@elizaos/plugin-openai 1.0.0-beta.3 → 1.0.0-beta.33

package/LICENSE

@@ -1,6 +1,6 @@
 MIT License
 
-Copyright (c) 2025 Shaw Walters, aka Moon aka @lalalune
+Copyright (c) 2025 Shaw Walters and elizaOS Contributors
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

package/README.md

@@ -19,7 +19,9 @@ The plugin requires these environment variables (can be set in .env file or char
   "OPENAI_API_KEY": "your_openai_api_key",
   "OPENAI_BASE_URL": "optional_custom_endpoint",
   "OPENAI_SMALL_MODEL": "gpt-4o-mini",
-  "OPENAI_LARGE_MODEL": "gpt-4o"
+  "OPENAI_LARGE_MODEL": "gpt-4o",
+  "OPENAI_EMBEDDING_MODEL": "text-embedding-3-small",
+  "OPENAI_EMBEDDING_DIMENSIONS": "1536"
 }
 ```
 
@@ -31,6 +33,8 @@ OPENAI_API_KEY=your_openai_api_key
 OPENAI_BASE_URL=optional_custom_endpoint
 OPENAI_SMALL_MODEL=gpt-4o-mini
 OPENAI_LARGE_MODEL=gpt-4o
+OPENAI_EMBEDDING_MODEL=text-embedding-3-small
+OPENAI_EMBEDDING_DIMENSIONS=1536
 ```
 
 ### Configuration Options
@@ -39,12 +43,14 @@ OPENAI_LARGE_MODEL=gpt-4o
 - `OPENAI_BASE_URL`: Custom API endpoint (default: https://api.openai.com/v1)
 - `OPENAI_SMALL_MODEL`: Defaults to GPT-4o Mini ("gpt-4o-mini")
 - `OPENAI_LARGE_MODEL`: Defaults to GPT-4o ("gpt-4o")
+- `OPENAI_EMBEDDING_MODEL`: Defaults to text-embedding-3-small ("text-embedding-3-small")
+- `OPENAI_EMBEDDING_DIMENSIONS`: Defaults to 1536 (1536)
 
 The plugin provides these model classes:
 
 - `TEXT_SMALL`: Optimized for fast, cost-effective responses
 - `TEXT_LARGE`: For complex tasks requiring deeper reasoning
-- `TEXT_EMBEDDING`: Text embedding model (text-embedding-3-small)
+- `TEXT_EMBEDDING`: Text embedding model (text-embedding-3-small by default)
 - `IMAGE`: DALL-E image generation
 - `IMAGE_DESCRIPTION`: GPT-4o image analysis
 - `TRANSCRIPTION`: Whisper audio transcription

package/dist/index.js

@@ -2,10 +2,34 @@
 import { createOpenAI } from "@ai-sdk/openai";
 import {
   ModelType,
-  logger
+  logger,
+  VECTOR_DIMS
 } from "@elizaos/core";
-import { generateObject, generateText } from "ai";
+import { generateObject, generateText, JSONParseError } from "ai";
 import { encodingForModel } from "js-tiktoken";
+import FormData from "form-data";
+import fetch from "node-fetch";
+function getSetting(runtime, key, defaultValue) {
+  return runtime.getSetting(key) ?? process.env[key] ?? defaultValue;
+}
+function getBaseURL(runtime) {
+  return getSetting(runtime, "OPENAI_BASE_URL", "https://api.openai.com/v1");
+}
+function getApiKey(runtime) {
+  return getSetting(runtime, "OPENAI_API_KEY");
+}
+function getSmallModel(runtime) {
+  return getSetting(runtime, "OPENAI_SMALL_MODEL") ?? getSetting(runtime, "SMALL_MODEL", "gpt-4o-mini");
+}
+function getLargeModel(runtime) {
+  return getSetting(runtime, "OPENAI_LARGE_MODEL") ?? getSetting(runtime, "LARGE_MODEL", "gpt-4o");
+}
+function createOpenAIClient(runtime) {
+  return createOpenAI({
+    apiKey: getApiKey(runtime),
+    baseURL: getBaseURL(runtime)
+  });
+}
 async function tokenizeText(model, prompt) {
   const modelName = model === ModelType.TEXT_SMALL ? process.env.OPENAI_SMALL_MODEL ?? process.env.SMALL_MODEL ?? "gpt-4o-mini" : process.env.LARGE_MODEL ?? "gpt-4o";
   const encoding = encodingForModel(modelName);
@@ -17,6 +41,69 @@ async function detokenizeText(model, tokens) {
   const encoding = encodingForModel(modelName);
   return encoding.decode(tokens);
 }
+async function generateObjectByModelType(runtime, params, modelType, getModelFn) {
+  const openai = createOpenAIClient(runtime);
+  const model = getModelFn(runtime);
+  try {
+    if (params.schema) {
+      logger.info(`Using ${modelType} without schema validation`);
+    }
+    const { object } = await generateObject({
+      model: openai.languageModel(model),
+      output: "no-schema",
+      prompt: params.prompt,
+      temperature: params.temperature,
+      experimental_repairText: getJsonRepairFunction()
+    });
+    return object;
+  } catch (error) {
+    logger.error(`Error generating object with ${modelType}:`, error);
+    throw error;
+  }
+}
+function getJsonRepairFunction() {
+  return async ({ text, error }) => {
+    try {
+      if (error instanceof JSONParseError) {
+        const cleanedText = text.replace(/```json\n|\n```|```/g, "");
+        JSON.parse(cleanedText);
+        return cleanedText;
+      }
+    } catch (jsonError) {
+      logger.warn("Failed to repair JSON text:", jsonError);
+      return null;
+    }
+  };
+}
+async function fetchTextToSpeech(runtime, text) {
+  const apiKey = getApiKey(runtime);
+  const model = getSetting(runtime, "OPENAI_TTS_MODEL", "gpt-4o-mini-tts");
+  const voice = getSetting(runtime, "OPENAI_TTS_VOICE", "nova");
+  const instructions = getSetting(runtime, "OPENAI_TTS_INSTRUCTIONS", "");
+  const baseURL = getBaseURL(runtime);
+  try {
+    const res = await fetch(`${baseURL}/audio/speech`, {
+      method: "POST",
+      headers: {
+        Authorization: `Bearer ${apiKey}`,
+        "Content-Type": "application/json"
+      },
+      body: JSON.stringify({
+        model,
+        voice,
+        input: text,
+        ...instructions && { instructions }
+      })
+    });
+    if (!res.ok) {
+      const err = await res.text();
+      throw new Error(`OpenAI TTS error ${res.status}: ${err}`);
+    }
+    return res.body;
+  } catch (err) {
+    throw new Error(`Failed to fetch speech from OpenAI TTS: ${err.message || err}`);
+  }
+}
 var openaiPlugin = {
   name: "openai",
   description: "OpenAI plugin",
@@ -26,20 +113,22 @@ var openaiPlugin = {
     OPENAI_SMALL_MODEL: process.env.OPENAI_SMALL_MODEL,
     OPENAI_LARGE_MODEL: process.env.OPENAI_LARGE_MODEL,
     SMALL_MODEL: process.env.SMALL_MODEL,
-    LARGE_MODEL: process.env.LARGE_MODEL
+    LARGE_MODEL: process.env.LARGE_MODEL,
+    OPENAI_EMBEDDING_MODEL: process.env.OPENAI_EMBEDDING_MODEL,
+    OPENAI_EMBEDDING_DIMENSIONS: process.env.OPENAI_EMBEDDING_DIMENSIONS
   },
-  async init(config) {
+  async init(_config, runtime) {
     try {
-      if (!process.env.OPENAI_API_KEY) {
+      if (!getApiKey(runtime)) {
         logger.warn(
           "OPENAI_API_KEY is not set in environment - OpenAI functionality will be limited"
         );
         return;
       }
       try {
-        const baseURL = process.env.OPENAI_BASE_URL ?? "https://api.openai.com/v1";
+        const baseURL = getBaseURL(runtime);
         const response = await fetch(`${baseURL}/models`, {
-          headers: { Authorization: `Bearer ${process.env.OPENAI_API_KEY}` }
+          headers: { Authorization: `Bearer ${getApiKey(runtime)}` }
         });
         if (!response.ok) {
           logger.warn(`OpenAI API key validation failed: ${response.statusText}`);
@@ -57,10 +146,21 @@ var openaiPlugin = {
     }
   },
   models: {
-    [ModelType.TEXT_EMBEDDING]: async (_runtime, params) => {
+    [ModelType.TEXT_EMBEDDING]: async (runtime, params) => {
+      const embeddingDimension = parseInt(
+        getSetting(runtime, "OPENAI_EMBEDDING_DIMENSIONS", "1536")
+      );
+      if (!Object.values(VECTOR_DIMS).includes(embeddingDimension)) {
+        logger.error(
+          `Invalid embedding dimension: ${embeddingDimension}. Must be one of: ${Object.values(VECTOR_DIMS).join(", ")}`
+        );
+        throw new Error(
+          `Invalid embedding dimension: ${embeddingDimension}. Must be one of: ${Object.values(VECTOR_DIMS).join(", ")}`
+        );
+      }
       if (params === null) {
         logger.debug("Creating test embedding for initialization");
-        const testVector = Array(1536).fill(0);
+        const testVector = Array(embeddingDimension).fill(0);
         testVector[0] = 0.1;
         return testVector;
       }
@@ -71,39 +171,39 @@ var openaiPlugin = {
         text = params.text;
       } else {
         logger.warn("Invalid input format for embedding");
-        const fallbackVector = Array(1536).fill(0);
+        const fallbackVector = Array(embeddingDimension).fill(0);
         fallbackVector[0] = 0.2;
         return fallbackVector;
       }
       if (!text.trim()) {
         logger.warn("Empty text for embedding");
-        const emptyVector = Array(1536).fill(0);
+        const emptyVector = Array(embeddingDimension).fill(0);
         emptyVector[0] = 0.3;
         return emptyVector;
       }
       try {
-        const baseURL = process.env.OPENAI_BASE_URL ?? "https://api.openai.com/v1";
+        const baseURL = getBaseURL(runtime);
         const response = await fetch(`${baseURL}/embeddings`, {
           method: "POST",
           headers: {
-            Authorization: `Bearer ${process.env.OPENAI_API_KEY}`,
+            Authorization: `Bearer ${getApiKey(runtime)}`,
             "Content-Type": "application/json"
           },
           body: JSON.stringify({
-            model: "text-embedding-3-small",
+            model: getSetting(runtime, "OPENAI_EMBEDDING_MODEL", "text-embedding-3-small"),
             input: text
           })
         });
         if (!response.ok) {
           logger.error(`OpenAI API error: ${response.status} - ${response.statusText}`);
-          const errorVector = Array(1536).fill(0);
+          const errorVector = Array(embeddingDimension).fill(0);
           errorVector[0] = 0.4;
           return errorVector;
         }
         const data = await response.json();
         if (!data?.data?.[0]?.embedding) {
           logger.error("API returned invalid structure");
-          const errorVector = Array(1536).fill(0);
+          const errorVector = Array(embeddingDimension).fill(0);
           errorVector[0] = 0.5;
           return errorVector;
         }
@@ -112,7 +212,7 @@ var openaiPlugin = {
         return embedding;
       } catch (error) {
         logger.error("Error generating embedding:", error);
-        const errorVector = Array(1536).fill(0);
+        const errorVector = Array(embeddingDimension).fill(0);
         errorVector[0] = 0.6;
         return errorVector;
       }
@@ -128,12 +228,8 @@ var openaiPlugin = {
       const frequency_penalty = 0.7;
       const presence_penalty = 0.7;
       const max_response_length = 8192;
-      const baseURL = runtime.getSetting("OPENAI_BASE_URL") ?? "https://api.openai.com/v1";
-      const openai = createOpenAI({
-        apiKey: runtime.getSetting("OPENAI_API_KEY"),
-        baseURL
-      });
-      const model = runtime.getSetting("OPENAI_SMALL_MODEL") ?? runtime.getSetting("SMALL_MODEL") ?? "gpt-4o-mini";
+      const openai = createOpenAIClient(runtime);
+      const model = getSmallModel(runtime);
       logger.log("generating text");
       logger.log(prompt);
       const { text: openaiResponse } = await generateText({
@@ -156,12 +252,8 @@ var openaiPlugin = {
       frequencyPenalty = 0.7,
       presencePenalty = 0.7
     }) => {
-      const baseURL = runtime.getSetting("OPENAI_BASE_URL") ?? "https://api.openai.com/v1";
-      const openai = createOpenAI({
-        apiKey: runtime.getSetting("OPENAI_API_KEY"),
-        baseURL
-      });
-      const model = runtime.getSetting("OPENAI_LARGE_MODEL") ?? runtime.getSetting("LARGE_MODEL") ?? "gpt-4o";
+      const openai = createOpenAIClient(runtime);
+      const model = getLargeModel(runtime);
       const { text: openaiResponse } = await generateText({
         model: openai.languageModel(model),
         prompt,
@@ -175,11 +267,11 @@ var openaiPlugin = {
       return openaiResponse;
     },
     [ModelType.IMAGE]: async (runtime, params) => {
-      const baseURL = runtime.getSetting("OPENAI_BASE_URL") ?? "https://api.openai.com/v1";
+      const baseURL = getBaseURL(runtime);
       const response = await fetch(`${baseURL}/images/generations`, {
         method: "POST",
         headers: {
-          Authorization: `Bearer ${runtime.getSetting("OPENAI_API_KEY")}`,
+          Authorization: `Bearer ${getApiKey(runtime)}`,
           "Content-Type": "application/json"
         },
         body: JSON.stringify({
@@ -206,8 +298,8 @@ var openaiPlugin = {
         prompt = params.prompt;
       }
       try {
-        const baseURL = process.env.OPENAI_BASE_URL ?? "https://api.openai.com/v1";
-        const apiKey = process.env.OPENAI_API_KEY;
+        const baseURL = getBaseURL(runtime);
+        const apiKey = getApiKey(runtime);
         if (!apiKey) {
           logger.error("OpenAI API key not set");
           return {
@@ -222,7 +314,7 @@ var openaiPlugin = {
             Authorization: `Bearer ${apiKey}`
           },
           body: JSON.stringify({
-            model: "gpt-4-vision-preview",
+            model: "gpt-4o-mini",
             messages: [
               {
                 role: "user",
@@ -266,15 +358,17 @@ var openaiPlugin = {
     },
     [ModelType.TRANSCRIPTION]: async (runtime, audioBuffer) => {
       logger.log("audioBuffer", audioBuffer);
-      const baseURL = runtime.getSetting("OPENAI_BASE_URL") ?? "https://api.openai.com/v1";
+      const baseURL = getBaseURL(runtime);
       const formData = new FormData();
-      formData.append("file", new Blob([audioBuffer], { type: "audio/mp3" }));
+      formData.append("file", audioBuffer, {
+        filename: "recording.mp3",
+        contentType: "audio/mp3"
+      });
       formData.append("model", "whisper-1");
       const response = await fetch(`${baseURL}/audio/transcriptions`, {
         method: "POST",
         headers: {
-          Authorization: `Bearer ${runtime.getSetting("OPENAI_API_KEY")}`
-          // Note: Do not set a Content-Type header—letting fetch set it for FormData is best
+          Authorization: `Bearer ${getApiKey(runtime)}`
         },
         body: formData
       });
@@ -285,65 +379,14 @@ var openaiPlugin = {
       const data = await response.json();
       return data.text;
     },
+    [ModelType.TEXT_TO_SPEECH]: async (runtime, text) => {
+      return await fetchTextToSpeech(runtime, text);
+    },
     [ModelType.OBJECT_SMALL]: async (runtime, params) => {
-      const baseURL = runtime.getSetting("OPENAI_BASE_URL") ?? "https://api.openai.com/v1";
-      const openai = createOpenAI({
-        apiKey: runtime.getSetting("OPENAI_API_KEY"),
-        baseURL
-      });
-      const model = runtime.getSetting("OPENAI_SMALL_MODEL") ?? runtime.getSetting("SMALL_MODEL") ?? "gpt-4o-mini";
-      try {
-        if (params.schema) {
-          logger.info("Using OBJECT_SMALL without schema validation");
-          const { object: object2 } = await generateObject({
-            model: openai.languageModel(model),
-            output: "no-schema",
-            prompt: params.prompt,
-            temperature: params.temperature
-          });
-          return object2;
-        }
-        const { object } = await generateObject({
-          model: openai.languageModel(model),
-          output: "no-schema",
-          prompt: params.prompt,
-          temperature: params.temperature
-        });
-        return object;
-      } catch (error) {
-        logger.error("Error generating object:", error);
-        throw error;
-      }
+      return generateObjectByModelType(runtime, params, ModelType.OBJECT_SMALL, getSmallModel);
     },
     [ModelType.OBJECT_LARGE]: async (runtime, params) => {
-      const baseURL = runtime.getSetting("OPENAI_BASE_URL") ?? "https://api.openai.com/v1";
-      const openai = createOpenAI({
-        apiKey: runtime.getSetting("OPENAI_API_KEY"),
-        baseURL
-      });
-      const model = runtime.getSetting("OPENAI_LARGE_MODEL") ?? runtime.getSetting("LARGE_MODEL") ?? "gpt-4o";
-      try {
-        if (params.schema) {
-          logger.info("Using OBJECT_LARGE without schema validation");
-          const { object: object2 } = await generateObject({
-            model: openai.languageModel(model),
-            output: "no-schema",
-            prompt: params.prompt,
-            temperature: params.temperature
-          });
-          return object2;
-        }
-        const { object } = await generateObject({
-          model: openai.languageModel(model),
-          output: "no-schema",
-          prompt: params.prompt,
-          temperature: params.temperature
-        });
-        return object;
-      } catch (error) {
-        logger.error("Error generating object:", error);
-        throw error;
-      }
+      return generateObjectByModelType(runtime, params, ModelType.OBJECT_LARGE, getLargeModel);
     }
   },
   tests: [
@@ -353,10 +396,10 @@ var openaiPlugin = {
         {
           name: "openai_test_url_and_api_key_validation",
           fn: async (runtime) => {
-            const baseURL = runtime.getSetting("OPENAI_BASE_URL") ?? "https://api.openai.com/v1";
+            const baseURL = getBaseURL(runtime);
             const response = await fetch(`${baseURL}/models`, {
               headers: {
-                Authorization: `Bearer ${runtime.getSetting("OPENAI_API_KEY")}`
+                Authorization: `Bearer ${getApiKey(runtime)}`
               }
             });
             const data = await response.json();
@@ -498,6 +541,22 @@ var openaiPlugin = {
             }
             logger.log("Decoded text:", decodedText);
           }
+        },
+        {
+          name: "openai_test_text_to_speech",
+          fn: async (runtime) => {
+            try {
+              const text = "Hello, this is a test for text-to-speech.";
+              const response = await fetchTextToSpeech(runtime, text);
+              if (!response) {
+                throw new Error("Failed to generate speech");
+              }
+              logger.log("Generated speech successfully");
+            } catch (error) {
+              logger.error("Error in openai_test_text_to_speech:", error);
+              throw error;
+            }
+          }
         }
       ]
     }

package/dist/index.js.map

@@ -1 +1 @@
-{"version":3,"sources":["../src/index.ts"],"sourcesContent":["import { createOpenAI } from '@ai-sdk/openai';\nimport type {\n  ImageDescriptionParams,\n  ModelTypeName,\n  ObjectGenerationParams,\n  Plugin,\n  TextEmbeddingParams,\n} from '@elizaos/core';\nimport {\n  type DetokenizeTextParams,\n  type GenerateTextParams,\n  ModelType,\n  type TokenizeTextParams,\n  logger,\n} from '@elizaos/core';\nimport { generateObject, generateText } from 'ai';\nimport { type TiktokenModel, encodingForModel } from 'js-tiktoken';\nimport { z } from 'zod';\n\n/**\n * Asynchronously tokenizes the given text based on the specified model and prompt.\n *\n * @param {ModelTypeName} model - The type of model to use for tokenization.\n * @param {string} prompt - The text prompt to tokenize.\n * @returns {number[]} - An array of tokens representing the encoded prompt.\n */\nasync function tokenizeText(model: ModelTypeName, prompt: string) {\n  const modelName =\n    model === ModelType.TEXT_SMALL\n      ? (process.env.OPENAI_SMALL_MODEL ?? process.env.SMALL_MODEL ?? 'gpt-4o-mini')\n      : (process.env.LARGE_MODEL ?? 'gpt-4o');\n  const encoding = encodingForModel(modelName as TiktokenModel);\n  const tokens = encoding.encode(prompt);\n  return tokens;\n}\n\n/**\n * Detokenize a sequence of tokens back into text using the specified model.\n *\n * @param {ModelTypeName} model - The type of model to use for detokenization.\n * @param {number[]} tokens - The sequence of tokens to detokenize.\n * @returns {string} The detokenized text.\n */\nasync function detokenizeText(model: ModelTypeName, tokens: number[]) {\n  const modelName =\n    model === ModelType.TEXT_SMALL\n      ? (process.env.OPENAI_SMALL_MODEL ?? process.env.SMALL_MODEL ?? 'gpt-4o-mini')\n      : (process.env.OPENAI_LARGE_MODEL ?? process.env.LARGE_MODEL ?? 'gpt-4o');\n  const encoding = encodingForModel(modelName as TiktokenModel);\n  return encoding.decode(tokens);\n}\n\n/**\n * Defines the OpenAI plugin with its name, description, and configuration options.\n * @type {Plugin}\n */\nexport const openaiPlugin: Plugin = {\n  name: 'openai',\n  description: 'OpenAI plugin',\n  config: {\n    OPENAI_API_KEY: process.env.OPENAI_API_KEY,\n    OPENAI_BASE_URL: process.env.OPENAI_BASE_URL,\n    OPENAI_SMALL_MODEL: process.env.OPENAI_SMALL_MODEL,\n    OPENAI_LARGE_MODEL: process.env.OPENAI_LARGE_MODEL,\n    SMALL_MODEL: process.env.SMALL_MODEL,\n    LARGE_MODEL: process.env.LARGE_MODEL,\n  },\n  async init(config: Record<string, string>) {\n    try {\n      // const validatedConfig = await configSchema.parseAsync(config);\n\n      // // Set all environment variables at once\n      // for (const [key, value] of Object.entries(validatedConfig)) {\n      // \tif (value) process.env[key] = value;\n      // }\n\n      // If API key is not set, we'll show a warning but continue\n      if (!process.env.OPENAI_API_KEY) {\n        logger.warn(\n          'OPENAI_API_KEY is not set in environment - OpenAI functionality will be limited'\n        );\n        // Return early without throwing an error\n        return;\n      }\n\n      // Verify API key only if we have one\n      try {\n        const baseURL = process.env.OPENAI_BASE_URL ?? 'https://api.openai.com/v1';\n        const response = await fetch(`${baseURL}/models`, {\n          headers: { Authorization: `Bearer ${process.env.OPENAI_API_KEY}` },\n        });\n\n        if (!response.ok) {\n          logger.warn(`OpenAI API key validation failed: ${response.statusText}`);\n          logger.warn('OpenAI functionality will be limited until a valid API key is provided');\n          // Continue execution instead of throwing\n        } else {\n          // logger.log(\"OpenAI API key validated successfully\");\n        }\n      } catch (fetchError) {\n        logger.warn(`Error validating OpenAI API key: ${fetchError}`);\n        logger.warn('OpenAI functionality will be limited until a valid API key is provided');\n        // Continue execution instead of throwing\n      }\n    } catch (error) {\n      // Convert to warning instead of error\n      logger.warn(\n        `OpenAI plugin configuration issue: ${error.errors\n          .map((e) => e.message)\n          .join(', ')} - You need to configure the OPENAI_API_KEY in your environment variables`\n      );\n    }\n  },\n  models: {\n    [ModelType.TEXT_EMBEDDING]: async (\n      _runtime,\n      params: TextEmbeddingParams | string | null\n    ): Promise<number[]> => {\n      // Handle null input (initialization case)\n      if (params === null) {\n        logger.debug('Creating test embedding for initialization');\n        // Return a consistent vector for null input\n        const testVector = Array(1536).fill(0);\n        testVector[0] = 0.1; // Make it non-zero\n        return testVector;\n      }\n\n      // Get the text from whatever format was provided\n      let text: string;\n      if (typeof params === 'string') {\n        text = params; // Direct string input\n      } else if (typeof params === 'object' && params.text) {\n        text = params.text; // Object with text property\n      } else {\n        logger.warn('Invalid input format for embedding');\n        // Return a fallback for invalid input\n        const fallbackVector = Array(1536).fill(0);\n        fallbackVector[0] = 0.2; // Different value for tracking\n        return fallbackVector;\n      }\n\n      // Skip API call for empty text\n      if (!text.trim()) {\n        logger.warn('Empty text for embedding');\n        const emptyVector = Array(1536).fill(0);\n        emptyVector[0] = 0.3; // Different value for tracking\n        return emptyVector;\n      }\n\n      try {\n        const baseURL = process.env.OPENAI_BASE_URL ?? 'https://api.openai.com/v1';\n\n        // Call the OpenAI API\n        const response = await fetch(`${baseURL}/embeddings`, {\n          method: 'POST',\n          headers: {\n            Authorization: `Bearer ${process.env.OPENAI_API_KEY}`,\n            'Content-Type': 'application/json',\n          },\n          body: JSON.stringify({\n            model: 'text-embedding-3-small',\n            input: text,\n          }),\n        });\n\n        if (!response.ok) {\n          logger.error(`OpenAI API error: ${response.status} - ${response.statusText}`);\n          const errorVector = Array(1536).fill(0);\n          errorVector[0] = 0.4; // Different value for tracking\n          return errorVector;\n        }\n\n        const data = (await response.json()) as {\n          data: [{ embedding: number[] }];\n        };\n\n        if (!data?.data?.[0]?.embedding) {\n          logger.error('API returned invalid structure');\n          const errorVector = Array(1536).fill(0);\n          errorVector[0] = 0.5; // Different value for tracking\n          return errorVector;\n        }\n\n        const embedding = data.data[0].embedding;\n        logger.log(`Got valid embedding with length ${embedding.length}`);\n        return embedding;\n      } catch (error) {\n        logger.error('Error generating embedding:', error);\n        const errorVector = Array(1536).fill(0);\n        errorVector[0] = 0.6; // Different value for tracking\n        return errorVector;\n      }\n    },\n    [ModelType.TEXT_TOKENIZER_ENCODE]: async (\n      _runtime,\n      { prompt, modelType = ModelType.TEXT_LARGE }: TokenizeTextParams\n    ) => {\n      return await tokenizeText(modelType ?? ModelType.TEXT_LARGE, prompt);\n    },\n    [ModelType.TEXT_TOKENIZER_DECODE]: async (\n      _runtime,\n      { tokens, modelType = ModelType.TEXT_LARGE }: DetokenizeTextParams\n    ) => {\n      return await detokenizeText(modelType ?? ModelType.TEXT_LARGE, tokens);\n    },\n    [ModelType.TEXT_SMALL]: async (runtime, { prompt, stopSequences = [] }: GenerateTextParams) => {\n      const temperature = 0.7;\n      const frequency_penalty = 0.7;\n      const presence_penalty = 0.7;\n      const max_response_length = 8192;\n\n      const baseURL = runtime.getSetting('OPENAI_BASE_URL') ?? 'https://api.openai.com/v1';\n\n      const openai = createOpenAI({\n        apiKey: runtime.getSetting('OPENAI_API_KEY'),\n        baseURL,\n      });\n\n      const model =\n        runtime.getSetting('OPENAI_SMALL_MODEL') ??\n        runtime.getSetting('SMALL_MODEL') ??\n        'gpt-4o-mini';\n\n      logger.log('generating text');\n      logger.log(prompt);\n\n      const { text: openaiResponse } = await generateText({\n        model: openai.languageModel(model),\n        prompt: prompt,\n        system: runtime.character.system ?? undefined,\n        temperature: temperature,\n        maxTokens: max_response_length,\n        frequencyPenalty: frequency_penalty,\n        presencePenalty: presence_penalty,\n        stopSequences: stopSequences,\n      });\n\n      return openaiResponse;\n    },\n    [ModelType.TEXT_LARGE]: async (\n      runtime,\n      {\n        prompt,\n        stopSequences = [],\n        maxTokens = 8192,\n        temperature = 0.7,\n        frequencyPenalty = 0.7,\n        presencePenalty = 0.7,\n      }: GenerateTextParams\n    ) => {\n      const baseURL = runtime.getSetting('OPENAI_BASE_URL') ?? 'https://api.openai.com/v1';\n\n      const openai = createOpenAI({\n        apiKey: runtime.getSetting('OPENAI_API_KEY'),\n        baseURL,\n      });\n\n      const model =\n        runtime.getSetting('OPENAI_LARGE_MODEL') ?? runtime.getSetting('LARGE_MODEL') ?? 'gpt-4o';\n\n      const { text: openaiResponse } = await generateText({\n        model: openai.languageModel(model),\n        prompt: prompt,\n        system: runtime.character.system ?? undefined,\n        temperature: temperature,\n        maxTokens: maxTokens,\n        frequencyPenalty: frequencyPenalty,\n        presencePenalty: presencePenalty,\n        stopSequences: stopSequences,\n      });\n\n      return openaiResponse;\n    },\n    [ModelType.IMAGE]: async (\n      runtime,\n      params: {\n        prompt: string;\n        n?: number;\n        size?: string;\n      }\n    ) => {\n      const baseURL = runtime.getSetting('OPENAI_BASE_URL') ?? 'https://api.openai.com/v1';\n      const response = await fetch(`${baseURL}/images/generations`, {\n        method: 'POST',\n        headers: {\n          Authorization: `Bearer ${runtime.getSetting('OPENAI_API_KEY')}`,\n          'Content-Type': 'application/json',\n        },\n        body: JSON.stringify({\n          prompt: params.prompt,\n          n: params.n || 1,\n          size: params.size || '1024x1024',\n        }),\n      });\n      if (!response.ok) {\n        throw new Error(`Failed to generate image: ${response.statusText}`);\n      }\n      const data = await response.json();\n      const typedData = data as { data: { url: string }[] };\n      return typedData.data;\n    },\n    [ModelType.IMAGE_DESCRIPTION]: async (runtime, params: ImageDescriptionParams | string) => {\n      // Handle string case (direct URL)\n      let imageUrl: string;\n      let prompt: string | undefined;\n\n      if (typeof params === 'string') {\n        imageUrl = params;\n        prompt = undefined;\n      } else {\n        // Object parameter case\n        imageUrl = params.imageUrl;\n        prompt = params.prompt;\n      }\n\n      try {\n        const baseURL = process.env.OPENAI_BASE_URL ?? 'https://api.openai.com/v1';\n        const apiKey = process.env.OPENAI_API_KEY;\n\n        if (!apiKey) {\n          logger.error('OpenAI API key not set');\n          return {\n            title: 'Failed to analyze image',\n            description: 'API key not configured',\n          };\n        }\n\n        // Call the GPT-4 Vision API\n        const response = await fetch(`${baseURL}/chat/completions`, {\n          method: 'POST',\n          headers: {\n            'Content-Type': 'application/json',\n            Authorization: `Bearer ${apiKey}`,\n          },\n          body: JSON.stringify({\n            model: 'gpt-4-vision-preview',\n            messages: [\n              {\n                role: 'user',\n                content: [\n                  {\n                    type: 'text',\n                    text:\n                      prompt ||\n                      'Please analyze this image and provide a title and detailed description.',\n                  },\n                  {\n                    type: 'image_url',\n                    image_url: { url: imageUrl },\n                  },\n                ],\n              },\n            ],\n            max_tokens: 300,\n          }),\n        });\n\n        if (!response.ok) {\n          throw new Error(`OpenAI API error: ${response.status}`);\n        }\n\n        const result: any = await response.json();\n        const content = result.choices?.[0]?.message?.content;\n\n        if (!content) {\n          return {\n            title: 'Failed to analyze image',\n            description: 'No response from API',\n          };\n        }\n\n        // Extract title and description\n        const titleMatch = content.match(/title[:\\s]+(.+?)(?:\\n|$)/i);\n        const title = titleMatch?.[1] || 'Image Analysis';\n\n        // Rest of content is the description\n        const description = content.replace(/title[:\\s]+(.+?)(?:\\n|$)/i, '').trim();\n\n        return { title, description };\n      } catch (error) {\n        logger.error('Error analyzing image:', error);\n        return {\n          title: 'Failed to analyze image',\n          description: `Error: ${error instanceof Error ? error.message : String(error)}`,\n        };\n      }\n    },\n    [ModelType.TRANSCRIPTION]: async (runtime, audioBuffer: Buffer) => {\n      logger.log('audioBuffer', audioBuffer);\n      const baseURL = runtime.getSetting('OPENAI_BASE_URL') ?? 'https://api.openai.com/v1';\n      const formData = new FormData();\n      formData.append('file', new Blob([audioBuffer], { type: 'audio/mp3' }));\n      formData.append('model', 'whisper-1');\n      const response = await fetch(`${baseURL}/audio/transcriptions`, {\n        method: 'POST',\n        headers: {\n          Authorization: `Bearer ${runtime.getSetting('OPENAI_API_KEY')}`,\n          // Note: Do not set a Content-Type header—letting fetch set it for FormData is best\n        },\n        body: formData,\n      });\n\n      logger.log('response', response);\n      if (!response.ok) {\n        throw new Error(`Failed to transcribe audio: ${response.statusText}`);\n      }\n      const data = (await response.json()) as { text: string };\n      return data.text;\n    },\n    [ModelType.OBJECT_SMALL]: async (runtime, params: ObjectGenerationParams) => {\n      const baseURL = runtime.getSetting('OPENAI_BASE_URL') ?? 'https://api.openai.com/v1';\n      const openai = createOpenAI({\n        apiKey: runtime.getSetting('OPENAI_API_KEY'),\n        baseURL,\n      });\n      const model =\n        runtime.getSetting('OPENAI_SMALL_MODEL') ??\n        runtime.getSetting('SMALL_MODEL') ??\n        'gpt-4o-mini';\n\n      try {\n        if (params.schema) {\n          // Skip zod validation and just use the generateObject without schema\n          logger.info('Using OBJECT_SMALL without schema validation');\n          const { object } = await generateObject({\n            model: openai.languageModel(model),\n            output: 'no-schema',\n            prompt: params.prompt,\n            temperature: params.temperature,\n          });\n          return object;\n        }\n\n        const { object } = await generateObject({\n          model: openai.languageModel(model),\n          output: 'no-schema',\n          prompt: params.prompt,\n          temperature: params.temperature,\n        });\n        return object;\n      } catch (error) {\n        logger.error('Error generating object:', error);\n        throw error;\n      }\n    },\n    [ModelType.OBJECT_LARGE]: async (runtime, params: ObjectGenerationParams) => {\n      const baseURL = runtime.getSetting('OPENAI_BASE_URL') ?? 'https://api.openai.com/v1';\n      const openai = createOpenAI({\n        apiKey: runtime.getSetting('OPENAI_API_KEY'),\n        baseURL,\n      });\n      const model =\n        runtime.getSetting('OPENAI_LARGE_MODEL') ?? runtime.getSetting('LARGE_MODEL') ?? 'gpt-4o';\n\n      try {\n        if (params.schema) {\n          // Skip zod validation and just use the generateObject without schema\n          logger.info('Using OBJECT_LARGE without schema validation');\n          const { object } = await generateObject({\n            model: openai.languageModel(model),\n            output: 'no-schema',\n            prompt: params.prompt,\n            temperature: params.temperature,\n          });\n          return object;\n        }\n\n        const { object } = await generateObject({\n          model: openai.languageModel(model),\n          output: 'no-schema',\n          prompt: params.prompt,\n          temperature: params.temperature,\n        });\n        return object;\n      } catch (error) {\n        logger.error('Error generating object:', error);\n        throw error;\n      }\n    },\n  },\n  tests: [\n    {\n      name: 'openai_plugin_tests',\n      tests: [\n        {\n          name: 'openai_test_url_and_api_key_validation',\n          fn: async (runtime) => {\n            const baseURL = runtime.getSetting('OPENAI_BASE_URL') ?? 'https://api.openai.com/v1';\n            const response = await fetch(`${baseURL}/models`, {\n              headers: {\n                Authorization: `Bearer ${runtime.getSetting('OPENAI_API_KEY')}`,\n              },\n            });\n            const data = await response.json();\n            logger.log('Models Available:', (data as any)?.data.length);\n            if (!response.ok) {\n              throw new Error(`Failed to validate OpenAI API key: ${response.statusText}`);\n            }\n          },\n        },\n        {\n          name: 'openai_test_text_embedding',\n          fn: async (runtime) => {\n            try {\n              const embedding = await runtime.useModel(ModelType.TEXT_EMBEDDING, {\n                text: 'Hello, world!',\n              });\n              logger.log('embedding', embedding);\n            } catch (error) {\n              logger.error('Error in test_text_embedding:', error);\n              throw error;\n            }\n          },\n        },\n        {\n          name: 'openai_test_text_large',\n          fn: async (runtime) => {\n            try {\n              const text = await runtime.useModel(ModelType.TEXT_LARGE, {\n                prompt: 'What is the nature of reality in 10 words?',\n              });\n              if (text.length === 0) {\n                throw new Error('Failed to generate text');\n              }\n              logger.log('generated with test_text_large:', text);\n            } catch (error) {\n              logger.error('Error in test_text_large:', error);\n              throw error;\n            }\n          },\n        },\n        {\n          name: 'openai_test_text_small',\n          fn: async (runtime) => {\n            try {\n              const text = await runtime.useModel(ModelType.TEXT_SMALL, {\n                prompt: 'What is the nature of reality in 10 words?',\n              });\n              if (text.length === 0) {\n                throw new Error('Failed to generate text');\n              }\n              logger.log('generated with test_text_small:', text);\n            } catch (error) {\n              logger.error('Error in test_text_small:', error);\n              throw error;\n            }\n          },\n        },\n        {\n          name: 'openai_test_image_generation',\n          fn: async (runtime) => {\n            logger.log('openai_test_image_generation');\n            try {\n              const image = await runtime.useModel(ModelType.IMAGE, {\n                prompt: 'A beautiful sunset over a calm ocean',\n                n: 1,\n                size: '1024x1024',\n              });\n              logger.log('generated with test_image_generation:', image);\n            } catch (error) {\n              logger.error('Error in test_image_generation:', error);\n              throw error;\n            }\n          },\n        },\n        {\n          name: 'image-description',\n          fn: async (runtime) => {\n            try {\n              logger.log('openai_test_image_description');\n              try {\n                const result = await runtime.useModel(\n                  ModelType.IMAGE_DESCRIPTION,\n                  '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'\n                );\n\n                // Check if result has the expected structure\n                if (\n                  result &&\n                  typeof result === 'object' &&\n                  'title' in result &&\n                  'description' in result\n                ) {\n                  logger.log('Image description:', result);\n                } else {\n                  logger.error('Invalid image description result format:', result);\n                }\n              } catch (e) {\n                logger.error('Error in image description test:', e);\n              }\n            } catch (e) {\n              logger.error('Error in openai_test_image_description:', e);\n            }\n          },\n        },\n        {\n          name: 'openai_test_transcription',\n          fn: async (runtime) => {\n            logger.log('openai_test_transcription');\n            try {\n              const response = await fetch(\n                'https://upload.wikimedia.org/wikipedia/en/4/40/Chris_Benoit_Voice_Message.ogg'\n              );\n              const arrayBuffer = await response.arrayBuffer();\n              const transcription = await runtime.useModel(\n                ModelType.TRANSCRIPTION,\n                Buffer.from(new Uint8Array(arrayBuffer))\n              );\n              logger.log('generated with test_transcription:', transcription);\n            } catch (error) {\n              logger.error('Error in test_transcription:', error);\n              throw error;\n            }\n          },\n        },\n        {\n          name: 'openai_test_text_tokenizer_encode',\n          fn: async (runtime) => {\n            const prompt = 'Hello tokenizer encode!';\n            const tokens = await runtime.useModel(ModelType.TEXT_TOKENIZER_ENCODE, { prompt });\n            if (!Array.isArray(tokens) || tokens.length === 0) {\n              throw new Error('Failed to tokenize text: expected non-empty array of tokens');\n            }\n            logger.log('Tokenized output:', tokens);\n          },\n        },\n        {\n          name: 'openai_test_text_tokenizer_decode',\n          fn: async (runtime) => {\n            const prompt = 'Hello tokenizer decode!';\n            // Encode the string into tokens first\n            const tokens = await runtime.useModel(ModelType.TEXT_TOKENIZER_ENCODE, { prompt });\n            // Now decode tokens back into text\n            const decodedText = await runtime.useModel(ModelType.TEXT_TOKENIZER_DECODE, { tokens });\n            if (decodedText !== prompt) {\n              throw new Error(\n                `Decoded text does not match original. Expected \"${prompt}\", got \"${decodedText}\"`\n              );\n            }\n            logger.log('Decoded text:', decodedText);\n          },\n        },\n      ],\n    },\n  ],\n};\nexport default openaiPlugin;\n"],"mappings":";AAAA,SAAS,oBAAoB;AAQ7B;AAAA,EAGE;AAAA,EAEA;AAAA,OACK;AACP,SAAS,gBAAgB,oBAAoB;AAC7C,SAA6B,wBAAwB;AAUrD,eAAe,aAAa,OAAsB,QAAgB;AAChE,QAAM,YACJ,UAAU,UAAU,aACf,QAAQ,IAAI,sBAAsB,QAAQ,IAAI,eAAe,gBAC7D,QAAQ,IAAI,eAAe;AAClC,QAAM,WAAW,iBAAiB,SAA0B;AAC5D,QAAM,SAAS,SAAS,OAAO,MAAM;AACrC,SAAO;AACT;AASA,eAAe,eAAe,OAAsB,QAAkB;AACpE,QAAM,YACJ,UAAU,UAAU,aACf,QAAQ,IAAI,sBAAsB,QAAQ,IAAI,eAAe,gBAC7D,QAAQ,IAAI,sBAAsB,QAAQ,IAAI,eAAe;AACpE,QAAM,WAAW,iBAAiB,SAA0B;AAC5D,SAAO,SAAS,OAAO,MAAM;AAC/B;AAMO,IAAM,eAAuB;AAAA,EAClC,MAAM;AAAA,EACN,aAAa;AAAA,EACb,QAAQ;AAAA,IACN,gBAAgB,QAAQ,IAAI;AAAA,IAC5B,iBAAiB,QAAQ,IAAI;AAAA,IAC7B,oBAAoB,QAAQ,IAAI;AAAA,IAChC,oBAAoB,QAAQ,IAAI;AAAA,IAChC,aAAa,QAAQ,IAAI;AAAA,IACzB,aAAa,QAAQ,IAAI;AAAA,EAC3B;AAAA,EACA,MAAM,KAAK,QAAgC;AACzC,QAAI;AASF,UAAI,CAAC,QAAQ,IAAI,gBAAgB;AAC/B,eAAO;AAAA,UACL;AAAA,QACF;AAEA;AAAA,MACF;AAGA,UAAI;AACF,cAAM,UAAU,QAAQ,IAAI,mBAAmB;AAC/C,cAAM,WAAW,MAAM,MAAM,GAAG,OAAO,WAAW;AAAA,UAChD,SAAS,EAAE,eAAe,UAAU,QAAQ,IAAI,cAAc,GAAG;AAAA,QACnE,CAAC;AAED,YAAI,CAAC,SAAS,IAAI;AAChB,iBAAO,KAAK,qCAAqC,SAAS,UAAU,EAAE;AACtE,iBAAO,KAAK,wEAAwE;AAAA,QAEtF,OAAO;AAAA,QAEP;AAAA,MACF,SAAS,YAAY;AACnB,eAAO,KAAK,oCAAoC,UAAU,EAAE;AAC5D,eAAO,KAAK,wEAAwE;AAAA,MAEtF;AAAA,IACF,SAAS,OAAO;AAEd,aAAO;AAAA,QACL,sCAAsC,MAAM,OACzC,IAAI,CAAC,MAAM,EAAE,OAAO,EACpB,KAAK,IAAI,CAAC;AAAA,MACf;AAAA,IACF;AAAA,EACF;AAAA,EACA,QAAQ;AAAA,IACN,CAAC,UAAU,cAAc,GAAG,OAC1B,UACA,WACsB;AAEtB,UAAI,WAAW,MAAM;AACnB,eAAO,MAAM,4CAA4C;AAEzD,cAAM,aAAa,MAAM,IAAI,EAAE,KAAK,CAAC;AACrC,mBAAW,CAAC,IAAI;AAChB,eAAO;AAAA,MACT;AAGA,UAAI;AACJ,UAAI,OAAO,WAAW,UAAU;AAC9B,eAAO;AAAA,MACT,WAAW,OAAO,WAAW,YAAY,OAAO,MAAM;AACpD,eAAO,OAAO;AAAA,MAChB,OAAO;AACL,eAAO,KAAK,oCAAoC;AAEhD,cAAM,iBAAiB,MAAM,IAAI,EAAE,KAAK,CAAC;AACzC,uBAAe,CAAC,IAAI;AACpB,eAAO;AAAA,MACT;AAGA,UAAI,CAAC,KAAK,KAAK,GAAG;AAChB,eAAO,KAAK,0BAA0B;AACtC,cAAM,cAAc,MAAM,IAAI,EAAE,KAAK,CAAC;AACtC,oBAAY,CAAC,IAAI;AACjB,eAAO;AAAA,MACT;AAEA,UAAI;AACF,cAAM,UAAU,QAAQ,IAAI,mBAAmB;AAG/C,cAAM,WAAW,MAAM,MAAM,GAAG,OAAO,eAAe;AAAA,UACpD,QAAQ;AAAA,UACR,SAAS;AAAA,YACP,eAAe,UAAU,QAAQ,IAAI,cAAc;AAAA,YACnD,gBAAgB;AAAA,UAClB;AAAA,UACA,MAAM,KAAK,UAAU;AAAA,YACnB,OAAO;AAAA,YACP,OAAO;AAAA,UACT,CAAC;AAAA,QACH,CAAC;AAED,YAAI,CAAC,SAAS,IAAI;AAChB,iBAAO,MAAM,qBAAqB,SAAS,MAAM,MAAM,SAAS,UAAU,EAAE;AAC5E,gBAAM,cAAc,MAAM,IAAI,EAAE,KAAK,CAAC;AACtC,sBAAY,CAAC,IAAI;AACjB,iBAAO;AAAA,QACT;AAEA,cAAM,OAAQ,MAAM,SAAS,KAAK;AAIlC,YAAI,CAAC,MAAM,OAAO,CAAC,GAAG,WAAW;AAC/B,iBAAO,MAAM,gCAAgC;AAC7C,gBAAM,cAAc,MAAM,IAAI,EAAE,KAAK,CAAC;AACtC,sBAAY,CAAC,IAAI;AACjB,iBAAO;AAAA,QACT;AAEA,cAAM,YAAY,KAAK,KAAK,CAAC,EAAE;AAC/B,eAAO,IAAI,mCAAmC,UAAU,MAAM,EAAE;AAChE,eAAO;AAAA,MACT,SAAS,OAAO;AACd,eAAO,MAAM,+BAA+B,KAAK;AACjD,cAAM,cAAc,MAAM,IAAI,EAAE,KAAK,CAAC;AACtC,oBAAY,CAAC,IAAI;AACjB,eAAO;AAAA,MACT;AAAA,IACF;AAAA,IACA,CAAC,UAAU,qBAAqB,GAAG,OACjC,UACA,EAAE,QAAQ,YAAY,UAAU,WAAW,MACxC;AACH,aAAO,MAAM,aAAa,aAAa,UAAU,YAAY,MAAM;AAAA,IACrE;AAAA,IACA,CAAC,UAAU,qBAAqB,GAAG,OACjC,UACA,EAAE,QAAQ,YAAY,UAAU,WAAW,MACxC;AACH,aAAO,MAAM,eAAe,aAAa,UAAU,YAAY,MAAM;AAAA,IACvE;AAAA,IACA,CAAC,UAAU,UAAU,GAAG,OAAO,SAAS,EAAE,QAAQ,gBAAgB,CAAC,EAAE,MAA0B;AAC7F,YAAM,cAAc;AACpB,YAAM,oBAAoB;AAC1B,YAAM,mBAAmB;AACzB,YAAM,sBAAsB;AAE5B,YAAM,UAAU,QAAQ,WAAW,iBAAiB,KAAK;AAEzD,YAAM,SAAS,aAAa;AAAA,QAC1B,QAAQ,QAAQ,WAAW,gBAAgB;AAAA,QAC3C;AAAA,MACF,CAAC;AAED,YAAM,QACJ,QAAQ,WAAW,oBAAoB,KACvC,QAAQ,WAAW,aAAa,KAChC;AAEF,aAAO,IAAI,iBAAiB;AAC5B,aAAO,IAAI,MAAM;AAEjB,YAAM,EAAE,MAAM,eAAe,IAAI,MAAM,aAAa;AAAA,QAClD,OAAO,OAAO,cAAc,KAAK;AAAA,QACjC;AAAA,QACA,QAAQ,QAAQ,UAAU,UAAU;AAAA,QACpC;AAAA,QACA,WAAW;AAAA,QACX,kBAAkB;AAAA,QAClB,iBAAiB;AAAA,QACjB;AAAA,MACF,CAAC;AAED,aAAO;AAAA,IACT;AAAA,IACA,CAAC,UAAU,UAAU,GAAG,OACtB,SACA;AAAA,MACE;AAAA,MACA,gBAAgB,CAAC;AAAA,MACjB,YAAY;AAAA,MACZ,cAAc;AAAA,MACd,mBAAmB;AAAA,MACnB,kBAAkB;AAAA,IACpB,MACG;AACH,YAAM,UAAU,QAAQ,WAAW,iBAAiB,KAAK;AAEzD,YAAM,SAAS,aAAa;AAAA,QAC1B,QAAQ,QAAQ,WAAW,gBAAgB;AAAA,QAC3C;AAAA,MACF,CAAC;AAED,YAAM,QACJ,QAAQ,WAAW,oBAAoB,KAAK,QAAQ,WAAW,aAAa,KAAK;AAEnF,YAAM,EAAE,MAAM,eAAe,IAAI,MAAM,aAAa;AAAA,QAClD,OAAO,OAAO,cAAc,KAAK;AAAA,QACjC;AAAA,QACA,QAAQ,QAAQ,UAAU,UAAU;AAAA,QACpC;AAAA,QACA;AAAA,QACA;AAAA,QACA;AAAA,QACA;AAAA,MACF,CAAC;AAED,aAAO;AAAA,IACT;AAAA,IACA,CAAC,UAAU,KAAK,GAAG,OACjB,SACA,WAKG;AACH,YAAM,UAAU,QAAQ,WAAW,iBAAiB,KAAK;AACzD,YAAM,WAAW,MAAM,MAAM,GAAG,OAAO,uBAAuB;AAAA,QAC5D,QAAQ;AAAA,QACR,SAAS;AAAA,UACP,eAAe,UAAU,QAAQ,WAAW,gBAAgB,CAAC;AAAA,UAC7D,gBAAgB;AAAA,QAClB;AAAA,QACA,MAAM,KAAK,UAAU;AAAA,UACnB,QAAQ,OAAO;AAAA,UACf,GAAG,OAAO,KAAK;AAAA,UACf,MAAM,OAAO,QAAQ;AAAA,QACvB,CAAC;AAAA,MACH,CAAC;AACD,UAAI,CAAC,SAAS,IAAI;AAChB,cAAM,IAAI,MAAM,6BAA6B,SAAS,UAAU,EAAE;AAAA,MACpE;AACA,YAAM,OAAO,MAAM,SAAS,KAAK;AACjC,YAAM,YAAY;AAClB,aAAO,UAAU;AAAA,IACnB;AAAA,IACA,CAAC,UAAU,iBAAiB,GAAG,OAAO,SAAS,WAA4C;AAEzF,UAAI;AACJ,UAAI;AAEJ,UAAI,OAAO,WAAW,UAAU;AAC9B,mBAAW;AACX,iBAAS;AAAA,MACX,OAAO;AAEL,mBAAW,OAAO;AAClB,iBAAS,OAAO;AAAA,MAClB;AAEA,UAAI;AACF,cAAM,UAAU,QAAQ,IAAI,mBAAmB;AAC/C,cAAM,SAAS,QAAQ,IAAI;AAE3B,YAAI,CAAC,QAAQ;AACX,iBAAO,MAAM,wBAAwB;AACrC,iBAAO;AAAA,YACL,OAAO;AAAA,YACP,aAAa;AAAA,UACf;AAAA,QACF;AAGA,cAAM,WAAW,MAAM,MAAM,GAAG,OAAO,qBAAqB;AAAA,UAC1D,QAAQ;AAAA,UACR,SAAS;AAAA,YACP,gBAAgB;AAAA,YAChB,eAAe,UAAU,MAAM;AAAA,UACjC;AAAA,UACA,MAAM,KAAK,UAAU;AAAA,YACnB,OAAO;AAAA,YACP,UAAU;AAAA,cACR;AAAA,gBACE,MAAM;AAAA,gBACN,SAAS;AAAA,kBACP;AAAA,oBACE,MAAM;AAAA,oBACN,MACE,UACA;AAAA,kBACJ;AAAA,kBACA;AAAA,oBACE,MAAM;AAAA,oBACN,WAAW,EAAE,KAAK,SAAS;AAAA,kBAC7B;AAAA,gBACF;AAAA,cACF;AAAA,YACF;AAAA,YACA,YAAY;AAAA,UACd,CAAC;AAAA,QACH,CAAC;AAED,YAAI,CAAC,SAAS,IAAI;AAChB,gBAAM,IAAI,MAAM,qBAAqB,SAAS,MAAM,EAAE;AAAA,QACxD;AAEA,cAAM,SAAc,MAAM,SAAS,KAAK;AACxC,cAAM,UAAU,OAAO,UAAU,CAAC,GAAG,SAAS;AAE9C,YAAI,CAAC,SAAS;AACZ,iBAAO;AAAA,YACL,OAAO;AAAA,YACP,aAAa;AAAA,UACf;AAAA,QACF;AAGA,cAAM,aAAa,QAAQ,MAAM,2BAA2B;AAC5D,cAAM,QAAQ,aAAa,CAAC,KAAK;AAGjC,cAAM,cAAc,QAAQ,QAAQ,6BAA6B,EAAE,EAAE,KAAK;AAE1E,eAAO,EAAE,OAAO,YAAY;AAAA,MAC9B,SAAS,OAAO;AACd,eAAO,MAAM,0BAA0B,KAAK;AAC5C,eAAO;AAAA,UACL,OAAO;AAAA,UACP,aAAa,UAAU,iBAAiB,QAAQ,MAAM,UAAU,OAAO,KAAK,CAAC;AAAA,QAC/E;AAAA,MACF;AAAA,IACF;AAAA,IACA,CAAC,UAAU,aAAa,GAAG,OAAO,SAAS,gBAAwB;AACjE,aAAO,IAAI,eAAe,WAAW;AACrC,YAAM,UAAU,QAAQ,WAAW,iBAAiB,KAAK;AACzD,YAAM,WAAW,IAAI,SAAS;AAC9B,eAAS,OAAO,QAAQ,IAAI,KAAK,CAAC,WAAW,GAAG,EAAE,MAAM,YAAY,CAAC,CAAC;AACtE,eAAS,OAAO,SAAS,WAAW;AACpC,YAAM,WAAW,MAAM,MAAM,GAAG,OAAO,yBAAyB;AAAA,QAC9D,QAAQ;AAAA,QACR,SAAS;AAAA,UACP,eAAe,UAAU,QAAQ,WAAW,gBAAgB,CAAC;AAAA;AAAA,QAE/D;AAAA,QACA,MAAM;AAAA,MACR,CAAC;AAED,aAAO,IAAI,YAAY,QAAQ;AAC/B,UAAI,CAAC,SAAS,IAAI;AAChB,cAAM,IAAI,MAAM,+BAA+B,SAAS,UAAU,EAAE;AAAA,MACtE;AACA,YAAM,OAAQ,MAAM,SAAS,KAAK;AAClC,aAAO,KAAK;AAAA,IACd;AAAA,IACA,CAAC,UAAU,YAAY,GAAG,OAAO,SAAS,WAAmC;AAC3E,YAAM,UAAU,QAAQ,WAAW,iBAAiB,KAAK;AACzD,YAAM,SAAS,aAAa;AAAA,QAC1B,QAAQ,QAAQ,WAAW,gBAAgB;AAAA,QAC3C;AAAA,MACF,CAAC;AACD,YAAM,QACJ,QAAQ,WAAW,oBAAoB,KACvC,QAAQ,WAAW,aAAa,KAChC;AAEF,UAAI;AACF,YAAI,OAAO,QAAQ;AAEjB,iBAAO,KAAK,8CAA8C;AAC1D,gBAAM,EAAE,QAAAA,QAAO,IAAI,MAAM,eAAe;AAAA,YACtC,OAAO,OAAO,cAAc,KAAK;AAAA,YACjC,QAAQ;AAAA,YACR,QAAQ,OAAO;AAAA,YACf,aAAa,OAAO;AAAA,UACtB,CAAC;AACD,iBAAOA;AAAA,QACT;AAEA,cAAM,EAAE,OAAO,IAAI,MAAM,eAAe;AAAA,UACtC,OAAO,OAAO,cAAc,KAAK;AAAA,UACjC,QAAQ;AAAA,UACR,QAAQ,OAAO;AAAA,UACf,aAAa,OAAO;AAAA,QACtB,CAAC;AACD,eAAO;AAAA,MACT,SAAS,OAAO;AACd,eAAO,MAAM,4BAA4B,KAAK;AAC9C,cAAM;AAAA,MACR;AAAA,IACF;AAAA,IACA,CAAC,UAAU,YAAY,GAAG,OAAO,SAAS,WAAmC;AAC3E,YAAM,UAAU,QAAQ,WAAW,iBAAiB,KAAK;AACzD,YAAM,SAAS,aAAa;AAAA,QAC1B,QAAQ,QAAQ,WAAW,gBAAgB;AAAA,QAC3C;AAAA,MACF,CAAC;AACD,YAAM,QACJ,QAAQ,WAAW,oBAAoB,KAAK,QAAQ,WAAW,aAAa,KAAK;AAEnF,UAAI;AACF,YAAI,OAAO,QAAQ;AAEjB,iBAAO,KAAK,8CAA8C;AAC1D,gBAAM,EAAE,QAAAA,QAAO,IAAI,MAAM,eAAe;AAAA,YACtC,OAAO,OAAO,cAAc,KAAK;AAAA,YACjC,QAAQ;AAAA,YACR,QAAQ,OAAO;AAAA,YACf,aAAa,OAAO;AAAA,UACtB,CAAC;AACD,iBAAOA;AAAA,QACT;AAEA,cAAM,EAAE,OAAO,IAAI,MAAM,eAAe;AAAA,UACtC,OAAO,OAAO,cAAc,KAAK;AAAA,UACjC,QAAQ;AAAA,UACR,QAAQ,OAAO;AAAA,UACf,aAAa,OAAO;AAAA,QACtB,CAAC;AACD,eAAO;AAAA,MACT,SAAS,OAAO;AACd,eAAO,MAAM,4BAA4B,KAAK;AAC9C,cAAM;AAAA,MACR;AAAA,IACF;AAAA,EACF;AAAA,EACA,OAAO;AAAA,IACL;AAAA,MACE,MAAM;AAAA,MACN,OAAO;AAAA,QACL;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,kBAAM,UAAU,QAAQ,WAAW,iBAAiB,KAAK;AACzD,kBAAM,WAAW,MAAM,MAAM,GAAG,OAAO,WAAW;AAAA,cAChD,SAAS;AAAA,gBACP,eAAe,UAAU,QAAQ,WAAW,gBAAgB,CAAC;AAAA,cAC/D;AAAA,YACF,CAAC;AACD,kBAAM,OAAO,MAAM,SAAS,KAAK;AACjC,mBAAO,IAAI,qBAAsB,MAAc,KAAK,MAAM;AAC1D,gBAAI,CAAC,SAAS,IAAI;AAChB,oBAAM,IAAI,MAAM,sCAAsC,SAAS,UAAU,EAAE;AAAA,YAC7E;AAAA,UACF;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,gBAAI;AACF,oBAAM,YAAY,MAAM,QAAQ,SAAS,UAAU,gBAAgB;AAAA,gBACjE,MAAM;AAAA,cACR,CAAC;AACD,qBAAO,IAAI,aAAa,SAAS;AAAA,YACnC,SAAS,OAAO;AACd,qBAAO,MAAM,iCAAiC,KAAK;AACnD,oBAAM;AAAA,YACR;AAAA,UACF;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,gBAAI;AACF,oBAAM,OAAO,MAAM,QAAQ,SAAS,UAAU,YAAY;AAAA,gBACxD,QAAQ;AAAA,cACV,CAAC;AACD,kBAAI,KAAK,WAAW,GAAG;AACrB,sBAAM,IAAI,MAAM,yBAAyB;AAAA,cAC3C;AACA,qBAAO,IAAI,mCAAmC,IAAI;AAAA,YACpD,SAAS,OAAO;AACd,qBAAO,MAAM,6BAA6B,KAAK;AAC/C,oBAAM;AAAA,YACR;AAAA,UACF;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,gBAAI;AACF,oBAAM,OAAO,MAAM,QAAQ,SAAS,UAAU,YAAY;AAAA,gBACxD,QAAQ;AAAA,cACV,CAAC;AACD,kBAAI,KAAK,WAAW,GAAG;AACrB,sBAAM,IAAI,MAAM,yBAAyB;AAAA,cAC3C;AACA,qBAAO,IAAI,mCAAmC,IAAI;AAAA,YACpD,SAAS,OAAO;AACd,qBAAO,MAAM,6BAA6B,KAAK;AAC/C,oBAAM;AAAA,YACR;AAAA,UACF;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,mBAAO,IAAI,8BAA8B;AACzC,gBAAI;AACF,oBAAM,QAAQ,MAAM,QAAQ,SAAS,UAAU,OAAO;AAAA,gBACpD,QAAQ;AAAA,gBACR,GAAG;AAAA,gBACH,MAAM;AAAA,cACR,CAAC;AACD,qBAAO,IAAI,yCAAyC,KAAK;AAAA,YAC3D,SAAS,OAAO;AACd,qBAAO,MAAM,mCAAmC,KAAK;AACrD,oBAAM;AAAA,YACR;AAAA,UACF;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,gBAAI;AACF,qBAAO,IAAI,+BAA+B;AAC1C,kBAAI;AACF,sBAAM,SAAS,MAAM,QAAQ;AAAA,kBAC3B,UAAU;AAAA,kBACV;AAAA,gBACF;AAGA,oBACE,UACA,OAAO,WAAW,YAClB,WAAW,UACX,iBAAiB,QACjB;AACA,yBAAO,IAAI,sBAAsB,MAAM;AAAA,gBACzC,OAAO;AACL,yBAAO,MAAM,4CAA4C,MAAM;AAAA,gBACjE;AAAA,cACF,SAAS,GAAG;AACV,uBAAO,MAAM,oCAAoC,CAAC;AAAA,cACpD;AAAA,YACF,SAAS,GAAG;AACV,qBAAO,MAAM,2CAA2C,CAAC;AAAA,YAC3D;AAAA,UACF;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,mBAAO,IAAI,2BAA2B;AACtC,gBAAI;AACF,oBAAM,WAAW,MAAM;AAAA,gBACrB;AAAA,cACF;AACA,oBAAM,cAAc,MAAM,SAAS,YAAY;AAC/C,oBAAM,gBAAgB,MAAM,QAAQ;AAAA,gBAClC,UAAU;AAAA,gBACV,OAAO,KAAK,IAAI,WAAW,WAAW,CAAC;AAAA,cACzC;AACA,qBAAO,IAAI,sCAAsC,aAAa;AAAA,YAChE,SAAS,OAAO;AACd,qBAAO,MAAM,gCAAgC,KAAK;AAClD,oBAAM;AAAA,YACR;AAAA,UACF;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,kBAAM,SAAS;AACf,kBAAM,SAAS,MAAM,QAAQ,SAAS,UAAU,uBAAuB,EAAE,OAAO,CAAC;AACjF,gBAAI,CAAC,MAAM,QAAQ,MAAM,KAAK,OAAO,WAAW,GAAG;AACjD,oBAAM,IAAI,MAAM,6DAA6D;AAAA,YAC/E;AACA,mBAAO,IAAI,qBAAqB,MAAM;AAAA,UACxC;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,kBAAM,SAAS;AAEf,kBAAM,SAAS,MAAM,QAAQ,SAAS,UAAU,uBAAuB,EAAE,OAAO,CAAC;AAEjF,kBAAM,cAAc,MAAM,QAAQ,SAAS,UAAU,uBAAuB,EAAE,OAAO,CAAC;AACtF,gBAAI,gBAAgB,QAAQ;AAC1B,oBAAM,IAAI;AAAA,gBACR,mDAAmD,MAAM,WAAW,WAAW;AAAA,cACjF;AAAA,YACF;AACA,mBAAO,IAAI,iBAAiB,WAAW;AAAA,UACzC;AAAA,QACF;AAAA,MACF;AAAA,IACF;AAAA,EACF;AACF;AACA,IAAO,gBAAQ;","names":["object"]}
+{"version":3,"sources":["../src/index.ts"],"sourcesContent":["import { createOpenAI } from '@ai-sdk/openai';\nimport type {\n  AgentRuntime,\n  ImageDescriptionParams,\n  ModelTypeName,\n  ObjectGenerationParams,\n  Plugin,\n  TextEmbeddingParams,\n} from '@elizaos/core';\nimport {\n  type DetokenizeTextParams,\n  type GenerateTextParams,\n  ModelType,\n  type TokenizeTextParams,\n  logger,\n  VECTOR_DIMS,\n} from '@elizaos/core';\nimport { generateObject, generateText, JSONParseError, JSONValue } from 'ai';\nimport { type TiktokenModel, encodingForModel } from 'js-tiktoken';\nimport FormData from 'form-data';\nimport fetch from 'node-fetch';\n\n/**\n * Helper function to get settings with fallback to process.env\n *\n * @param runtime The runtime context\n * @param key The setting key to retrieve\n * @param defaultValue Optional default value if not found\n * @returns The setting value with proper fallbacks\n */\nfunction getSetting(runtime: any, key: string, defaultValue?: string): string | undefined {\n  return runtime.getSetting(key) ?? process.env[key] ?? defaultValue;\n}\n\n/**\n * Helper function to get the base URL for OpenAI API\n *\n * @param runtime The runtime context\n * @returns The configured base URL or default\n */\nfunction getBaseURL(runtime: any): string {\n  return getSetting(runtime, 'OPENAI_BASE_URL', 'https://api.openai.com/v1');\n}\n\n/**\n * Helper function to get the API key for OpenAI\n *\n * @param runtime The runtime context\n * @returns The configured API key\n */\nfunction getApiKey(runtime: any): string | undefined {\n  return getSetting(runtime, 'OPENAI_API_KEY');\n}\n\n/**\n * Helper function to get the small model name with fallbacks\n *\n * @param runtime The runtime context\n * @returns The configured small model name\n */\nfunction getSmallModel(runtime: any): string {\n  return (\n    getSetting(runtime, 'OPENAI_SMALL_MODEL') ?? getSetting(runtime, 'SMALL_MODEL', 'gpt-4o-mini')\n  );\n}\n\n/**\n * Helper function to get the large model name with fallbacks\n *\n * @param runtime The runtime context\n * @returns The configured large model name\n */\nfunction getLargeModel(runtime: any): string {\n  return getSetting(runtime, 'OPENAI_LARGE_MODEL') ?? getSetting(runtime, 'LARGE_MODEL', 'gpt-4o');\n}\n\n/**\n * Create an OpenAI client with proper configuration\n *\n * @param runtime The runtime context\n * @returns Configured OpenAI client\n */\nfunction createOpenAIClient(runtime: any) {\n  return createOpenAI({\n    apiKey: getApiKey(runtime),\n    baseURL: getBaseURL(runtime),\n  });\n}\n\n/**\n * Asynchronously tokenizes the given text based on the specified model and prompt.\n *\n * @param {ModelTypeName} model - The type of model to use for tokenization.\n * @param {string} prompt - The text prompt to tokenize.\n * @returns {number[]} - An array of tokens representing the encoded prompt.\n */\nasync function tokenizeText(model: ModelTypeName, prompt: string) {\n  const modelName =\n    model === ModelType.TEXT_SMALL\n      ? (process.env.OPENAI_SMALL_MODEL ?? process.env.SMALL_MODEL ?? 'gpt-4o-mini')\n      : (process.env.LARGE_MODEL ?? 'gpt-4o');\n  const encoding = encodingForModel(modelName as TiktokenModel);\n  const tokens = encoding.encode(prompt);\n  return tokens;\n}\n\n/**\n * Detokenize a sequence of tokens back into text using the specified model.\n *\n * @param {ModelTypeName} model - The type of model to use for detokenization.\n * @param {number[]} tokens - The sequence of tokens to detokenize.\n * @returns {string} The detokenized text.\n */\nasync function detokenizeText(model: ModelTypeName, tokens: number[]) {\n  const modelName =\n    model === ModelType.TEXT_SMALL\n      ? (process.env.OPENAI_SMALL_MODEL ?? process.env.SMALL_MODEL ?? 'gpt-4o-mini')\n      : (process.env.OPENAI_LARGE_MODEL ?? process.env.LARGE_MODEL ?? 'gpt-4o');\n  const encoding = encodingForModel(modelName as TiktokenModel);\n  return encoding.decode(tokens);\n}\n\n/**\n * Helper function to generate objects using specified model type\n */\nasync function generateObjectByModelType(\n  runtime: AgentRuntime,\n  params: ObjectGenerationParams,\n  modelType: string,\n  getModelFn: (runtime: AgentRuntime) => string\n): Promise<JSONValue> {\n  const openai = createOpenAIClient(runtime);\n  const model = getModelFn(runtime);\n\n  try {\n    if (params.schema) {\n      // Skip zod validation and just use the generateObject without schema\n      logger.info(`Using ${modelType} without schema validation`);\n    }\n\n    const { object } = await generateObject({\n      model: openai.languageModel(model),\n      output: 'no-schema',\n      prompt: params.prompt,\n      temperature: params.temperature,\n      experimental_repairText: getJsonRepairFunction(),\n    });\n    return object;\n  } catch (error) {\n    logger.error(`Error generating object with ${modelType}:`, error);\n    throw error;\n  }\n}\n\n/**\n * Returns a function to repair JSON text\n */\nfunction getJsonRepairFunction(): (params: {\n  text: string;\n  error: unknown;\n}) => Promise<string | null> {\n  return async ({ text, error }: { text: string; error: unknown }) => {\n    try {\n      if (error instanceof JSONParseError) {\n        const cleanedText = text.replace(/```json\\n|\\n```|```/g, '');\n\n        JSON.parse(cleanedText);\n        return cleanedText;\n      }\n    } catch (jsonError) {\n      logger.warn('Failed to repair JSON text:', jsonError);\n      return null;\n    }\n  };\n}\n\n/**\n * function for text-to-speech\n */\nasync function fetchTextToSpeech(runtime: AgentRuntime, text: string) {\n  const apiKey = getApiKey(runtime);\n  const model = getSetting(runtime, 'OPENAI_TTS_MODEL', 'gpt-4o-mini-tts');\n  const voice = getSetting(runtime, 'OPENAI_TTS_VOICE', 'nova');\n  const instructions = getSetting(runtime, 'OPENAI_TTS_INSTRUCTIONS', '');\n  const baseURL = getBaseURL(runtime);\n\n  try {\n    const res = await fetch(`${baseURL}/audio/speech`, {\n      method: 'POST',\n      headers: {\n        Authorization: `Bearer ${apiKey}`,\n        'Content-Type': 'application/json',\n      },\n      body: JSON.stringify({\n        model,\n        voice,\n        input: text,\n        ...(instructions && { instructions }),\n      }),\n    });\n\n    if (!res.ok) {\n      const err = await res.text();\n      throw new Error(`OpenAI TTS error ${res.status}: ${err}`);\n    }\n\n    return res.body;\n  } catch (err: any) {\n    throw new Error(`Failed to fetch speech from OpenAI TTS: ${err.message || err}`);\n  }\n}\n\n/**\n * Defines the OpenAI plugin with its name, description, and configuration options.\n * @type {Plugin}\n */\nexport const openaiPlugin: Plugin = {\n  name: 'openai',\n  description: 'OpenAI plugin',\n  config: {\n    OPENAI_API_KEY: process.env.OPENAI_API_KEY,\n    OPENAI_BASE_URL: process.env.OPENAI_BASE_URL,\n    OPENAI_SMALL_MODEL: process.env.OPENAI_SMALL_MODEL,\n    OPENAI_LARGE_MODEL: process.env.OPENAI_LARGE_MODEL,\n    SMALL_MODEL: process.env.SMALL_MODEL,\n    LARGE_MODEL: process.env.LARGE_MODEL,\n    OPENAI_EMBEDDING_MODEL: process.env.OPENAI_EMBEDDING_MODEL,\n    OPENAI_EMBEDDING_DIMENSIONS: process.env.OPENAI_EMBEDDING_DIMENSIONS,\n  },\n  async init(_config, runtime) {\n    try {\n      // const validatedConfig = await configSchema.parseAsync(config);\n\n      // // Set all environment variables at once\n      // for (const [key, value] of Object.entries(validatedConfig)) {\n      // \tif (value) process.env[key] = value;\n      // }\n\n      // If API key is not set, we'll show a warning but continue\n      if (!getApiKey(runtime)) {\n        logger.warn(\n          'OPENAI_API_KEY is not set in environment - OpenAI functionality will be limited'\n        );\n        // Return early without throwing an error\n        return;\n      }\n\n      // Verify API key only if we have one\n      try {\n        const baseURL = getBaseURL(runtime);\n        const response = await fetch(`${baseURL}/models`, {\n          headers: { Authorization: `Bearer ${getApiKey(runtime)}` },\n        });\n\n        if (!response.ok) {\n          logger.warn(`OpenAI API key validation failed: ${response.statusText}`);\n          logger.warn('OpenAI functionality will be limited until a valid API key is provided');\n          // Continue execution instead of throwing\n        } else {\n          // logger.log(\"OpenAI API key validated successfully\");\n        }\n      } catch (fetchError) {\n        logger.warn(`Error validating OpenAI API key: ${fetchError}`);\n        logger.warn('OpenAI functionality will be limited until a valid API key is provided');\n        // Continue execution instead of throwing\n      }\n    } catch (error) {\n      // Convert to warning instead of error\n      logger.warn(\n        `OpenAI plugin configuration issue: ${error.errors\n          .map((e) => e.message)\n          .join(', ')} - You need to configure the OPENAI_API_KEY in your environment variables`\n      );\n    }\n  },\n  models: {\n    [ModelType.TEXT_EMBEDDING]: async (\n      runtime,\n      params: TextEmbeddingParams | string | null\n    ): Promise<number[]> => {\n      const embeddingDimension = parseInt(\n        getSetting(runtime, 'OPENAI_EMBEDDING_DIMENSIONS', '1536')\n      ) as (typeof VECTOR_DIMS)[keyof typeof VECTOR_DIMS];\n\n      // Validate embedding dimension\n      if (!Object.values(VECTOR_DIMS).includes(embeddingDimension)) {\n        logger.error(\n          `Invalid embedding dimension: ${embeddingDimension}. Must be one of: ${Object.values(VECTOR_DIMS).join(', ')}`\n        );\n        throw new Error(\n          `Invalid embedding dimension: ${embeddingDimension}. Must be one of: ${Object.values(VECTOR_DIMS).join(', ')}`\n        );\n      }\n\n      // Handle null input (initialization case)\n      if (params === null) {\n        logger.debug('Creating test embedding for initialization');\n        // Return a consistent vector for null input\n        const testVector = Array(embeddingDimension).fill(0);\n        testVector[0] = 0.1; // Make it non-zero\n        return testVector;\n      }\n\n      // Get the text from whatever format was provided\n      let text: string;\n      if (typeof params === 'string') {\n        text = params; // Direct string input\n      } else if (typeof params === 'object' && params.text) {\n        text = params.text; // Object with text property\n      } else {\n        logger.warn('Invalid input format for embedding');\n        // Return a fallback for invalid input\n        const fallbackVector = Array(embeddingDimension).fill(0);\n        fallbackVector[0] = 0.2; // Different value for tracking\n        return fallbackVector;\n      }\n\n      // Skip API call for empty text\n      if (!text.trim()) {\n        logger.warn('Empty text for embedding');\n        const emptyVector = Array(embeddingDimension).fill(0);\n        emptyVector[0] = 0.3; // Different value for tracking\n        return emptyVector;\n      }\n\n      try {\n        const baseURL = getBaseURL(runtime);\n\n        // Call the OpenAI API\n        const response = await fetch(`${baseURL}/embeddings`, {\n          method: 'POST',\n          headers: {\n            Authorization: `Bearer ${getApiKey(runtime)}`,\n            'Content-Type': 'application/json',\n          },\n          body: JSON.stringify({\n            model: getSetting(runtime, 'OPENAI_EMBEDDING_MODEL', 'text-embedding-3-small'),\n            input: text,\n          }),\n        });\n\n        if (!response.ok) {\n          logger.error(`OpenAI API error: ${response.status} - ${response.statusText}`);\n          const errorVector = Array(embeddingDimension).fill(0);\n          errorVector[0] = 0.4; // Different value for tracking\n          return errorVector;\n        }\n\n        const data = (await response.json()) as {\n          data: [{ embedding: number[] }];\n        };\n\n        if (!data?.data?.[0]?.embedding) {\n          logger.error('API returned invalid structure');\n          const errorVector = Array(embeddingDimension).fill(0);\n          errorVector[0] = 0.5; // Different value for tracking\n          return errorVector;\n        }\n\n        const embedding = data.data[0].embedding;\n        logger.log(`Got valid embedding with length ${embedding.length}`);\n        return embedding;\n      } catch (error) {\n        logger.error('Error generating embedding:', error);\n        const errorVector = Array(embeddingDimension).fill(0);\n        errorVector[0] = 0.6; // Different value for tracking\n        return errorVector;\n      }\n    },\n    [ModelType.TEXT_TOKENIZER_ENCODE]: async (\n      _runtime,\n      { prompt, modelType = ModelType.TEXT_LARGE }: TokenizeTextParams\n    ) => {\n      return await tokenizeText(modelType ?? ModelType.TEXT_LARGE, prompt);\n    },\n    [ModelType.TEXT_TOKENIZER_DECODE]: async (\n      _runtime,\n      { tokens, modelType = ModelType.TEXT_LARGE }: DetokenizeTextParams\n    ) => {\n      return await detokenizeText(modelType ?? ModelType.TEXT_LARGE, tokens);\n    },\n    [ModelType.TEXT_SMALL]: async (runtime, { prompt, stopSequences = [] }: GenerateTextParams) => {\n      const temperature = 0.7;\n      const frequency_penalty = 0.7;\n      const presence_penalty = 0.7;\n      const max_response_length = 8192;\n\n      const openai = createOpenAIClient(runtime);\n      const model = getSmallModel(runtime);\n\n      logger.log('generating text');\n      logger.log(prompt);\n\n      const { text: openaiResponse } = await generateText({\n        model: openai.languageModel(model),\n        prompt: prompt,\n        system: runtime.character.system ?? undefined,\n        temperature: temperature,\n        maxTokens: max_response_length,\n        frequencyPenalty: frequency_penalty,\n        presencePenalty: presence_penalty,\n        stopSequences: stopSequences,\n      });\n\n      return openaiResponse;\n    },\n    [ModelType.TEXT_LARGE]: async (\n      runtime,\n      {\n        prompt,\n        stopSequences = [],\n        maxTokens = 8192,\n        temperature = 0.7,\n        frequencyPenalty = 0.7,\n        presencePenalty = 0.7,\n      }: GenerateTextParams\n    ) => {\n      const openai = createOpenAIClient(runtime);\n      const model = getLargeModel(runtime);\n\n      const { text: openaiResponse } = await generateText({\n        model: openai.languageModel(model),\n        prompt: prompt,\n        system: runtime.character.system ?? undefined,\n        temperature: temperature,\n        maxTokens: maxTokens,\n        frequencyPenalty: frequencyPenalty,\n        presencePenalty: presencePenalty,\n        stopSequences: stopSequences,\n      });\n\n      return openaiResponse;\n    },\n    [ModelType.IMAGE]: async (\n      runtime,\n      params: {\n        prompt: string;\n        n?: number;\n        size?: string;\n      }\n    ) => {\n      const baseURL = getBaseURL(runtime);\n      const response = await fetch(`${baseURL}/images/generations`, {\n        method: 'POST',\n        headers: {\n          Authorization: `Bearer ${getApiKey(runtime)}`,\n          'Content-Type': 'application/json',\n        },\n        body: JSON.stringify({\n          prompt: params.prompt,\n          n: params.n || 1,\n          size: params.size || '1024x1024',\n        }),\n      });\n      if (!response.ok) {\n        throw new Error(`Failed to generate image: ${response.statusText}`);\n      }\n      const data = await response.json();\n      const typedData = data as { data: { url: string }[] };\n      return typedData.data;\n    },\n    [ModelType.IMAGE_DESCRIPTION]: async (runtime, params: ImageDescriptionParams | string) => {\n      // Handle string case (direct URL)\n      let imageUrl: string;\n      let prompt: string | undefined;\n\n      if (typeof params === 'string') {\n        imageUrl = params;\n        prompt = undefined;\n      } else {\n        // Object parameter case\n        imageUrl = params.imageUrl;\n        prompt = params.prompt;\n      }\n\n      try {\n        const baseURL = getBaseURL(runtime);\n        const apiKey = getApiKey(runtime);\n\n        if (!apiKey) {\n          logger.error('OpenAI API key not set');\n          return {\n            title: 'Failed to analyze image',\n            description: 'API key not configured',\n          };\n        }\n\n        // Call the GPT-4 Vision API\n        const response = await fetch(`${baseURL}/chat/completions`, {\n          method: 'POST',\n          headers: {\n            'Content-Type': 'application/json',\n            Authorization: `Bearer ${apiKey}`,\n          },\n          body: JSON.stringify({\n            model: 'gpt-4o-mini',\n            messages: [\n              {\n                role: 'user',\n                content: [\n                  {\n                    type: 'text',\n                    text:\n                      prompt ||\n                      'Please analyze this image and provide a title and detailed description.',\n                  },\n                  {\n                    type: 'image_url',\n                    image_url: { url: imageUrl },\n                  },\n                ],\n              },\n            ],\n            max_tokens: 300,\n          }),\n        });\n\n        if (!response.ok) {\n          throw new Error(`OpenAI API error: ${response.status}`);\n        }\n\n        const result: any = await response.json();\n        const content = result.choices?.[0]?.message?.content;\n\n        if (!content) {\n          return {\n            title: 'Failed to analyze image',\n            description: 'No response from API',\n          };\n        }\n\n        // Extract title and description\n        const titleMatch = content.match(/title[:\\s]+(.+?)(?:\\n|$)/i);\n        const title = titleMatch?.[1] || 'Image Analysis';\n\n        // Rest of content is the description\n        const description = content.replace(/title[:\\s]+(.+?)(?:\\n|$)/i, '').trim();\n\n        return { title, description };\n      } catch (error) {\n        logger.error('Error analyzing image:', error);\n        return {\n          title: 'Failed to analyze image',\n          description: `Error: ${error instanceof Error ? error.message : String(error)}`,\n        };\n      }\n    },\n    [ModelType.TRANSCRIPTION]: async (runtime, audioBuffer: Buffer) => {\n      logger.log('audioBuffer', audioBuffer);\n      const baseURL = getBaseURL(runtime);\n\n      const formData = new FormData();\n      formData.append('file', audioBuffer, {\n        filename: 'recording.mp3',\n        contentType: 'audio/mp3',\n      });\n      formData.append('model', 'whisper-1');\n\n      const response = await fetch(`${baseURL}/audio/transcriptions`, {\n        method: 'POST',\n        headers: {\n          Authorization: `Bearer ${getApiKey(runtime)}`,\n        },\n        body: formData,\n      });\n\n      logger.log('response', response);\n      if (!response.ok) {\n        throw new Error(`Failed to transcribe audio: ${response.statusText}`);\n      }\n\n      const data = (await response.json()) as { text: string };\n      return data.text;\n    },\n    [ModelType.TEXT_TO_SPEECH]: async (runtime: AgentRuntime, text: string) => {\n      return await fetchTextToSpeech(runtime, text);\n    },\n\n    [ModelType.OBJECT_SMALL]: async (runtime, params: ObjectGenerationParams) => {\n      return generateObjectByModelType(runtime, params, ModelType.OBJECT_SMALL, getSmallModel);\n    },\n    [ModelType.OBJECT_LARGE]: async (runtime, params: ObjectGenerationParams) => {\n      return generateObjectByModelType(runtime, params, ModelType.OBJECT_LARGE, getLargeModel);\n    },\n  },\n  tests: [\n    {\n      name: 'openai_plugin_tests',\n      tests: [\n        {\n          name: 'openai_test_url_and_api_key_validation',\n          fn: async (runtime) => {\n            const baseURL = getBaseURL(runtime);\n            const response = await fetch(`${baseURL}/models`, {\n              headers: {\n                Authorization: `Bearer ${getApiKey(runtime)}`,\n              },\n            });\n            const data = await response.json();\n            logger.log('Models Available:', (data as any)?.data.length);\n            if (!response.ok) {\n              throw new Error(`Failed to validate OpenAI API key: ${response.statusText}`);\n            }\n          },\n        },\n        {\n          name: 'openai_test_text_embedding',\n          fn: async (runtime) => {\n            try {\n              const embedding = await runtime.useModel(ModelType.TEXT_EMBEDDING, {\n                text: 'Hello, world!',\n              });\n              logger.log('embedding', embedding);\n            } catch (error) {\n              logger.error('Error in test_text_embedding:', error);\n              throw error;\n            }\n          },\n        },\n        {\n          name: 'openai_test_text_large',\n          fn: async (runtime) => {\n            try {\n              const text = await runtime.useModel(ModelType.TEXT_LARGE, {\n                prompt: 'What is the nature of reality in 10 words?',\n              });\n              if (text.length === 0) {\n                throw new Error('Failed to generate text');\n              }\n              logger.log('generated with test_text_large:', text);\n            } catch (error) {\n              logger.error('Error in test_text_large:', error);\n              throw error;\n            }\n          },\n        },\n        {\n          name: 'openai_test_text_small',\n          fn: async (runtime) => {\n            try {\n              const text = await runtime.useModel(ModelType.TEXT_SMALL, {\n                prompt: 'What is the nature of reality in 10 words?',\n              });\n              if (text.length === 0) {\n                throw new Error('Failed to generate text');\n              }\n              logger.log('generated with test_text_small:', text);\n            } catch (error) {\n              logger.error('Error in test_text_small:', error);\n              throw error;\n            }\n          },\n        },\n        {\n          name: 'openai_test_image_generation',\n          fn: async (runtime) => {\n            logger.log('openai_test_image_generation');\n            try {\n              const image = await runtime.useModel(ModelType.IMAGE, {\n                prompt: 'A beautiful sunset over a calm ocean',\n                n: 1,\n                size: '1024x1024',\n              });\n              logger.log('generated with test_image_generation:', image);\n            } catch (error) {\n              logger.error('Error in test_image_generation:', error);\n              throw error;\n            }\n          },\n        },\n        {\n          name: 'image-description',\n          fn: async (runtime) => {\n            try {\n              logger.log('openai_test_image_description');\n              try {\n                const result = await runtime.useModel(\n                  ModelType.IMAGE_DESCRIPTION,\n                  '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'\n                );\n\n                // Check if result has the expected structure\n                if (\n                  result &&\n                  typeof result === 'object' &&\n                  'title' in result &&\n                  'description' in result\n                ) {\n                  logger.log('Image description:', result);\n                } else {\n                  logger.error('Invalid image description result format:', result);\n                }\n              } catch (e) {\n                logger.error('Error in image description test:', e);\n              }\n            } catch (e) {\n              logger.error('Error in openai_test_image_description:', e);\n            }\n          },\n        },\n        {\n          name: 'openai_test_transcription',\n          fn: async (runtime) => {\n            logger.log('openai_test_transcription');\n            try {\n              const response = await fetch(\n                'https://upload.wikimedia.org/wikipedia/en/4/40/Chris_Benoit_Voice_Message.ogg'\n              );\n              const arrayBuffer = await response.arrayBuffer();\n              const transcription = await runtime.useModel(\n                ModelType.TRANSCRIPTION,\n                Buffer.from(new Uint8Array(arrayBuffer))\n              );\n              logger.log('generated with test_transcription:', transcription);\n            } catch (error) {\n              logger.error('Error in test_transcription:', error);\n              throw error;\n            }\n          },\n        },\n        {\n          name: 'openai_test_text_tokenizer_encode',\n          fn: async (runtime) => {\n            const prompt = 'Hello tokenizer encode!';\n            const tokens = await runtime.useModel(ModelType.TEXT_TOKENIZER_ENCODE, { prompt });\n            if (!Array.isArray(tokens) || tokens.length === 0) {\n              throw new Error('Failed to tokenize text: expected non-empty array of tokens');\n            }\n            logger.log('Tokenized output:', tokens);\n          },\n        },\n        {\n          name: 'openai_test_text_tokenizer_decode',\n          fn: async (runtime) => {\n            const prompt = 'Hello tokenizer decode!';\n            // Encode the string into tokens first\n            const tokens = await runtime.useModel(ModelType.TEXT_TOKENIZER_ENCODE, { prompt });\n            // Now decode tokens back into text\n            const decodedText = await runtime.useModel(ModelType.TEXT_TOKENIZER_DECODE, { tokens });\n            if (decodedText !== prompt) {\n              throw new Error(\n                `Decoded text does not match original. Expected \"${prompt}\", got \"${decodedText}\"`\n              );\n            }\n            logger.log('Decoded text:', decodedText);\n          },\n        },\n        {\n          name: 'openai_test_text_to_speech',\n          fn: async (runtime: AgentRuntime) => {\n            try {\n              const text = 'Hello, this is a test for text-to-speech.';\n              const response = await fetchTextToSpeech(runtime, text);\n              if (!response) {\n                throw new Error('Failed to generate speech');\n              }\n              logger.log('Generated speech successfully');\n            } catch (error) {\n              logger.error('Error in openai_test_text_to_speech:', error);\n              throw error;\n            }\n          },\n        },\n      ],\n    },\n  ],\n};\nexport default openaiPlugin;\n"],"mappings":";AAAA,SAAS,oBAAoB;AAS7B;AAAA,EAGE;AAAA,EAEA;AAAA,EACA;AAAA,OACK;AACP,SAAS,gBAAgB,cAAc,sBAAiC;AACxE,SAA6B,wBAAwB;AACrD,OAAO,cAAc;AACrB,OAAO,WAAW;AAUlB,SAAS,WAAW,SAAc,KAAa,cAA2C;AACxF,SAAO,QAAQ,WAAW,GAAG,KAAK,QAAQ,IAAI,GAAG,KAAK;AACxD;AAQA,SAAS,WAAW,SAAsB;AACxC,SAAO,WAAW,SAAS,mBAAmB,2BAA2B;AAC3E;AAQA,SAAS,UAAU,SAAkC;AACnD,SAAO,WAAW,SAAS,gBAAgB;AAC7C;AAQA,SAAS,cAAc,SAAsB;AAC3C,SACE,WAAW,SAAS,oBAAoB,KAAK,WAAW,SAAS,eAAe,aAAa;AAEjG;AAQA,SAAS,cAAc,SAAsB;AAC3C,SAAO,WAAW,SAAS,oBAAoB,KAAK,WAAW,SAAS,eAAe,QAAQ;AACjG;AAQA,SAAS,mBAAmB,SAAc;AACxC,SAAO,aAAa;AAAA,IAClB,QAAQ,UAAU,OAAO;AAAA,IACzB,SAAS,WAAW,OAAO;AAAA,EAC7B,CAAC;AACH;AASA,eAAe,aAAa,OAAsB,QAAgB;AAChE,QAAM,YACJ,UAAU,UAAU,aACf,QAAQ,IAAI,sBAAsB,QAAQ,IAAI,eAAe,gBAC7D,QAAQ,IAAI,eAAe;AAClC,QAAM,WAAW,iBAAiB,SAA0B;AAC5D,QAAM,SAAS,SAAS,OAAO,MAAM;AACrC,SAAO;AACT;AASA,eAAe,eAAe,OAAsB,QAAkB;AACpE,QAAM,YACJ,UAAU,UAAU,aACf,QAAQ,IAAI,sBAAsB,QAAQ,IAAI,eAAe,gBAC7D,QAAQ,IAAI,sBAAsB,QAAQ,IAAI,eAAe;AACpE,QAAM,WAAW,iBAAiB,SAA0B;AAC5D,SAAO,SAAS,OAAO,MAAM;AAC/B;AAKA,eAAe,0BACb,SACA,QACA,WACA,YACoB;AACpB,QAAM,SAAS,mBAAmB,OAAO;AACzC,QAAM,QAAQ,WAAW,OAAO;AAEhC,MAAI;AACF,QAAI,OAAO,QAAQ;AAEjB,aAAO,KAAK,SAAS,SAAS,4BAA4B;AAAA,IAC5D;AAEA,UAAM,EAAE,OAAO,IAAI,MAAM,eAAe;AAAA,MACtC,OAAO,OAAO,cAAc,KAAK;AAAA,MACjC,QAAQ;AAAA,MACR,QAAQ,OAAO;AAAA,MACf,aAAa,OAAO;AAAA,MACpB,yBAAyB,sBAAsB;AAAA,IACjD,CAAC;AACD,WAAO;AAAA,EACT,SAAS,OAAO;AACd,WAAO,MAAM,gCAAgC,SAAS,KAAK,KAAK;AAChE,UAAM;AAAA,EACR;AACF;AAKA,SAAS,wBAGoB;AAC3B,SAAO,OAAO,EAAE,MAAM,MAAM,MAAwC;AAClE,QAAI;AACF,UAAI,iBAAiB,gBAAgB;AACnC,cAAM,cAAc,KAAK,QAAQ,wBAAwB,EAAE;AAE3D,aAAK,MAAM,WAAW;AACtB,eAAO;AAAA,MACT;AAAA,IACF,SAAS,WAAW;AAClB,aAAO,KAAK,+BAA+B,SAAS;AACpD,aAAO;AAAA,IACT;AAAA,EACF;AACF;AAKA,eAAe,kBAAkB,SAAuB,MAAc;AACpE,QAAM,SAAS,UAAU,OAAO;AAChC,QAAM,QAAQ,WAAW,SAAS,oBAAoB,iBAAiB;AACvE,QAAM,QAAQ,WAAW,SAAS,oBAAoB,MAAM;AAC5D,QAAM,eAAe,WAAW,SAAS,2BAA2B,EAAE;AACtE,QAAM,UAAU,WAAW,OAAO;AAElC,MAAI;AACF,UAAM,MAAM,MAAM,MAAM,GAAG,OAAO,iBAAiB;AAAA,MACjD,QAAQ;AAAA,MACR,SAAS;AAAA,QACP,eAAe,UAAU,MAAM;AAAA,QAC/B,gBAAgB;AAAA,MAClB;AAAA,MACA,MAAM,KAAK,UAAU;AAAA,QACnB;AAAA,QACA;AAAA,QACA,OAAO;AAAA,QACP,GAAI,gBAAgB,EAAE,aAAa;AAAA,MACrC,CAAC;AAAA,IACH,CAAC;AAED,QAAI,CAAC,IAAI,IAAI;AACX,YAAM,MAAM,MAAM,IAAI,KAAK;AAC3B,YAAM,IAAI,MAAM,oBAAoB,IAAI,MAAM,KAAK,GAAG,EAAE;AAAA,IAC1D;AAEA,WAAO,IAAI;AAAA,EACb,SAAS,KAAU;AACjB,UAAM,IAAI,MAAM,2CAA2C,IAAI,WAAW,GAAG,EAAE;AAAA,EACjF;AACF;AAMO,IAAM,eAAuB;AAAA,EAClC,MAAM;AAAA,EACN,aAAa;AAAA,EACb,QAAQ;AAAA,IACN,gBAAgB,QAAQ,IAAI;AAAA,IAC5B,iBAAiB,QAAQ,IAAI;AAAA,IAC7B,oBAAoB,QAAQ,IAAI;AAAA,IAChC,oBAAoB,QAAQ,IAAI;AAAA,IAChC,aAAa,QAAQ,IAAI;AAAA,IACzB,aAAa,QAAQ,IAAI;AAAA,IACzB,wBAAwB,QAAQ,IAAI;AAAA,IACpC,6BAA6B,QAAQ,IAAI;AAAA,EAC3C;AAAA,EACA,MAAM,KAAK,SAAS,SAAS;AAC3B,QAAI;AASF,UAAI,CAAC,UAAU,OAAO,GAAG;AACvB,eAAO;AAAA,UACL;AAAA,QACF;AAEA;AAAA,MACF;AAGA,UAAI;AACF,cAAM,UAAU,WAAW,OAAO;AAClC,cAAM,WAAW,MAAM,MAAM,GAAG,OAAO,WAAW;AAAA,UAChD,SAAS,EAAE,eAAe,UAAU,UAAU,OAAO,CAAC,GAAG;AAAA,QAC3D,CAAC;AAED,YAAI,CAAC,SAAS,IAAI;AAChB,iBAAO,KAAK,qCAAqC,SAAS,UAAU,EAAE;AACtE,iBAAO,KAAK,wEAAwE;AAAA,QAEtF,OAAO;AAAA,QAEP;AAAA,MACF,SAAS,YAAY;AACnB,eAAO,KAAK,oCAAoC,UAAU,EAAE;AAC5D,eAAO,KAAK,wEAAwE;AAAA,MAEtF;AAAA,IACF,SAAS,OAAO;AAEd,aAAO;AAAA,QACL,sCAAsC,MAAM,OACzC,IAAI,CAAC,MAAM,EAAE,OAAO,EACpB,KAAK,IAAI,CAAC;AAAA,MACf;AAAA,IACF;AAAA,EACF;AAAA,EACA,QAAQ;AAAA,IACN,CAAC,UAAU,cAAc,GAAG,OAC1B,SACA,WACsB;AACtB,YAAM,qBAAqB;AAAA,QACzB,WAAW,SAAS,+BAA+B,MAAM;AAAA,MAC3D;AAGA,UAAI,CAAC,OAAO,OAAO,WAAW,EAAE,SAAS,kBAAkB,GAAG;AAC5D,eAAO;AAAA,UACL,gCAAgC,kBAAkB,qBAAqB,OAAO,OAAO,WAAW,EAAE,KAAK,IAAI,CAAC;AAAA,QAC9G;AACA,cAAM,IAAI;AAAA,UACR,gCAAgC,kBAAkB,qBAAqB,OAAO,OAAO,WAAW,EAAE,KAAK,IAAI,CAAC;AAAA,QAC9G;AAAA,MACF;AAGA,UAAI,WAAW,MAAM;AACnB,eAAO,MAAM,4CAA4C;AAEzD,cAAM,aAAa,MAAM,kBAAkB,EAAE,KAAK,CAAC;AACnD,mBAAW,CAAC,IAAI;AAChB,eAAO;AAAA,MACT;AAGA,UAAI;AACJ,UAAI,OAAO,WAAW,UAAU;AAC9B,eAAO;AAAA,MACT,WAAW,OAAO,WAAW,YAAY,OAAO,MAAM;AACpD,eAAO,OAAO;AAAA,MAChB,OAAO;AACL,eAAO,KAAK,oCAAoC;AAEhD,cAAM,iBAAiB,MAAM,kBAAkB,EAAE,KAAK,CAAC;AACvD,uBAAe,CAAC,IAAI;AACpB,eAAO;AAAA,MACT;AAGA,UAAI,CAAC,KAAK,KAAK,GAAG;AAChB,eAAO,KAAK,0BAA0B;AACtC,cAAM,cAAc,MAAM,kBAAkB,EAAE,KAAK,CAAC;AACpD,oBAAY,CAAC,IAAI;AACjB,eAAO;AAAA,MACT;AAEA,UAAI;AACF,cAAM,UAAU,WAAW,OAAO;AAGlC,cAAM,WAAW,MAAM,MAAM,GAAG,OAAO,eAAe;AAAA,UACpD,QAAQ;AAAA,UACR,SAAS;AAAA,YACP,eAAe,UAAU,UAAU,OAAO,CAAC;AAAA,YAC3C,gBAAgB;AAAA,UAClB;AAAA,UACA,MAAM,KAAK,UAAU;AAAA,YACnB,OAAO,WAAW,SAAS,0BAA0B,wBAAwB;AAAA,YAC7E,OAAO;AAAA,UACT,CAAC;AAAA,QACH,CAAC;AAED,YAAI,CAAC,SAAS,IAAI;AAChB,iBAAO,MAAM,qBAAqB,SAAS,MAAM,MAAM,SAAS,UAAU,EAAE;AAC5E,gBAAM,cAAc,MAAM,kBAAkB,EAAE,KAAK,CAAC;AACpD,sBAAY,CAAC,IAAI;AACjB,iBAAO;AAAA,QACT;AAEA,cAAM,OAAQ,MAAM,SAAS,KAAK;AAIlC,YAAI,CAAC,MAAM,OAAO,CAAC,GAAG,WAAW;AAC/B,iBAAO,MAAM,gCAAgC;AAC7C,gBAAM,cAAc,MAAM,kBAAkB,EAAE,KAAK,CAAC;AACpD,sBAAY,CAAC,IAAI;AACjB,iBAAO;AAAA,QACT;AAEA,cAAM,YAAY,KAAK,KAAK,CAAC,EAAE;AAC/B,eAAO,IAAI,mCAAmC,UAAU,MAAM,EAAE;AAChE,eAAO;AAAA,MACT,SAAS,OAAO;AACd,eAAO,MAAM,+BAA+B,KAAK;AACjD,cAAM,cAAc,MAAM,kBAAkB,EAAE,KAAK,CAAC;AACpD,oBAAY,CAAC,IAAI;AACjB,eAAO;AAAA,MACT;AAAA,IACF;AAAA,IACA,CAAC,UAAU,qBAAqB,GAAG,OACjC,UACA,EAAE,QAAQ,YAAY,UAAU,WAAW,MACxC;AACH,aAAO,MAAM,aAAa,aAAa,UAAU,YAAY,MAAM;AAAA,IACrE;AAAA,IACA,CAAC,UAAU,qBAAqB,GAAG,OACjC,UACA,EAAE,QAAQ,YAAY,UAAU,WAAW,MACxC;AACH,aAAO,MAAM,eAAe,aAAa,UAAU,YAAY,MAAM;AAAA,IACvE;AAAA,IACA,CAAC,UAAU,UAAU,GAAG,OAAO,SAAS,EAAE,QAAQ,gBAAgB,CAAC,EAAE,MAA0B;AAC7F,YAAM,cAAc;AACpB,YAAM,oBAAoB;AAC1B,YAAM,mBAAmB;AACzB,YAAM,sBAAsB;AAE5B,YAAM,SAAS,mBAAmB,OAAO;AACzC,YAAM,QAAQ,cAAc,OAAO;AAEnC,aAAO,IAAI,iBAAiB;AAC5B,aAAO,IAAI,MAAM;AAEjB,YAAM,EAAE,MAAM,eAAe,IAAI,MAAM,aAAa;AAAA,QAClD,OAAO,OAAO,cAAc,KAAK;AAAA,QACjC;AAAA,QACA,QAAQ,QAAQ,UAAU,UAAU;AAAA,QACpC;AAAA,QACA,WAAW;AAAA,QACX,kBAAkB;AAAA,QAClB,iBAAiB;AAAA,QACjB;AAAA,MACF,CAAC;AAED,aAAO;AAAA,IACT;AAAA,IACA,CAAC,UAAU,UAAU,GAAG,OACtB,SACA;AAAA,MACE;AAAA,MACA,gBAAgB,CAAC;AAAA,MACjB,YAAY;AAAA,MACZ,cAAc;AAAA,MACd,mBAAmB;AAAA,MACnB,kBAAkB;AAAA,IACpB,MACG;AACH,YAAM,SAAS,mBAAmB,OAAO;AACzC,YAAM,QAAQ,cAAc,OAAO;AAEnC,YAAM,EAAE,MAAM,eAAe,IAAI,MAAM,aAAa;AAAA,QAClD,OAAO,OAAO,cAAc,KAAK;AAAA,QACjC;AAAA,QACA,QAAQ,QAAQ,UAAU,UAAU;AAAA,QACpC;AAAA,QACA;AAAA,QACA;AAAA,QACA;AAAA,QACA;AAAA,MACF,CAAC;AAED,aAAO;AAAA,IACT;AAAA,IACA,CAAC,UAAU,KAAK,GAAG,OACjB,SACA,WAKG;AACH,YAAM,UAAU,WAAW,OAAO;AAClC,YAAM,WAAW,MAAM,MAAM,GAAG,OAAO,uBAAuB;AAAA,QAC5D,QAAQ;AAAA,QACR,SAAS;AAAA,UACP,eAAe,UAAU,UAAU,OAAO,CAAC;AAAA,UAC3C,gBAAgB;AAAA,QAClB;AAAA,QACA,MAAM,KAAK,UAAU;AAAA,UACnB,QAAQ,OAAO;AAAA,UACf,GAAG,OAAO,KAAK;AAAA,UACf,MAAM,OAAO,QAAQ;AAAA,QACvB,CAAC;AAAA,MACH,CAAC;AACD,UAAI,CAAC,SAAS,IAAI;AAChB,cAAM,IAAI,MAAM,6BAA6B,SAAS,UAAU,EAAE;AAAA,MACpE;AACA,YAAM,OAAO,MAAM,SAAS,KAAK;AACjC,YAAM,YAAY;AAClB,aAAO,UAAU;AAAA,IACnB;AAAA,IACA,CAAC,UAAU,iBAAiB,GAAG,OAAO,SAAS,WAA4C;AAEzF,UAAI;AACJ,UAAI;AAEJ,UAAI,OAAO,WAAW,UAAU;AAC9B,mBAAW;AACX,iBAAS;AAAA,MACX,OAAO;AAEL,mBAAW,OAAO;AAClB,iBAAS,OAAO;AAAA,MAClB;AAEA,UAAI;AACF,cAAM,UAAU,WAAW,OAAO;AAClC,cAAM,SAAS,UAAU,OAAO;AAEhC,YAAI,CAAC,QAAQ;AACX,iBAAO,MAAM,wBAAwB;AACrC,iBAAO;AAAA,YACL,OAAO;AAAA,YACP,aAAa;AAAA,UACf;AAAA,QACF;AAGA,cAAM,WAAW,MAAM,MAAM,GAAG,OAAO,qBAAqB;AAAA,UAC1D,QAAQ;AAAA,UACR,SAAS;AAAA,YACP,gBAAgB;AAAA,YAChB,eAAe,UAAU,MAAM;AAAA,UACjC;AAAA,UACA,MAAM,KAAK,UAAU;AAAA,YACnB,OAAO;AAAA,YACP,UAAU;AAAA,cACR;AAAA,gBACE,MAAM;AAAA,gBACN,SAAS;AAAA,kBACP;AAAA,oBACE,MAAM;AAAA,oBACN,MACE,UACA;AAAA,kBACJ;AAAA,kBACA;AAAA,oBACE,MAAM;AAAA,oBACN,WAAW,EAAE,KAAK,SAAS;AAAA,kBAC7B;AAAA,gBACF;AAAA,cACF;AAAA,YACF;AAAA,YACA,YAAY;AAAA,UACd,CAAC;AAAA,QACH,CAAC;AAED,YAAI,CAAC,SAAS,IAAI;AAChB,gBAAM,IAAI,MAAM,qBAAqB,SAAS,MAAM,EAAE;AAAA,QACxD;AAEA,cAAM,SAAc,MAAM,SAAS,KAAK;AACxC,cAAM,UAAU,OAAO,UAAU,CAAC,GAAG,SAAS;AAE9C,YAAI,CAAC,SAAS;AACZ,iBAAO;AAAA,YACL,OAAO;AAAA,YACP,aAAa;AAAA,UACf;AAAA,QACF;AAGA,cAAM,aAAa,QAAQ,MAAM,2BAA2B;AAC5D,cAAM,QAAQ,aAAa,CAAC,KAAK;AAGjC,cAAM,cAAc,QAAQ,QAAQ,6BAA6B,EAAE,EAAE,KAAK;AAE1E,eAAO,EAAE,OAAO,YAAY;AAAA,MAC9B,SAAS,OAAO;AACd,eAAO,MAAM,0BAA0B,KAAK;AAC5C,eAAO;AAAA,UACL,OAAO;AAAA,UACP,aAAa,UAAU,iBAAiB,QAAQ,MAAM,UAAU,OAAO,KAAK,CAAC;AAAA,QAC/E;AAAA,MACF;AAAA,IACF;AAAA,IACA,CAAC,UAAU,aAAa,GAAG,OAAO,SAAS,gBAAwB;AACjE,aAAO,IAAI,eAAe,WAAW;AACrC,YAAM,UAAU,WAAW,OAAO;AAElC,YAAM,WAAW,IAAI,SAAS;AAC9B,eAAS,OAAO,QAAQ,aAAa;AAAA,QACnC,UAAU;AAAA,QACV,aAAa;AAAA,MACf,CAAC;AACD,eAAS,OAAO,SAAS,WAAW;AAEpC,YAAM,WAAW,MAAM,MAAM,GAAG,OAAO,yBAAyB;AAAA,QAC9D,QAAQ;AAAA,QACR,SAAS;AAAA,UACP,eAAe,UAAU,UAAU,OAAO,CAAC;AAAA,QAC7C;AAAA,QACA,MAAM;AAAA,MACR,CAAC;AAED,aAAO,IAAI,YAAY,QAAQ;AAC/B,UAAI,CAAC,SAAS,IAAI;AAChB,cAAM,IAAI,MAAM,+BAA+B,SAAS,UAAU,EAAE;AAAA,MACtE;AAEA,YAAM,OAAQ,MAAM,SAAS,KAAK;AAClC,aAAO,KAAK;AAAA,IACd;AAAA,IACA,CAAC,UAAU,cAAc,GAAG,OAAO,SAAuB,SAAiB;AACzE,aAAO,MAAM,kBAAkB,SAAS,IAAI;AAAA,IAC9C;AAAA,IAEA,CAAC,UAAU,YAAY,GAAG,OAAO,SAAS,WAAmC;AAC3E,aAAO,0BAA0B,SAAS,QAAQ,UAAU,cAAc,aAAa;AAAA,IACzF;AAAA,IACA,CAAC,UAAU,YAAY,GAAG,OAAO,SAAS,WAAmC;AAC3E,aAAO,0BAA0B,SAAS,QAAQ,UAAU,cAAc,aAAa;AAAA,IACzF;AAAA,EACF;AAAA,EACA,OAAO;AAAA,IACL;AAAA,MACE,MAAM;AAAA,MACN,OAAO;AAAA,QACL;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,kBAAM,UAAU,WAAW,OAAO;AAClC,kBAAM,WAAW,MAAM,MAAM,GAAG,OAAO,WAAW;AAAA,cAChD,SAAS;AAAA,gBACP,eAAe,UAAU,UAAU,OAAO,CAAC;AAAA,cAC7C;AAAA,YACF,CAAC;AACD,kBAAM,OAAO,MAAM,SAAS,KAAK;AACjC,mBAAO,IAAI,qBAAsB,MAAc,KAAK,MAAM;AAC1D,gBAAI,CAAC,SAAS,IAAI;AAChB,oBAAM,IAAI,MAAM,sCAAsC,SAAS,UAAU,EAAE;AAAA,YAC7E;AAAA,UACF;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,gBAAI;AACF,oBAAM,YAAY,MAAM,QAAQ,SAAS,UAAU,gBAAgB;AAAA,gBACjE,MAAM;AAAA,cACR,CAAC;AACD,qBAAO,IAAI,aAAa,SAAS;AAAA,YACnC,SAAS,OAAO;AACd,qBAAO,MAAM,iCAAiC,KAAK;AACnD,oBAAM;AAAA,YACR;AAAA,UACF;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,gBAAI;AACF,oBAAM,OAAO,MAAM,QAAQ,SAAS,UAAU,YAAY;AAAA,gBACxD,QAAQ;AAAA,cACV,CAAC;AACD,kBAAI,KAAK,WAAW,GAAG;AACrB,sBAAM,IAAI,MAAM,yBAAyB;AAAA,cAC3C;AACA,qBAAO,IAAI,mCAAmC,IAAI;AAAA,YACpD,SAAS,OAAO;AACd,qBAAO,MAAM,6BAA6B,KAAK;AAC/C,oBAAM;AAAA,YACR;AAAA,UACF;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,gBAAI;AACF,oBAAM,OAAO,MAAM,QAAQ,SAAS,UAAU,YAAY;AAAA,gBACxD,QAAQ;AAAA,cACV,CAAC;AACD,kBAAI,KAAK,WAAW,GAAG;AACrB,sBAAM,IAAI,MAAM,yBAAyB;AAAA,cAC3C;AACA,qBAAO,IAAI,mCAAmC,IAAI;AAAA,YACpD,SAAS,OAAO;AACd,qBAAO,MAAM,6BAA6B,KAAK;AAC/C,oBAAM;AAAA,YACR;AAAA,UACF;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,mBAAO,IAAI,8BAA8B;AACzC,gBAAI;AACF,oBAAM,QAAQ,MAAM,QAAQ,SAAS,UAAU,OAAO;AAAA,gBACpD,QAAQ;AAAA,gBACR,GAAG;AAAA,gBACH,MAAM;AAAA,cACR,CAAC;AACD,qBAAO,IAAI,yCAAyC,KAAK;AAAA,YAC3D,SAAS,OAAO;AACd,qBAAO,MAAM,mCAAmC,KAAK;AACrD,oBAAM;AAAA,YACR;AAAA,UACF;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,gBAAI;AACF,qBAAO,IAAI,+BAA+B;AAC1C,kBAAI;AACF,sBAAM,SAAS,MAAM,QAAQ;AAAA,kBAC3B,UAAU;AAAA,kBACV;AAAA,gBACF;AAGA,oBACE,UACA,OAAO,WAAW,YAClB,WAAW,UACX,iBAAiB,QACjB;AACA,yBAAO,IAAI,sBAAsB,MAAM;AAAA,gBACzC,OAAO;AACL,yBAAO,MAAM,4CAA4C,MAAM;AAAA,gBACjE;AAAA,cACF,SAAS,GAAG;AACV,uBAAO,MAAM,oCAAoC,CAAC;AAAA,cACpD;AAAA,YACF,SAAS,GAAG;AACV,qBAAO,MAAM,2CAA2C,CAAC;AAAA,YAC3D;AAAA,UACF;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,mBAAO,IAAI,2BAA2B;AACtC,gBAAI;AACF,oBAAM,WAAW,MAAM;AAAA,gBACrB;AAAA,cACF;AACA,oBAAM,cAAc,MAAM,SAAS,YAAY;AAC/C,oBAAM,gBAAgB,MAAM,QAAQ;AAAA,gBAClC,UAAU;AAAA,gBACV,OAAO,KAAK,IAAI,WAAW,WAAW,CAAC;AAAA,cACzC;AACA,qBAAO,IAAI,sCAAsC,aAAa;AAAA,YAChE,SAAS,OAAO;AACd,qBAAO,MAAM,gCAAgC,KAAK;AAClD,oBAAM;AAAA,YACR;AAAA,UACF;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,kBAAM,SAAS;AACf,kBAAM,SAAS,MAAM,QAAQ,SAAS,UAAU,uBAAuB,EAAE,OAAO,CAAC;AACjF,gBAAI,CAAC,MAAM,QAAQ,MAAM,KAAK,OAAO,WAAW,GAAG;AACjD,oBAAM,IAAI,MAAM,6DAA6D;AAAA,YAC/E;AACA,mBAAO,IAAI,qBAAqB,MAAM;AAAA,UACxC;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAAY;AACrB,kBAAM,SAAS;AAEf,kBAAM,SAAS,MAAM,QAAQ,SAAS,UAAU,uBAAuB,EAAE,OAAO,CAAC;AAEjF,kBAAM,cAAc,MAAM,QAAQ,SAAS,UAAU,uBAAuB,EAAE,OAAO,CAAC;AACtF,gBAAI,gBAAgB,QAAQ;AAC1B,oBAAM,IAAI;AAAA,gBACR,mDAAmD,MAAM,WAAW,WAAW;AAAA,cACjF;AAAA,YACF;AACA,mBAAO,IAAI,iBAAiB,WAAW;AAAA,UACzC;AAAA,QACF;AAAA,QACA;AAAA,UACE,MAAM;AAAA,UACN,IAAI,OAAO,YAA0B;AACnC,gBAAI;AACF,oBAAM,OAAO;AACb,oBAAM,WAAW,MAAM,kBAAkB,SAAS,IAAI;AACtD,kBAAI,CAAC,UAAU;AACb,sBAAM,IAAI,MAAM,2BAA2B;AAAA,cAC7C;AACA,qBAAO,IAAI,+BAA+B;AAAA,YAC5C,SAAS,OAAO;AACd,qBAAO,MAAM,wCAAwC,KAAK;AAC1D,oBAAM;AAAA,YACR;AAAA,UACF;AAAA,QACF;AAAA,MACF;AAAA,IACF;AAAA,EACF;AACF;AACA,IAAO,gBAAQ;","names":[]}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@elizaos/plugin-openai",
-  "version": "1.0.0-beta.3",
+  "version": "1.0.0-beta.33",
   "type": "module",
   "main": "dist/index.js",
   "module": "dist/index.js",
@@ -24,7 +24,7 @@
   "dependencies": {
     "@ai-sdk/openai": "^1.1.9",
     "@ai-sdk/ui-utils": "1.1.9",
-    "@elizaos/core": "^1.0.0-beta.3",
+    "@elizaos/core": "^1.0.0-beta.33",
     "ai": "^4.1.25",
     "js-tiktoken": "^1.0.18",
     "tsup": "8.4.0"
@@ -49,7 +49,7 @@
       }
     }
   },
-  "gitHead": "7fafcbac799e40bd30f1e2acd3239262e79bbb2f",
+  "gitHead": "9da6c2edde9ccc19112c6aa5b028810fb8d2da54",
   "devDependencies": {
     "prettier": "3.5.3"
   }