations }] : undefined;
-}
-function normalizeToolConfigForGoogle(toolChoice) {
-  if (!toolChoice)
-    return;
-  if (toolChoice === "auto") {
-    return { functionCallingConfig: { mode: "AUTO" } };
-  }
-  if (toolChoice === "required") {
-    return { functionCallingConfig: { mode: "ANY" } };
-  }
-  if (toolChoice === "none") {
-    return { functionCallingConfig: { mode: "NONE" } };
-  }
-  let toolName;
-  if ("type" in toolChoice) {
-    toolName = toolChoice.type === "function" ? toolChoice.function.name : toolChoice.toolName ?? toolChoice.name;
-  } else {
-    toolName = toolChoice.name;
-  }
-  if (toolName) {
-    return {
-      functionCallingConfig: {
-        mode: "ANY",
-        allowedFunctionNames: [toolName]
-      }
-    };
-  }
-  return;
-}
-function resolveResponseJsonSchema(responseSchema) {
-  if (!responseSchema)
-    return;
-  if ("schema" in responseSchema && responseSchema.schema) {
-    return responseSchema.schema;
-  }
-  return responseSchema;
-}
-function buildPromptParts(prompt, attachments) {
-  const parts = [{ text: prompt }];
-  for (const attachment of attachments ?? []) {
-    if (attachment.data instanceof URL) {
-      parts.push({
-        fileData: {
-          mimeType: attachment.mediaType,
-          fileUri: attachment.data.toString()
-        }
-      });
-      continue;
-    }
-    if (typeof attachment.data === "string" && /^https?:\/\//i.test(attachment.data)) {
-      parts.push({
-        fileData: {
-          mimeType: attachment.mediaType,
-          fileUri: attachment.data
-        }
-      });
-      continue;
-    }
-    if (typeof attachment.data === "string") {
-      const dataUrlMatch = attachment.data.match(/^data:([^;,]+);base64,(.+)$/i);
-      parts.push({
-        inlineData: {
-          mimeType: dataUrlMatch?.[1] ?? attachment.mediaType,
-          data: dataUrlMatch?.[2] ?? attachment.data
-        }
-      });
-      continue;
-    }
-    parts.push({
-      inlineData: {
-        mimeType: attachment.mediaType,
-        data: Buffer.from(attachment.data).toString("base64")
-      }
-    });
-  }
-  return parts;
-}
-function resolveGoogleSystemInstruction(runtime, params) {
-  return import_core5.resolveEffectiveSystemPrompt({
-    params,
-    fallback: import_core5.buildCanonicalSystemPrompt({ character: runtime.character })
-  });
-}
-function resolveGooglePrompt(params, systemInstruction) {
-  return import_core5.renderChatMessagesForPrompt(params.messages, {
-    omitDuplicateSystem: systemInstruction
-  }) ?? params.prompt;
-}
-function getModelNameForType(runtime, modelType) {
-  switch (modelType) {
-    case TEXT_NANO_MODEL_TYPE:
-      return getNanoModel(runtime);
-    case TEXT_MEDIUM_MODEL_TYPE:
-      return getMediumModel(runtime);
-    case TEXT_SMALL_MODEL_TYPE:
-      return getSmallModel(runtime);
-    case TEXT_LARGE_MODEL_TYPE:
-      return getLargeModel(runtime);
-    case TEXT_MEGA_MODEL_TYPE:
-      return getMegaModel(runtime);
-    case RESPONSE_HANDLER_MODEL_TYPE:
-      return getResponseHandlerModel(runtime);
-    case ACTION_PLANNER_MODEL_TYPE:
-      return getActionPlannerModel(runtime);
-    default:
-      return getLargeModel(runtime);
-  }
-}
-function buildGoogleGenerationConfig(params, systemInstruction, temperature, maxTokens, stopSequences) {
-  const tools = normalizeToolsForGoogle(params.tools);
-  const toolConfig = normalizeToolConfigForGoogle(params.toolChoice);
-  const responseJsonSchema = resolveResponseJsonSchema(params.responseSchema);
-  const baseConfig = {
-    temperature,
-    topK: 40,
-    topP: 0.95,
-    maxOutputTokens: maxTokens,
-    stopSequences,
-    safetySettings: getSafetySettings(),
-    ...systemInstruction && { systemInstruction },
-    ...tools ? { tools } : {},
-    ...toolConfig ? { toolConfig } : {},
-    ...responseJsonSchema ? {
-      responseMimeType: "application/json",
-      responseJsonSchema
-    } : {}
-  };
-  return baseConfig;
-}
-function createLlmCallDetails(modelName, modelType, prompt, systemInstruction, temperature, maxTokens) {
-  return {
-    model: modelName,
-    systemPrompt: systemInstruction ?? "",
-    userPrompt: prompt,
-    temperature,
-    maxTokens,
-    purpose: "external_llm",
-    actionType: `google-genai.${modelType}.generateContent`
-  };
-}
-async function generateContentWithTrajectory(runtime, genAI, modelName, modelType, prompt, systemInstruction, temperature, maxTokens, request) {
-  const details = createLlmCallDetails(modelName, modelType, prompt, systemInstruction, temperature, maxTokens);
-  const response = await import_core5.recordLlmCall(runtime, details, async () => {
-    const result = await genAI.models.generateContent(request);
-    const text2 = result.text || "";
-    details.response = text2;
-    details.promptTokens = await countTokens(prompt);
-    details.completionTokens = await countTokens(text2);
-    return result;
-  });
-  const text = response.text || "";
-  const promptTokens = details.promptTokens ?? await countTokens(prompt);
-  const completionTokens = details.completionTokens ?? await countTokens(text);
-  emitModelUsageEvent(runtime, modelType, prompt, {
-    promptTokens,
-    completionTokens,
-    totalTokens: promptTokens + completionTokens
-  });
-  return text;
-}
-async function handleTextSmall(runtime, params) {
-  const {
-    stopSequences = [],
-    maxTokens = 8192,
-    temperature = 0.7,
-    attachments
-  } = params;
-  const genAI = createGoogleGenAI(runtime);
-  if (!genAI) {
-    throw new Error("Google Generative AI client not initialized");
-  }
-  const modelName = getModelNameForType(runtime, TEXT_SMALL_MODEL_TYPE);
-  import_core5.logger.log(`[TEXT_SMALL] Using model: ${modelName}`);
-  try {
-    const systemInstruction = resolveGoogleSystemInstruction(runtime, params);
-    const promptText = resolveGooglePrompt(params, systemInstruction);
-    return await generateContentWithTrajectory(runtime, genAI, modelName, TEXT_SMALL_MODEL_TYPE, promptText, systemInstruction, temperature, maxTokens, {
-      model: modelName,
-      contents: (attachments?.length ?? 0) > 0 ? [
-        {
-          role: "user",
-          parts: buildPromptParts(promptText, attachments)
-        }
-      ] : promptText,
-      config: buildGoogleGenerationConfig(params, systemInstruction, temperature, maxTokens, stopSequences)
-    });
-  } catch (error) {
-    import_core5.logger.error(`[TEXT_SMALL] Error: ${error instanceof Error ? error.message : String(error)}`);
-    throw error;
-  }
-}
-async function handleTextLarge(runtime, params) {
-  const {
-    stopSequences = [],
-    maxTokens = 8192,
-    temperature = 0.7,
-    attachments
-  } = params;
-  const genAI = createGoogleGenAI(runtime);
-  if (!genAI) {
-    throw new Error("Google Generative AI client not initialized");
-  }
-  const modelName = getModelNameForType(runtime, TEXT_LARGE_MODEL_TYPE);
-  import_core5.logger.log(`[TEXT_LARGE] Using model: ${modelName}`);
-  try {
-    const systemInstruction = resolveGoogleSystemInstruction(runtime, params);
-    const promptText = resolveGooglePrompt(params, systemInstruction);
-    return await generateContentWithTrajectory(runtime, genAI, modelName, TEXT_LARGE_MODEL_TYPE, promptText, systemInstruction, temperature, maxTokens, {
-      model: modelName,
-      contents: (attachments?.length ?? 0) > 0 ? [
-        {
-          role: "user",
-          parts: buildPromptParts(promptText, attachments)
-        }
-      ] : promptText,
-      config: buildGoogleGenerationConfig(params, systemInstruction, temperature, maxTokens, stopSequences)
-    });
-  } catch (error) {
-    import_core5.logger.error(`[TEXT_LARGE] Error: ${error instanceof Error ? error.message : String(error)}`);
-    throw error;
-  }
-}
-async function handleTextNano(runtime, params) {
-  return handleTextWithType(runtime, TEXT_NANO_MODEL_TYPE, params);
-}
-async function handleTextMedium(runtime, params) {
-  return handleTextWithType(runtime, TEXT_MEDIUM_MODEL_TYPE, params);
-}
-async function handleTextMega(runtime, params) {
-  return handleTextWithType(runtime, TEXT_MEGA_MODEL_TYPE, params);
-}
-async function handleResponseHandler(runtime, params) {
-  return handleTextWithType(runtime, RESPONSE_HANDLER_MODEL_TYPE, params);
-}
-async function handleActionPlanner(runtime, params) {
-  return handleTextWithType(runtime, ACTION_PLANNER_MODEL_TYPE, params);
-}
-async function handleTextWithType(runtime, modelType, params) {
-  const {
-    stopSequences = [],
-    maxTokens = 8192,
-    temperature = 0.7,
-    attachments
-  } = params;
-  const genAI = createGoogleGenAI(runtime);
-  if (!genAI) {
-    throw new Error("Google Generative AI client not initialized");
-  }
-  const modelName = getModelNameForType(runtime, modelType);
-  import_core5.logger.log(`[${modelType}] Using model: ${modelName}`);
-  try {
-    const systemInstruction = resolveGoogleSystemInstruction(runtime, params);
-    const promptText = resolveGooglePrompt(params, systemInstruction);
-    return await generateContentWithTrajectory(runtime, genAI, modelName, modelType, promptText, systemInstruction, temperature, maxTokens, {
-      model: modelName,
-      contents: (attachments?.length ?? 0) > 0 ? [
-        {
-          role: "user",
-          parts: buildPromptParts(promptText, attachments)
-        }
-      ] : promptText,
-      config: buildGoogleGenerationConfig(params, systemInstruction, temperature, maxTokens, stopSequences)
-    });
-  } catch (error) {
-    import_core5.logger.error(`[${modelType}] Error: ${error instanceof Error ? error.message : String(error)}`);
-    throw error;
-  }
-}
-// index.ts
-var CORE_MODEL_TYPES2 = ElizaCore3.ModelType ?? {};
-var TEXT_NANO_MODEL_TYPE2 = CORE_MODEL_TYPES2.TEXT_NANO ?? "TEXT_NANO";
-var TEXT_MEDIUM_MODEL_TYPE2 = CORE_MODEL_TYPES2.TEXT_MEDIUM ?? "TEXT_MEDIUM";
-var TEXT_SMALL_MODEL_TYPE2 = CORE_MODEL_TYPES2.TEXT_SMALL ?? "TEXT_SMALL";
-var TEXT_LARGE_MODEL_TYPE2 = CORE_MODEL_TYPES2.TEXT_LARGE ?? "TEXT_LARGE";
-var TEXT_EMBEDDING_MODEL_TYPE2 = CORE_MODEL_TYPES2.TEXT_EMBEDDING ?? "TEXT_EMBEDDING";
-var IMAGE_DESCRIPTION_MODEL_TYPE = CORE_MODEL_TYPES2.IMAGE_DESCRIPTION ?? "IMAGE_DESCRIPTION";
-var TEXT_MEGA_MODEL_TYPE2 = CORE_MODEL_TYPES2.TEXT_MEGA ?? "TEXT_MEGA";
-var RESPONSE_HANDLER_MODEL_TYPE2 = CORE_MODEL_TYPES2.RESPONSE_HANDLER ?? "RESPONSE_HANDLER";
-var ACTION_PLANNER_MODEL_TYPE2 = CORE_MODEL_TYPES2.ACTION_PLANNER ?? "ACTION_PLANNER";
-var pluginTests = [
-  {
-    name: "google_genai_plugin_tests",
-    tests: [
-      {
-        name: "google_test_api_key_validation",
-        fn: async (runtime) => {
-          const apiKey = getApiKey(runtime);
-          if (!apiKey) {
-            throw new Error("GOOGLE_GENERATIVE_AI_API_KEY not set");
-          }
-          const genAI = new import_genai3.GoogleGenAI({ apiKey });
-          const modelList = await genAI.models.list();
-          const models = [];
-          for await (const model of modelList) {
-            models.push(model);
-          }
-          import_core6.logger.log(`Available models: ${models.length}`);
-        }
-      },
-      {
-        name: "google_test_text_embedding",
-        fn: async (runtime) => {
-          try {
-            const embedding = await runtime.useModel(import_core6.ModelType.TEXT_EMBEDDING, {
-              text: "Hello, world!"
-            });
-            import_core6.logger.log(`Embedding dimension: ${embedding.length}`);
-            if (embedding.length === 0) {
-              throw new Error("Failed to generate embedding");
-            }
-          } catch (error) {
-            import_core6.logger.error(`Error in test_text_embedding: ${error instanceof Error ? error.message : String(error)}`);
-            throw error;
-          }
-        }
-      },
-      {
-        name: "google_test_text_small",
-        fn: async (runtime) => {
-          try {
-            const text = await runtime.useModel(import_core6.ModelType.TEXT_SMALL, {
-              prompt: "What is the nature of reality in 10 words?"
-            });
-            if (text.length === 0) {
-              throw new Error("Failed to generate text");
-            }
-            import_core6.logger.log("Generated with TEXT_SMALL:", text);
-          } catch (error) {
-            import_core6.logger.error(`Error in test_text_small: ${error instanceof Error ? error.message : String(error)}`);
-            throw error;
-          }
-        }
-      },
-      {
-        name: "google_test_text_large",
-        fn: async (runtime) => {
-          try {
-            const text = await runtime.useModel(import_core6.ModelType.TEXT_LARGE, {
-              prompt: "Explain quantum mechanics in simple terms."
-            });
-            if (text.length === 0) {
-              throw new Error("Failed to generate text");
-            }
-            import_core6.logger.log("Generated with TEXT_LARGE:", `${text.substring(0, 100)}...`);
-          } catch (error) {
-            import_core6.logger.error(`Error in test_text_large: ${error instanceof Error ? error.message : String(error)}`);
-            throw error;
-          }
-        }
-      },
-      {
-        name: "google_test_image_description",
-        fn: async (runtime) => {
-          try {
-            const result = await runtime.useModel(import_core6.ModelType.IMAGE_DESCRIPTION, "https://upload.wikimedia.org/wikipedia/commons/thumb/1/1c/Vitalik_Buterin_TechCrunch_London_2015_%28cropped%29.jpg/537px-Vitalik_Buterin_TechCrunch_London_2015_%28cropped%29.jpg");
-            if (result != null && typeof result === "object" && "title" in result && "description" in result) {
-              import_core6.logger.log("Image description:", JSON.stringify(result));
-            } else {
-              import_core6.logger.error(`Invalid image description result format: ${JSON.stringify(result)}`);
-            }
-          } catch (error) {
-            import_core6.logger.error(`Error in test_image_description: ${error instanceof Error ? error.message : String(error)}`);
-            throw error;
-          }
-        }
-      },
-      {
-        name: "google_test_structured_output_via_text_large",
-        fn: async (runtime) => {
-          try {
-            const schema = {
-              type: "object",
-              properties: {
-                name: { type: "string" },
-                age: { type: "number" },
-                hobbies: { type: "array", items: { type: "string" } }
-              },
-              required: ["name", "age", "hobbies"]
-            };
-            const result = await runtime.useModel(import_core6.ModelType.TEXT_LARGE, {
-              prompt: "Generate a person profile with name, age, and hobbies.",
-              responseSchema: schema
-            });
-            import_core6.logger.log("Generated structured output:", JSON.stringify(result));
-            if (!result) {
-              throw new Error("Generated structured output is empty");
-            }
-          } catch (error) {
-            import_core6.logger.error(`Error in test_structured_output_via_text_large: ${error instanceof Error ? error.message : String(error)}`);
-            throw error;
-          }
-        }
-      }
-    ]
-  }
-];
-function getProcessEnv() {
-  if (typeof process === "undefined") {
-    return {};
-  }
-  return process.env;
-}
-var env = getProcessEnv();
-var googleGenAIPlugin = {
-  name: "google-genai",
-  description: "Google Generative AI plugin for Gemini models",
-  autoEnable: {
-    envKeys: ["GOOGLE_API_KEY", "GOOGLE_GENERATIVE_AI_API_KEY"]
-  },
-  config: {
-    GOOGLE_GENERATIVE_AI_API_KEY: env.GOOGLE_GENERATIVE_AI_API_KEY ?? null,
-    GOOGLE_NANO_MODEL: env.GOOGLE_NANO_MODEL ?? null,
-    GOOGLE_MEDIUM_MODEL: env.GOOGLE_MEDIUM_MODEL ?? null,
-    GOOGLE_SMALL_MODEL: env.GOOGLE_SMALL_MODEL ?? null,
-    GOOGLE_LARGE_MODEL: env.GOOGLE_LARGE_MODEL ?? null,
-    GOOGLE_MEGA_MODEL: env.GOOGLE_MEGA_MODEL ?? null,
-    GOOGLE_RESPONSE_HANDLER_MODEL: env.GOOGLE_RESPONSE_HANDLER_MODEL ?? null,
-    GOOGLE_SHOULD_RESPOND_MODEL: env.GOOGLE_SHOULD_RESPOND_MODEL ?? null,
-    GOOGLE_ACTION_PLANNER_MODEL: env.GOOGLE_ACTION_PLANNER_MODEL ?? null,
-    GOOGLE_PLANNER_MODEL: env.GOOGLE_PLANNER_MODEL ?? null,
-    GOOGLE_IMAGE_MODEL: env.GOOGLE_IMAGE_MODEL ?? null,
-    GOOGLE_EMBEDDING_MODEL: env.GOOGLE_EMBEDDING_MODEL ?? null,
-    NANO_MODEL: env.NANO_MODEL ?? null,
-    MEDIUM_MODEL: env.MEDIUM_MODEL ?? null,
-    SMALL_MODEL: env.SMALL_MODEL ?? null,
-    LARGE_MODEL: env.LARGE_MODEL ?? null,
-    MEGA_MODEL: env.MEGA_MODEL ?? null,
-    RESPONSE_HANDLER_MODEL: env.RESPONSE_HANDLER_MODEL ?? null,
-    SHOULD_RESPOND_MODEL: env.SHOULD_RESPOND_MODEL ?? null,
-    ACTION_PLANNER_MODEL: env.ACTION_PLANNER_MODEL ?? null,
-    PLANNER_MODEL: env.PLANNER_MODEL ?? null,
-    IMAGE_MODEL: env.IMAGE_MODEL ?? null
-  },
-  async init(config, runtime) {
-    initializeGoogleGenAI(config, runtime);
-  },
-  models: {
-    [TEXT_NANO_MODEL_TYPE2]: async (runtime, params) => {
-      return handleTextNano(runtime, params);
-    },
-    [TEXT_MEDIUM_MODEL_TYPE2]: async (runtime, params) => {
-      return handleTextMedium(runtime, params);
-    },
-    [TEXT_SMALL_MODEL_TYPE2]: async (runtime, params) => {
-      return handleTextSmall(runtime, params);
-    },
-    [TEXT_LARGE_MODEL_TYPE2]: async (runtime, params) => {
-      return handleTextLarge(runtime, params);
-    },
-    [TEXT_MEGA_MODEL_TYPE2]: async (runtime, params) => {
-      return handleTextMega(runtime, params);
-    },
-    [RESPONSE_HANDLER_MODEL_TYPE2]: async (runtime, params) => {
-      return handleResponseHandler(runtime, params);
-    },
-    [ACTION_PLANNER_MODEL_TYPE2]: async (runtime, params) => {
-      return handleActionPlanner(runtime, params);
-    },
-    [TEXT_EMBEDDING_MODEL_TYPE2]: async (runtime, params) => {
-      return handleTextEmbedding(runtime, params);
-    },
-    [IMAGE_DESCRIPTION_MODEL_TYPE]: async (runtime, params) => {
-      return handleImageDescription(runtime, params);
-    }
-  },
-  tests: pluginTests
-};
-var plugin_google_genai_default = googleGenAIPlugin;
-
-// index.node.ts
-var index_node_default = plugin_google_genai_default;
-
-//# debugId=E1E56B528121F63364756E2164756E21