chore: update CHANGELOG for release v0.7.9-rc1

* 🔍 feat: Introduce Web Search Functionality for OpenAI API

- Added a new web_search parameter to enable web search capabilities in the OpenAI configuration.
- Updated the DynamicSlider component for improved styling.
- Enhanced the useSetIndexOptions hook to auto-enable the Responses API when web search is activated.
- Modified relevant schemas, types, and translation files to support the new web search feature.

* chore: remove comments

* refactor: tool handling in initializeAgent for better clarity and functionality and reflection of openai features

---------

Co-authored-by: Danny Avila <danny@librechat.ai>

.env.example

@@ -58,7 +58,7 @@ DEBUG_CONSOLE=false
 #                     Endpoints                     #
 #===================================================#
 
-# ENDPOINTS=openAI,assistants,azureOpenAI,google,gptPlugins,anthropic
+# ENDPOINTS=openAI,assistants,azureOpenAI,google,anthropic
 
 PROXY=
 
@@ -142,10 +142,10 @@ GOOGLE_KEY=user_provided
 # GOOGLE_AUTH_HEADER=true
 
 # Gemini API (AI Studio)
-# GOOGLE_MODELS=gemini-2.5-pro-preview-05-06,gemini-2.5-flash-preview-04-17,gemini-2.0-flash-001,gemini-2.0-flash-exp,gemini-2.0-flash-lite-001,gemini-1.5-pro-002,gemini-1.5-flash-002
+# GOOGLE_MODELS=gemini-2.5-pro,gemini-2.5-flash,gemini-2.5-flash-lite-preview-06-17,gemini-2.0-flash,gemini-2.0-flash-lite
 
 # Vertex AI
-# GOOGLE_MODELS=gemini-2.5-pro-preview-05-06,gemini-2.5-flash-preview-04-17,gemini-2.0-flash-001,gemini-2.0-flash-exp,gemini-2.0-flash-lite-001,gemini-1.5-pro-002,gemini-1.5-flash-002
+# GOOGLE_MODELS=gemini-2.5-pro,gemini-2.5-flash,gemini-2.5-flash-lite-preview-06-17,gemini-2.0-flash-001,gemini-2.0-flash-lite-001
 
 # GOOGLE_TITLE_MODEL=gemini-2.0-flash-lite-001
 
@@ -453,8 +453,8 @@ OPENID_REUSE_TOKENS=
 OPENID_JWKS_URL_CACHE_ENABLED=
 OPENID_JWKS_URL_CACHE_TIME= # 600000 ms eq to 10 minutes leave empty to disable caching
 #Set to true to trigger token exchange flow to acquire access token for the userinfo endpoint.
-OPENID_ON_BEHALF_FLOW_FOR_USERINFRO_REQUIRED=
-OPENID_ON_BEHALF_FLOW_USERINFRO_SCOPE = "user.read" # example for Scope Needed for Microsoft Graph API
+OPENID_ON_BEHALF_FLOW_FOR_USERINFO_REQUIRED=
+OPENID_ON_BEHALF_FLOW_USERINFO_SCOPE="user.read" # example for Scope Needed for Microsoft Graph API
 # Set to true to use the OpenID Connect end session endpoint for logout
 OPENID_USE_END_SESSION_ENDPOINT=
 
@@ -657,4 +657,4 @@ OPENWEATHER_API_KEY=
 # Reranker (Required)
 # JINA_API_KEY=your_jina_api_key
 # or
-# COHERE_API_KEY=your_cohere_api_key
+# COHERE_API_KEY=your_cohere_api_key

CHANGELOG.md

@@ -7,7 +7,9 @@ All notable changes to this project will be documented in this file.
 
 
 
-## [Unreleased]
+## [v0.7.9-rc1] - 
+
+Changes from v0.7.8 to v0.7.9-rc1.
 
 ### ✨ New Features
 
@@ -35,6 +37,7 @@ All notable changes to this project will be documented in this file.
 - 🌘 fix: artifact of preview text is illegible in dark mode by **@nhtruong** in [#7405](https://github.com/danny-avila/LibreChat/pull/7405)
 - 🛡️ fix: Temporarily Remove CSP until Configurable by **@danny-avila** in [#7419](https://github.com/danny-avila/LibreChat/pull/7419)
 - 💽 fix: Exclude index page `/` from static cache settings by **@sbruel** in [#7382](https://github.com/danny-avila/LibreChat/pull/7382)
+-   📂 fix: Prevent Null Reference Errors in File Process by **@sbruel** in [#8084](https://github.com/danny-avila/LibreChat/pull/8084)
 
 ### ⚙️ Other Changes
 
@@ -49,6 +52,10 @@ All notable changes to this project will be documented in this file.
 
 
 
+[See full release details][release-v0.7.9-rc1]
+
+[release-v0.7.9-rc1]: https://github.com/danny-avila/LibreChat/releases/tag/v0.7.9-rc1
+
 ---
 ## [v0.7.8] - 
 
@@ -91,6 +98,7 @@ Changes from v0.7.8-rc1 to v0.7.8.
 
 ---
 ## [v0.7.8-rc1] - 
+## [v0.7.8-rc1] - 
 
 Changes from v0.7.7 to v0.7.8-rc1.
 

Dockerfile

@@ -1,4 +1,4 @@
-# v0.7.8
+# v0.7.9-rc1
 
 # Base node image
 FROM node:20-alpine AS node

Dockerfile.multi

@@ -1,5 +1,5 @@
 # Dockerfile.multi
-# v0.7.8
+# v0.7.9-rc1
 
 # Base for all builds
 FROM node:20-alpine AS base-min

api/app/clients/BaseClient.js

@@ -13,7 +13,6 @@ const {
 const { getMessages, saveMessage, updateMessage, saveConvo, getConvo } = require('~/models');
 const { checkBalance } = require('~/models/balanceMethods');
 const { truncateToolCallOutputs } = require('./prompts');
-const { addSpaceIfNeeded } = require('~/server/utils');
 const { getFiles } = require('~/models/File');
 const TextStream = require('./TextStream');
 const { logger } = require('~/config');
@@ -572,7 +571,7 @@ class BaseClient {
       });
     }
 
-    const { generation = '' } = opts;
+    const { editedContent } = opts;
 
     // It's not necessary to push to currentMessages
     // depending on subclass implementation of handling messages
@@ -587,11 +586,21 @@ class BaseClient {
           isCreatedByUser: false,
           model: this.modelOptions?.model ?? this.model,
           sender: this.sender,
-          text: generation,
         };
         this.currentMessages.push(userMessage, latestMessage);
-      } else {
-        latestMessage.text = generation;
+      } else if (editedContent != null) {
+        // Handle editedContent for content parts
+        if (editedContent && latestMessage.content && Array.isArray(latestMessage.content)) {
+          const { index, text, type } = editedContent;
+          if (index >= 0 && index < latestMessage.content.length) {
+            const contentPart = latestMessage.content[index];
+            if (type === ContentTypes.THINK && contentPart.type === ContentTypes.THINK) {
+              contentPart[ContentTypes.THINK] = text;
+            } else if (type === ContentTypes.TEXT && contentPart.type === ContentTypes.TEXT) {
+              contentPart[ContentTypes.TEXT] = text;
+            }
+          }
+        }
       }
       this.continued = true;
     } else {
@@ -672,16 +681,32 @@ class BaseClient {
     };
 
     if (typeof completion === 'string') {
-      responseMessage.text = addSpaceIfNeeded(generation) + completion;
+      responseMessage.text = completion;
     } else if (
       Array.isArray(completion) &&
       (this.clientName === EModelEndpoint.agents ||
         isParamEndpoint(this.options.endpoint, this.options.endpointType))
     ) {
       responseMessage.text = '';
-      responseMessage.content = completion;
+
+      if (!opts.editedContent || this.currentMessages.length === 0) {
+        responseMessage.content = completion;
+      } else {
+        const latestMessage = this.currentMessages[this.currentMessages.length - 1];
+        if (!latestMessage?.content) {
+          responseMessage.content = completion;
+        } else {
+          const existingContent = [...latestMessage.content];
+          const { type: editedType } = opts.editedContent;
+          responseMessage.content = this.mergeEditedContent(
+            existingContent,
+            completion,
+            editedType,
+          );
+        }
+      }
     } else if (Array.isArray(completion)) {
-      responseMessage.text = addSpaceIfNeeded(generation) + completion.join('');
+      responseMessage.text = completion.join('');
     }
 
     if (
@@ -792,7 +817,8 @@ class BaseClient {
 
     userMessage.tokenCount = userMessageTokenCount;
     /*
-      Note: `AskController` saves the user message, so we update the count of its `userMessage` reference
+      Note: `AgentController` saves the user message if not saved here
+      (noted by `savedMessageIds`), so we update the count of its `userMessage` reference
     */
     if (typeof opts?.getReqData === 'function') {
       opts.getReqData({
@@ -801,7 +827,8 @@ class BaseClient {
     }
     /*
       Note: we update the user message to be sure it gets the calculated token count;
-      though `AskController` saves the user message, EditController does not
+      though `AgentController` saves the user message if not saved here
+      (noted by `savedMessageIds`), EditController does not
     */
     await userMessagePromise;
     await this.updateMessageInDatabase({
@@ -1093,6 +1120,50 @@ class BaseClient {
     return numTokens;
   }
 
+  /**
+   * Merges completion content with existing content when editing TEXT or THINK types
+   * @param {Array} existingContent - The existing content array
+   * @param {Array} newCompletion - The new completion content
+   * @param {string} editedType - The type of content being edited
+   * @returns {Array} The merged content array
+   */
+  mergeEditedContent(existingContent, newCompletion, editedType) {
+    if (!newCompletion.length) {
+      return existingContent.concat(newCompletion);
+    }
+
+    if (editedType !== ContentTypes.TEXT && editedType !== ContentTypes.THINK) {
+      return existingContent.concat(newCompletion);
+    }
+
+    const lastIndex = existingContent.length - 1;
+    const lastExisting = existingContent[lastIndex];
+    const firstNew = newCompletion[0];
+
+    if (lastExisting?.type !== firstNew?.type || firstNew?.type !== editedType) {
+      return existingContent.concat(newCompletion);
+    }
+
+    const mergedContent = [...existingContent];
+    if (editedType === ContentTypes.TEXT) {
+      mergedContent[lastIndex] = {
+        ...mergedContent[lastIndex],
+        [ContentTypes.TEXT]:
+          (mergedContent[lastIndex][ContentTypes.TEXT] || '') + (firstNew[ContentTypes.TEXT] || ''),
+      };
+    } else {
+      mergedContent[lastIndex] = {
+        ...mergedContent[lastIndex],
+        [ContentTypes.THINK]:
+          (mergedContent[lastIndex][ContentTypes.THINK] || '') +
+          (firstNew[ContentTypes.THINK] || ''),
+      };
+    }
+
+    // Add remaining completion items
+    return mergedContent.concat(newCompletion.slice(1));
+  }
+
   async sendPayload(payload, opts = {}) {
     if (opts && typeof opts === 'object') {
       this.setOptions(opts);

api/app/clients/ChatGPTClient.js

@@ -1,804 +0,0 @@
-const { Keyv } = require('keyv');
-const crypto = require('crypto');
-const { CohereClient } = require('cohere-ai');
-const { fetchEventSource } = require('@waylaidwanderer/fetch-event-source');
-const { constructAzureURL, genAzureChatCompletion } = require('@librechat/api');
-const { encoding_for_model: encodingForModel, get_encoding: getEncoding } = require('tiktoken');
-const {
-  ImageDetail,
-  EModelEndpoint,
-  resolveHeaders,
-  CohereConstants,
-  mapModelToAzureConfig,
-} = require('librechat-data-provider');
-const { createContextHandlers } = require('./prompts');
-const { createCoherePayload } = require('./llm');
-const { extractBaseURL } = require('~/utils');
-const BaseClient = require('./BaseClient');
-const { logger } = require('~/config');
-
-const CHATGPT_MODEL = 'gpt-3.5-turbo';
-const tokenizersCache = {};
-
-class ChatGPTClient extends BaseClient {
-  constructor(apiKey, options = {}, cacheOptions = {}) {
-    super(apiKey, options, cacheOptions);
-
-    cacheOptions.namespace = cacheOptions.namespace || 'chatgpt';
-    this.conversationsCache = new Keyv(cacheOptions);
-    this.setOptions(options);
-  }
-
-  setOptions(options) {
-    if (this.options && !this.options.replaceOptions) {
-      // nested options aren't spread properly, so we need to do this manually
-      this.options.modelOptions = {
-        ...this.options.modelOptions,
-        ...options.modelOptions,
-      };
-      delete options.modelOptions;
-      // now we can merge options
-      this.options = {
-        ...this.options,
-        ...options,
-      };
-    } else {
-      this.options = options;
-    }
-
-    if (this.options.openaiApiKey) {
-      this.apiKey = this.options.openaiApiKey;
-    }
-
-    const modelOptions = this.options.modelOptions || {};
-    this.modelOptions = {
-      ...modelOptions,
-      // set some good defaults (check for undefined in some cases because they may be 0)
-      model: modelOptions.model || CHATGPT_MODEL,
-      temperature: typeof modelOptions.temperature === 'undefined' ? 0.8 : modelOptions.temperature,
-      top_p: typeof modelOptions.top_p === 'undefined' ? 1 : modelOptions.top_p,
-      presence_penalty:
-        typeof modelOptions.presence_penalty === 'undefined' ? 1 : modelOptions.presence_penalty,
-      stop: modelOptions.stop,
-    };
-
-    this.isChatGptModel = this.modelOptions.model.includes('gpt-');
-    const { isChatGptModel } = this;
-    this.isUnofficialChatGptModel =
-      this.modelOptions.model.startsWith('text-chat') ||
-      this.modelOptions.model.startsWith('text-davinci-002-render');
-    const { isUnofficialChatGptModel } = this;
-
-    // Davinci models have a max context length of 4097 tokens.
-    this.maxContextTokens = this.options.maxContextTokens || (isChatGptModel ? 4095 : 4097);
-    // I decided to reserve 1024 tokens for the response.
-    // The max prompt tokens is determined by the max context tokens minus the max response tokens.
-    // Earlier messages will be dropped until the prompt is within the limit.
-    this.maxResponseTokens = this.modelOptions.max_tokens || 1024;
-    this.maxPromptTokens =
-      this.options.maxPromptTokens || this.maxContextTokens - this.maxResponseTokens;
-
-    if (this.maxPromptTokens + this.maxResponseTokens > this.maxContextTokens) {
-      throw new Error(
-        `maxPromptTokens + max_tokens (${this.maxPromptTokens} + ${this.maxResponseTokens} = ${
-          this.maxPromptTokens + this.maxResponseTokens
-        }) must be less than or equal to maxContextTokens (${this.maxContextTokens})`,
-      );
-    }
-
-    this.userLabel = this.options.userLabel || 'User';
-    this.chatGptLabel = this.options.chatGptLabel || 'ChatGPT';
-
-    if (isChatGptModel) {
-      // Use these faux tokens to help the AI understand the context since we are building the chat log ourselves.
-      // Trying to use "<|im_start|>" causes the AI to still generate "<" or "<|" at the end sometimes for some reason,
-      // without tripping the stop sequences, so I'm using "||>" instead.
-      this.startToken = '||>';
-      this.endToken = '';
-      this.gptEncoder = this.constructor.getTokenizer('cl100k_base');
-    } else if (isUnofficialChatGptModel) {
-      this.startToken = '<|im_start|>';
-      this.endToken = '<|im_end|>';
-      this.gptEncoder = this.constructor.getTokenizer('text-davinci-003', true, {
-        '<|im_start|>': 100264,
-        '<|im_end|>': 100265,
-      });
-    } else {
-      // Previously I was trying to use "<|endoftext|>" but there seems to be some bug with OpenAI's token counting
-      // system that causes only the first "<|endoftext|>" to be counted as 1 token, and the rest are not treated
-      // as a single token. So we're using this instead.
-      this.startToken = '||>';
-      this.endToken = '';
-      try {
-        this.gptEncoder = this.constructor.getTokenizer(this.modelOptions.model, true);
-      } catch {
-        this.gptEncoder = this.constructor.getTokenizer('text-davinci-003', true);
-      }
-    }
-
-    if (!this.modelOptions.stop) {
-      const stopTokens = [this.startToken];
-      if (this.endToken && this.endToken !== this.startToken) {
-        stopTokens.push(this.endToken);
-      }
-      stopTokens.push(`\n${this.userLabel}:`);
-      stopTokens.push('<|diff_marker|>');
-      // I chose not to do one for `chatGptLabel` because I've never seen it happen
-      this.modelOptions.stop = stopTokens;
-    }
-
-    if (this.options.reverseProxyUrl) {
-      this.completionsUrl = this.options.reverseProxyUrl;
-    } else if (isChatGptModel) {
-      this.completionsUrl = 'https://api.openai.com/v1/chat/completions';
-    } else {
-      this.completionsUrl = 'https://api.openai.com/v1/completions';
-    }
-
-    return this;
-  }
-
-  static getTokenizer(encoding, isModelName = false, extendSpecialTokens = {}) {
-    if (tokenizersCache[encoding]) {
-      return tokenizersCache[encoding];
-    }
-    let tokenizer;
-    if (isModelName) {
-      tokenizer = encodingForModel(encoding, extendSpecialTokens);
-    } else {
-      tokenizer = getEncoding(encoding, extendSpecialTokens);
-    }
-    tokenizersCache[encoding] = tokenizer;
-    return tokenizer;
-  }
-
-  /** @type {getCompletion} */
-  async getCompletion(input, onProgress, onTokenProgress, abortController = null) {
-    if (!abortController) {
-      abortController = new AbortController();
-    }
-
-    let modelOptions = { ...this.modelOptions };
-    if (typeof onProgress === 'function') {
-      modelOptions.stream = true;
-    }
-    if (this.isChatGptModel) {
-      modelOptions.messages = input;
-    } else {
-      modelOptions.prompt = input;
-    }
-
-    if (this.useOpenRouter && modelOptions.prompt) {
-      delete modelOptions.stop;
-    }
-
-    const { debug } = this.options;
-    let baseURL = this.completionsUrl;
-    if (debug) {
-      console.debug();
-      console.debug(baseURL);
-      console.debug(modelOptions);
-      console.debug();
-    }
-
-    const opts = {
-      method: 'POST',
-      headers: {
-        'Content-Type': 'application/json',
-      },
-    };
-
-    if (this.isVisionModel) {
-      modelOptions.max_tokens = 4000;
-    }
-
-    /** @type {TAzureConfig | undefined} */
-    const azureConfig = this.options?.req?.app?.locals?.[EModelEndpoint.azureOpenAI];
-
-    const isAzure = this.azure || this.options.azure;
-    if (
-      (isAzure && this.isVisionModel && azureConfig) ||
-      (azureConfig && this.isVisionModel && this.options.endpoint === EModelEndpoint.azureOpenAI)
-    ) {
-      const { modelGroupMap, groupMap } = azureConfig;
-      const {
-        azureOptions,
-        baseURL,
-        headers = {},
-        serverless,
-      } = mapModelToAzureConfig({
-        modelName: modelOptions.model,
-        modelGroupMap,
-        groupMap,
-      });
-      opts.headers = resolveHeaders(headers);
-      this.langchainProxy = extractBaseURL(baseURL);
-      this.apiKey = azureOptions.azureOpenAIApiKey;
-
-      const groupName = modelGroupMap[modelOptions.model].group;
-      this.options.addParams = azureConfig.groupMap[groupName].addParams;
-      this.options.dropParams = azureConfig.groupMap[groupName].dropParams;
-      // Note: `forcePrompt` not re-assigned as only chat models are vision models
-
-      this.azure = !serverless && azureOptions;
-      this.azureEndpoint =
-        !serverless && genAzureChatCompletion(this.azure, modelOptions.model, this);
-      if (serverless === true) {
-        this.options.defaultQuery = azureOptions.azureOpenAIApiVersion
-          ? { 'api-version': azureOptions.azureOpenAIApiVersion }
-          : undefined;
-        this.options.headers['api-key'] = this.apiKey;
-      }
-    }
-
-    if (this.options.defaultQuery) {
-      opts.defaultQuery = this.options.defaultQuery;
-    }
-
-    if (this.options.headers) {
-      opts.headers = { ...opts.headers, ...this.options.headers };
-    }
-
-    if (isAzure) {
-      // Azure does not accept `model` in the body, so we need to remove it.
-      delete modelOptions.model;
-
-      baseURL = this.langchainProxy
-        ? constructAzureURL({
-            baseURL: this.langchainProxy,
-            azureOptions: this.azure,
-          })
-        : this.azureEndpoint.split(/(?<!\/)\/(chat|completion)\//)[0];
-
-      if (this.options.forcePrompt) {
-        baseURL += '/completions';
-      } else {
-        baseURL += '/chat/completions';
-      }
-
-      opts.defaultQuery = { 'api-version': this.azure.azureOpenAIApiVersion };
-      opts.headers = { ...opts.headers, 'api-key': this.apiKey };
-    } else if (this.apiKey) {
-      opts.headers.Authorization = `Bearer ${this.apiKey}`;
-    }
-
-    if (process.env.OPENAI_ORGANIZATION) {
-      opts.headers['OpenAI-Organization'] = process.env.OPENAI_ORGANIZATION;
-    }
-
-    if (this.useOpenRouter) {
-      opts.headers['HTTP-Referer'] = 'https://librechat.ai';
-      opts.headers['X-Title'] = 'LibreChat';
-    }
-
-    /* hacky fixes for Mistral AI API:
-      - Re-orders system message to the top of the messages payload, as not allowed anywhere else
-      - If there is only one message and it's a system message, change the role to user
-      */
-    if (baseURL.includes('https://api.mistral.ai/v1') && modelOptions.messages) {
-      const { messages } = modelOptions;
-
-      const systemMessageIndex = messages.findIndex((msg) => msg.role === 'system');
-
-      if (systemMessageIndex > 0) {
-        const [systemMessage] = messages.splice(systemMessageIndex, 1);
-        messages.unshift(systemMessage);
-      }
-
-      modelOptions.messages = messages;
-
-      if (messages.length === 1 && messages[0].role === 'system') {
-        modelOptions.messages[0].role = 'user';
-      }
-    }
-
-    if (this.options.addParams && typeof this.options.addParams === 'object') {
-      modelOptions = {
-        ...modelOptions,
-        ...this.options.addParams,
-      };
-      logger.debug('[ChatGPTClient] chatCompletion: added params', {
-        addParams: this.options.addParams,
-        modelOptions,
-      });
-    }
-
-    if (this.options.dropParams && Array.isArray(this.options.dropParams)) {
-      this.options.dropParams.forEach((param) => {
-        delete modelOptions[param];
-      });
-      logger.debug('[ChatGPTClient] chatCompletion: dropped params', {
-        dropParams: this.options.dropParams,
-        modelOptions,
-      });
-    }
-
-    if (baseURL.startsWith(CohereConstants.API_URL)) {
-      const payload = createCoherePayload({ modelOptions });
-      return await this.cohereChatCompletion({ payload, onTokenProgress });
-    }
-
-    if (baseURL.includes('v1') && !baseURL.includes('/completions') && !this.isChatCompletion) {
-      baseURL = baseURL.split('v1')[0] + 'v1/completions';
-    } else if (
-      baseURL.includes('v1') &&
-      !baseURL.includes('/chat/completions') &&
-      this.isChatCompletion
-    ) {
-      baseURL = baseURL.split('v1')[0] + 'v1/chat/completions';
-    }
-
-    const BASE_URL = new URL(baseURL);
-    if (opts.defaultQuery) {
-      Object.entries(opts.defaultQuery).forEach(([key, value]) => {
-        BASE_URL.searchParams.append(key, value);
-      });
-      delete opts.defaultQuery;
-    }
-
-    const completionsURL = BASE_URL.toString();
-    opts.body = JSON.stringify(modelOptions);
-
-    if (modelOptions.stream) {
-      return new Promise(async (resolve, reject) => {
-        try {
-          let done = false;
-          await fetchEventSource(completionsURL, {
-            ...opts,
-            signal: abortController.signal,
-            async onopen(response) {
-              if (response.status === 200) {
-                return;
-              }
-              if (debug) {
-                console.debug(response);
-              }
-              let error;
-              try {
-                const body = await response.text();
-                error = new Error(`Failed to send message. HTTP ${response.status} - ${body}`);
-                error.status = response.status;
-                error.json = JSON.parse(body);
-              } catch {
-                error = error || new Error(`Failed to send message. HTTP ${response.status}`);
-              }
-              throw error;
-            },
-            onclose() {
-              if (debug) {
-                console.debug('Server closed the connection unexpectedly, returning...');
-              }
-              // workaround for private API not sending [DONE] event
-              if (!done) {
-                onProgress('[DONE]');
-                resolve();
-              }
-            },
-            onerror(err) {
-              if (debug) {
-                console.debug(err);
-              }
-              // rethrow to stop the operation
-              throw err;
-            },
-            onmessage(message) {
-              if (debug) {
-                console.debug(message);
-              }
-              if (!message.data || message.event === 'ping') {
-                return;
-              }
-              if (message.data === '[DONE]') {
-                onProgress('[DONE]');
-                resolve();
-                done = true;
-                return;
-              }
-              onProgress(JSON.parse(message.data));
-            },
-          });
-        } catch (err) {
-          reject(err);
-        }
-      });
-    }
-    const response = await fetch(completionsURL, {
-      ...opts,
-      signal: abortController.signal,
-    });
-    if (response.status !== 200) {
-      const body = await response.text();
-      const error = new Error(`Failed to send message. HTTP ${response.status} - ${body}`);
-      error.status = response.status;
-      try {
-        error.json = JSON.parse(body);
-      } catch {
-        error.body = body;
-      }
-      throw error;
-    }
-    return response.json();
-  }
-
-  /** @type {cohereChatCompletion} */
-  async cohereChatCompletion({ payload, onTokenProgress }) {
-    const cohere = new CohereClient({
-      token: this.apiKey,
-      environment: this.completionsUrl,
-    });
-
-    if (!payload.stream) {
-      const chatResponse = await cohere.chat(payload);
-      return chatResponse.text;
-    }
-
-    const chatStream = await cohere.chatStream(payload);
-    let reply = '';
-    for await (const message of chatStream) {
-      if (!message) {
-        continue;
-      }
-
-      if (message.eventType === 'text-generation' && message.text) {
-        onTokenProgress(message.text);
-        reply += message.text;
-      }
-      /*
-      Cohere API Chinese Unicode character replacement hotfix.
-      Should be un-commented when the following issue is resolved:
-      https://github.com/cohere-ai/cohere-typescript/issues/151
-
-      else if (message.eventType === 'stream-end' && message.response) {
-        reply = message.response.text;
-      }
-      */
-    }
-
-    return reply;
-  }
-
-  async generateTitle(userMessage, botMessage) {
-    const instructionsPayload = {
-      role: 'system',
-      content: `Write an extremely concise subtitle for this conversation with no more than a few words. All words should be capitalized. Exclude punctuation.
-
-||>Message:
-${userMessage.message}
-||>Response:
-${botMessage.message}
-
-||>Title:`,
-    };
-
-    const titleGenClientOptions = JSON.parse(JSON.stringify(this.options));
-    titleGenClientOptions.modelOptions = {
-      model: 'gpt-3.5-turbo',
-      temperature: 0,
-      presence_penalty: 0,
-      frequency_penalty: 0,
-    };
-    const titleGenClient = new ChatGPTClient(this.apiKey, titleGenClientOptions);
-    const result = await titleGenClient.getCompletion([instructionsPayload], null);
-    // remove any non-alphanumeric characters, replace multiple spaces with 1, and then trim
-    return result.choices[0].message.content
-      .replace(/[^a-zA-Z0-9' ]/g, '')
-      .replace(/\s+/g, ' ')
-      .trim();
-  }
-
-  async sendMessage(message, opts = {}) {
-    if (opts.clientOptions && typeof opts.clientOptions === 'object') {
-      this.setOptions(opts.clientOptions);
-    }
-
-    const conversationId = opts.conversationId || crypto.randomUUID();
-    const parentMessageId = opts.parentMessageId || crypto.randomUUID();
-
-    let conversation =
-      typeof opts.conversation === 'object'
-        ? opts.conversation
-        : await this.conversationsCache.get(conversationId);
-
-    let isNewConversation = false;
-    if (!conversation) {
-      conversation = {
-        messages: [],
-        createdAt: Date.now(),
-      };
-      isNewConversation = true;
-    }
-
-    const shouldGenerateTitle = opts.shouldGenerateTitle && isNewConversation;
-
-    const userMessage = {
-      id: crypto.randomUUID(),
-      parentMessageId,
-      role: 'User',
-      message,
-    };
-    conversation.messages.push(userMessage);
-
-    // Doing it this way instead of having each message be a separate element in the array seems to be more reliable,
-    // especially when it comes to keeping the AI in character. It also seems to improve coherency and context retention.
-    const { prompt: payload, context } = await this.buildPrompt(
-      conversation.messages,
-      userMessage.id,
-      {
-        isChatGptModel: this.isChatGptModel,
-        promptPrefix: opts.promptPrefix,
-      },
-    );
-
-    if (this.options.keepNecessaryMessagesOnly) {
-      conversation.messages = context;
-    }
-
-    let reply = '';
-    let result = null;
-    if (typeof opts.onProgress === 'function') {
-      await this.getCompletion(
-        payload,
-        (progressMessage) => {
-          if (progressMessage === '[DONE]') {
-            return;
-          }
-          const token = this.isChatGptModel
-            ? progressMessage.choices[0].delta.content
-            : progressMessage.choices[0].text;
-          // first event's delta content is always undefined
-          if (!token) {
-            return;
-          }
-          if (this.options.debug) {
-            console.debug(token);
-          }
-          if (token === this.endToken) {
-            return;
-          }
-          opts.onProgress(token);
-          reply += token;
-        },
-        opts.abortController || new AbortController(),
-      );
-    } else {
-      result = await this.getCompletion(
-        payload,
-        null,
-        opts.abortController || new AbortController(),
-      );
-      if (this.options.debug) {
-        console.debug(JSON.stringify(result));
-      }
-      if (this.isChatGptModel) {
-        reply = result.choices[0].message.content;
-      } else {
-        reply = result.choices[0].text.replace(this.endToken, '');
-      }
-    }
-
-    // avoids some rendering issues when using the CLI app
-    if (this.options.debug) {
-      console.debug();
-    }
-
-    reply = reply.trim();
-
-    const replyMessage = {
-      id: crypto.randomUUID(),
-      parentMessageId: userMessage.id,
-      role: 'ChatGPT',
-      message: reply,
-    };
-    conversation.messages.push(replyMessage);
-
-    const returnData = {
-      response: replyMessage.message,
-      conversationId,
-      parentMessageId: replyMessage.parentMessageId,
-      messageId: replyMessage.id,
-      details: result || {},
-    };
-
-    if (shouldGenerateTitle) {
-      conversation.title = await this.generateTitle(userMessage, replyMessage);
-      returnData.title = conversation.title;
-    }
-
-    await this.conversationsCache.set(conversationId, conversation);
-
-    if (this.options.returnConversation) {
-      returnData.conversation = conversation;
-    }
-
-    return returnData;
-  }
-
-  async buildPrompt(messages, { isChatGptModel = false, promptPrefix = null }) {
-    promptPrefix = (promptPrefix || this.options.promptPrefix || '').trim();
-
-    // Handle attachments and create augmentedPrompt
-    if (this.options.attachments) {
-      const attachments = await this.options.attachments;
-      const lastMessage = messages[messages.length - 1];
-
-      if (this.message_file_map) {
-        this.message_file_map[lastMessage.messageId] = attachments;
-      } else {
-        this.message_file_map = {
-          [lastMessage.messageId]: attachments,
-        };
-      }
-
-      const files = await this.addImageURLs(lastMessage, attachments);
-      this.options.attachments = files;
-
-      this.contextHandlers = createContextHandlers(this.options.req, lastMessage.text);
-    }
-
-    if (this.message_file_map) {
-      this.contextHandlers = createContextHandlers(
-        this.options.req,
-        messages[messages.length - 1].text,
-      );
-    }
-
-    // Calculate image token cost and process embedded files
-    messages.forEach((message, i) => {
-      if (this.message_file_map && this.message_file_map[message.messageId]) {
-        const attachments = this.message_file_map[message.messageId];
-        for (const file of attachments) {
-          if (file.embedded) {
-            this.contextHandlers?.processFile(file);
-            continue;
-          }
-
-          messages[i].tokenCount =
-            (messages[i].tokenCount || 0) +
-            this.calculateImageTokenCost({
-              width: file.width,
-              height: file.height,
-              detail: this.options.imageDetail ?? ImageDetail.auto,
-            });
-        }
-      }
-    });
-
-    if (this.contextHandlers) {
-      this.augmentedPrompt = await this.contextHandlers.createContext();
-      promptPrefix = this.augmentedPrompt + promptPrefix;
-    }
-
-    if (promptPrefix) {
-      // If the prompt prefix doesn't end with the end token, add it.
-      if (!promptPrefix.endsWith(`${this.endToken}`)) {
-        promptPrefix = `${promptPrefix.trim()}${this.endToken}\n\n`;
-      }
-      promptPrefix = `${this.startToken}Instructions:\n${promptPrefix}`;
-    }
-    const promptSuffix = `${this.startToken}${this.chatGptLabel}:\n`; // Prompt ChatGPT to respond.
-
-    const instructionsPayload = {
-      role: 'system',
-      content: promptPrefix,
-    };
-
-    const messagePayload = {
-      role: 'system',
-      content: promptSuffix,
-    };
-
-    let currentTokenCount;
-    if (isChatGptModel) {
-      currentTokenCount =
-        this.getTokenCountForMessage(instructionsPayload) +
-        this.getTokenCountForMessage(messagePayload);
-    } else {
-      currentTokenCount = this.getTokenCount(`${promptPrefix}${promptSuffix}`);
-    }
-    let promptBody = '';
-    const maxTokenCount = this.maxPromptTokens;
-
-    const context = [];
-
-    // Iterate backwards through the messages, adding them to the prompt until we reach the max token count.
-    // Do this within a recursive async function so that it doesn't block the event loop for too long.
-    const buildPromptBody = async () => {
-      if (currentTokenCount < maxTokenCount && messages.length > 0) {
-        const message = messages.pop();
-        const roleLabel =
-          message?.isCreatedByUser || message?.role?.toLowerCase() === 'user'
-            ? this.userLabel
-            : this.chatGptLabel;
-        const messageString = `${this.startToken}${roleLabel}:\n${
-          message?.text ?? message?.message
-        }${this.endToken}\n`;
-        let newPromptBody;
-        if (promptBody || isChatGptModel) {
-          newPromptBody = `${messageString}${promptBody}`;
-        } else {
-          // Always insert prompt prefix before the last user message, if not gpt-3.5-turbo.
-          // This makes the AI obey the prompt instructions better, which is important for custom instructions.
-          // After a bunch of testing, it doesn't seem to cause the AI any confusion, even if you ask it things
-          // like "what's the last thing I wrote?".
-          newPromptBody = `${promptPrefix}${messageString}${promptBody}`;
-        }
-
-        context.unshift(message);
-
-        const tokenCountForMessage = this.getTokenCount(messageString);
-        const newTokenCount = currentTokenCount + tokenCountForMessage;
-        if (newTokenCount > maxTokenCount) {
-          if (promptBody) {
-            // This message would put us over the token limit, so don't add it.
-            return false;
-          }
-          // This is the first message, so we can't add it. Just throw an error.
-          throw new Error(
-            `Prompt is too long. Max token count is ${maxTokenCount}, but prompt is ${newTokenCount} tokens long.`,
-          );
-        }
-        promptBody = newPromptBody;
-        currentTokenCount = newTokenCount;
-        // wait for next tick to avoid blocking the event loop
-        await new Promise((resolve) => setImmediate(resolve));
-        return buildPromptBody();
-      }
-      return true;
-    };
-
-    await buildPromptBody();
-
-    const prompt = `${promptBody}${promptSuffix}`;
-    if (isChatGptModel) {
-      messagePayload.content = prompt;
-      // Add 3 tokens for Assistant Label priming after all messages have been counted.
-      currentTokenCount += 3;
-    }
-
-    // Use up to `this.maxContextTokens` tokens (prompt + response), but try to leave `this.maxTokens` tokens for the response.
-    this.modelOptions.max_tokens = Math.min(
-      this.maxContextTokens - currentTokenCount,
-      this.maxResponseTokens,
-    );
-
-    if (isChatGptModel) {
-      return { prompt: [instructionsPayload, messagePayload], context };
-    }
-    return { prompt, context, promptTokens: currentTokenCount };
-  }
-
-  getTokenCount(text) {
-    return this.gptEncoder.encode(text, 'all').length;
-  }
-
-  /**
-   * Algorithm adapted from "6. Counting tokens for chat API calls" of
-   * https://github.com/openai/openai-cookbook/blob/main/examples/How_to_count_tokens_with_tiktoken.ipynb
-   *
-   * An additional 3 tokens need to be added for assistant label priming after all messages have been counted.
-   *
-   * @param {Object} message
-   */
-  getTokenCountForMessage(message) {
-    // Note: gpt-3.5-turbo and gpt-4 may update over time. Use default for these as well as for unknown models
-    let tokensPerMessage = 3;
-    let tokensPerName = 1;
-
-    if (this.modelOptions.model === 'gpt-3.5-turbo-0301') {
-      tokensPerMessage = 4;
-      tokensPerName = -1;
-    }
-
-    let numTokens = tokensPerMessage;
-    for (let [key, value] of Object.entries(message)) {
-      numTokens += this.getTokenCount(value);
-      if (key === 'name') {
-        numTokens += tokensPerName;
-      }
-    }
-
-    return numTokens;
-  }
-}
-
-module.exports = ChatGPTClient;

api/app/clients/GoogleClient.js

@@ -1,7 +1,7 @@
 const { google } = require('googleapis');
-const { Tokenizer } = require('@librechat/api');
 const { concat } = require('@langchain/core/utils/stream');
 const { ChatVertexAI } = require('@langchain/google-vertexai');
+const { Tokenizer, getSafetySettings } = require('@librechat/api');
 const { ChatGoogleGenerativeAI } = require('@langchain/google-genai');
 const { GoogleGenerativeAI: GenAI } = require('@google/generative-ai');
 const { HumanMessage, SystemMessage } = require('@langchain/core/messages');
@@ -12,13 +12,13 @@ const {
   endpointSettings,
   parseTextParts,
   EModelEndpoint,
+  googleSettings,
   ContentTypes,
   VisionModes,
   ErrorTypes,
   Constants,
   AuthKeys,
 } = require('librechat-data-provider');
-const { getSafetySettings } = require('~/server/services/Endpoints/google/llm');
 const { encodeAndFormat } = require('~/server/services/Files/images');
 const { spendTokens } = require('~/models/spendTokens');
 const { getModelMaxTokens } = require('~/utils');
@@ -166,6 +166,16 @@ class GoogleClient extends BaseClient {
       );
     }
 
+    // Add thinking configuration
+    this.modelOptions.thinkingConfig = {
+      thinkingBudget:
+        (this.modelOptions.thinking ?? googleSettings.thinking.default)
+          ? this.modelOptions.thinkingBudget
+          : 0,
+    };
+    delete this.modelOptions.thinking;
+    delete this.modelOptions.thinkingBudget;
+
     this.sender =
       this.options.sender ??
       getResponseSender({

api/app/clients/OpenAIClient.js

@@ -5,6 +5,7 @@ const {
   isEnabled,
   Tokenizer,
   createFetch,
+  resolveHeaders,
   constructAzureURL,
   genAzureChatCompletion,
   createStreamEventHandlers,
@@ -15,7 +16,6 @@ const {
   ContentTypes,
   parseTextParts,
   EModelEndpoint,
-  resolveHeaders,
   KnownEndpoints,
   openAISettings,
   ImageDetailCost,
@@ -37,7 +37,6 @@ const { addSpaceIfNeeded, sleep } = require('~/server/utils');
 const { spendTokens } = require('~/models/spendTokens');
 const { handleOpenAIErrors } = require('./tools/util');
 const { createLLM, RunManager } = require('./llm');
-const ChatGPTClient = require('./ChatGPTClient');
 const { summaryBuffer } = require('./memory');
 const { runTitleChain } = require('./chains');
 const { tokenSplit } = require('./document');
@@ -47,12 +46,6 @@ const { logger } = require('~/config');
 class OpenAIClient extends BaseClient {
   constructor(apiKey, options = {}) {
     super(apiKey, options);
-    this.ChatGPTClient = new ChatGPTClient();
-    this.buildPrompt = this.ChatGPTClient.buildPrompt.bind(this);
-    /** @type {getCompletion} */
-    this.getCompletion = this.ChatGPTClient.getCompletion.bind(this);
-    /** @type {cohereChatCompletion} */
-    this.cohereChatCompletion = this.ChatGPTClient.cohereChatCompletion.bind(this);
     this.contextStrategy = options.contextStrategy
       ? options.contextStrategy.toLowerCase()
       : 'discard';
@@ -379,23 +372,12 @@ class OpenAIClient extends BaseClient {
     return files;
   }
 
-  async buildMessages(
-    messages,
-    parentMessageId,
-    { isChatCompletion = false, promptPrefix = null },
-    opts,
-  ) {
+  async buildMessages(messages, parentMessageId, { promptPrefix = null }, opts) {
     let orderedMessages = this.constructor.getMessagesForConversation({
       messages,
       parentMessageId,
       summary: this.shouldSummarize,
     });
-    if (!isChatCompletion) {
-      return await this.buildPrompt(orderedMessages, {
-        isChatGptModel: isChatCompletion,
-        promptPrefix,
-      });
-    }
 
     let payload;
     let instructions;

api/app/clients/PluginsClient.js

@@ -1,542 +0,0 @@
-const OpenAIClient = require('./OpenAIClient');
-const { CallbackManager } = require('@langchain/core/callbacks/manager');
-const { BufferMemory, ChatMessageHistory } = require('langchain/memory');
-const { addImages, buildErrorInput, buildPromptPrefix } = require('./output_parsers');
-const { initializeCustomAgent, initializeFunctionsAgent } = require('./agents');
-const { processFileURL } = require('~/server/services/Files/process');
-const { EModelEndpoint } = require('librechat-data-provider');
-const { checkBalance } = require('~/models/balanceMethods');
-const { formatLangChainMessages } = require('./prompts');
-const { extractBaseURL } = require('~/utils');
-const { loadTools } = require('./tools/util');
-const { logger } = require('~/config');
-
-class PluginsClient extends OpenAIClient {
-  constructor(apiKey, options = {}) {
-    super(apiKey, options);
-    this.sender = options.sender ?? 'Assistant';
-    this.tools = [];
-    this.actions = [];
-    this.setOptions(options);
-    this.openAIApiKey = this.apiKey;
-    this.executor = null;
-  }
-
-  setOptions(options) {
-    this.agentOptions = { ...options.agentOptions };
-    this.functionsAgent = this.agentOptions?.agent === 'functions';
-    this.agentIsGpt3 = this.agentOptions?.model?.includes('gpt-3');
-
-    super.setOptions(options);
-
-    this.isGpt3 = this.modelOptions?.model?.includes('gpt-3');
-
-    if (this.options.reverseProxyUrl) {
-      this.langchainProxy = extractBaseURL(this.options.reverseProxyUrl);
-    }
-  }
-
-  getSaveOptions() {
-    return {
-      artifacts: this.options.artifacts,
-      chatGptLabel: this.options.chatGptLabel,
-      modelLabel: this.options.modelLabel,
-      promptPrefix: this.options.promptPrefix,
-      tools: this.options.tools,
-      ...this.modelOptions,
-      agentOptions: this.agentOptions,
-      iconURL: this.options.iconURL,
-      greeting: this.options.greeting,
-      spec: this.options.spec,
-    };
-  }
-
-  saveLatestAction(action) {
-    this.actions.push(action);
-  }
-
-  getFunctionModelName(input) {
-    if (/-(?!0314)\d{4}/.test(input)) {
-      return input;
-    } else if (input.includes('gpt-3.5-turbo')) {
-      return 'gpt-3.5-turbo';
-    } else if (input.includes('gpt-4')) {
-      return 'gpt-4';
-    } else {
-      return 'gpt-3.5-turbo';
-    }
-  }
-
-  getBuildMessagesOptions(opts) {
-    return {
-      isChatCompletion: true,
-      promptPrefix: opts.promptPrefix,
-      abortController: opts.abortController,
-    };
-  }
-
-  async initialize({ user, message, onAgentAction, onChainEnd, signal }) {
-    const modelOptions = {
-      modelName: this.agentOptions.model,
-      temperature: this.agentOptions.temperature,
-    };
-
-    const model = this.initializeLLM({
-      ...modelOptions,
-      context: 'plugins',
-      initialMessageCount: this.currentMessages.length + 1,
-    });
-
-    logger.debug(
-      `[PluginsClient] Agent Model: ${model.modelName} | Temp: ${model.temperature} | Functions: ${this.functionsAgent}`,
-    );
-
-    // Map Messages to Langchain format
-    const pastMessages = formatLangChainMessages(this.currentMessages.slice(0, -1), {
-      userName: this.options?.name,
-    });
-    logger.debug('[PluginsClient] pastMessages: ' + pastMessages.length);
-
-    // TODO: use readOnly memory, TokenBufferMemory? (both unavailable in LangChainJS)
-    const memory = new BufferMemory({
-      llm: model,
-      chatHistory: new ChatMessageHistory(pastMessages),
-    });
-
-    const { loadedTools } = await loadTools({
-      user,
-      model,
-      tools: this.options.tools,
-      functions: this.functionsAgent,
-      options: {
-        memory,
-        signal: this.abortController.signal,
-        openAIApiKey: this.openAIApiKey,
-        conversationId: this.conversationId,
-        fileStrategy: this.options.req.app.locals.fileStrategy,
-        processFileURL,
-        message,
-      },
-      useSpecs: true,
-    });
-
-    if (loadedTools.length === 0) {
-      return;
-    }
-
-    this.tools = loadedTools;
-
-    logger.debug('[PluginsClient] Requested Tools', this.options.tools);
-    logger.debug(
-      '[PluginsClient] Loaded Tools',
-      this.tools.map((tool) => tool.name),
-    );
-
-    const handleAction = (action, runId, callback = null) => {
-      this.saveLatestAction(action);
-
-      logger.debug('[PluginsClient] Latest Agent Action ', this.actions[this.actions.length - 1]);
-
-      if (typeof callback === 'function') {
-        callback(action, runId);
-      }
-    };
-
-    // initialize agent
-    const initializer = this.functionsAgent ? initializeFunctionsAgent : initializeCustomAgent;
-
-    let customInstructions = (this.options.promptPrefix ?? '').trim();
-    if (typeof this.options.artifactsPrompt === 'string' && this.options.artifactsPrompt) {
-      customInstructions = `${customInstructions ?? ''}\n${this.options.artifactsPrompt}`.trim();
-    }
-
-    this.executor = await initializer({
-      model,
-      signal,
-      pastMessages,
-      tools: this.tools,
-      customInstructions,
-      verbose: this.options.debug,
-      returnIntermediateSteps: true,
-      customName: this.options.chatGptLabel,
-      currentDateString: this.currentDateString,
-      callbackManager: CallbackManager.fromHandlers({
-        async handleAgentAction(action, runId) {
-          handleAction(action, runId, onAgentAction);
-        },
-        async handleChainEnd(action) {
-          if (typeof onChainEnd === 'function') {
-            onChainEnd(action);
-          }
-        },
-      }),
-    });
-
-    logger.debug('[PluginsClient] Loaded agent.');
-  }
-
-  async executorCall(message, { signal, stream, onToolStart, onToolEnd }) {
-    let errorMessage = '';
-    const maxAttempts = 1;
-
-    for (let attempts = 1; attempts <= maxAttempts; attempts++) {
-      const errorInput = buildErrorInput({
-        message,
-        errorMessage,
-        actions: this.actions,
-        functionsAgent: this.functionsAgent,
-      });
-      const input = attempts > 1 ? errorInput : message;
-
-      logger.debug(`[PluginsClient] Attempt ${attempts} of ${maxAttempts}`);
-
-      if (errorMessage.length > 0) {
-        logger.debug('[PluginsClient] Caught error, input: ' + JSON.stringify(input));
-      }
-
-      try {
-        this.result = await this.executor.call({ input, signal }, [
-          {
-            async handleToolStart(...args) {
-              await onToolStart(...args);
-            },
-            async handleToolEnd(...args) {
-              await onToolEnd(...args);
-            },
-            async handleLLMEnd(output) {
-              const { generations } = output;
-              const { text } = generations[0][0];
-              if (text && typeof stream === 'function') {
-                await stream(text);
-              }
-            },
-          },
-        ]);
-        break; // Exit the loop if the function call is successful
-      } catch (err) {
-        logger.error('[PluginsClient] executorCall error:', err);
-        if (attempts === maxAttempts) {
-          const { run } = this.runManager.getRunByConversationId(this.conversationId);
-          const defaultOutput = `Encountered an error while attempting to respond: ${err.message}`;
-          this.result.output = run && run.error ? run.error : defaultOutput;
-          this.result.errorMessage = run && run.error ? run.error : err.message;
-          this.result.intermediateSteps = this.actions;
-          break;
-        }
-      }
-    }
-  }
-
-  /**
-   *
-   * @param {TMessage} responseMessage
-   * @param {Partial<TMessage>} saveOptions
-   * @param {string} user
-   * @returns
-   */
-  async handleResponseMessage(responseMessage, saveOptions, user) {
-    const { output, errorMessage, ...result } = this.result;
-    logger.debug('[PluginsClient][handleResponseMessage] Output:', {
-      output,
-      errorMessage,
-      ...result,
-    });
-    const { error } = responseMessage;
-    if (!error) {
-      responseMessage.tokenCount = this.getTokenCountForResponse(responseMessage);
-      responseMessage.completionTokens = this.getTokenCount(responseMessage.text);
-    }
-
-    // Record usage only when completion is skipped as it is already recorded in the agent phase.
-    if (!this.agentOptions.skipCompletion && !error) {
-      await this.recordTokenUsage(responseMessage);
-    }
-
-    const databasePromise = this.saveMessageToDatabase(responseMessage, saveOptions, user);
-    delete responseMessage.tokenCount;
-    return { ...responseMessage, ...result, databasePromise };
-  }
-
-  async sendMessage(message, opts = {}) {
-    /** @type {Promise<TMessage>} */
-    let userMessagePromise;
-    /** @type {{ filteredTools: string[], includedTools: string[] }} */
-    const { filteredTools = [], includedTools = [] } = this.options.req.app.locals;
-
-    if (includedTools.length > 0) {
-      const tools = this.options.tools.filter((plugin) => includedTools.includes(plugin));
-      this.options.tools = tools;
-    } else {
-      const tools = this.options.tools.filter((plugin) => !filteredTools.includes(plugin));
-      this.options.tools = tools;
-    }
-
-    // If a message is edited, no tools can be used.
-    const completionMode = this.options.tools.length === 0 || opts.isEdited;
-    if (completionMode) {
-      this.setOptions(opts);
-      return super.sendMessage(message, opts);
-    }
-
-    logger.debug('[PluginsClient] sendMessage', { userMessageText: message, opts });
-    const {
-      user,
-      conversationId,
-      responseMessageId,
-      saveOptions,
-      userMessage,
-      onAgentAction,
-      onChainEnd,
-      onToolStart,
-      onToolEnd,
-    } = await this.handleStartMethods(message, opts);
-
-    if (opts.progressCallback) {
-      opts.onProgress = opts.progressCallback.call(null, {
-        ...(opts.progressOptions ?? {}),
-        parentMessageId: userMessage.messageId,
-        messageId: responseMessageId,
-      });
-    }
-
-    this.currentMessages.push(userMessage);
-
-    let {
-      prompt: payload,
-      tokenCountMap,
-      promptTokens,
-    } = await this.buildMessages(
-      this.currentMessages,
-      userMessage.messageId,
-      this.getBuildMessagesOptions({
-        promptPrefix: null,
-        abortController: this.abortController,
-      }),
-    );
-
-    if (tokenCountMap) {
-      logger.debug('[PluginsClient] tokenCountMap', { tokenCountMap });
-      if (tokenCountMap[userMessage.messageId]) {
-        userMessage.tokenCount = tokenCountMap[userMessage.messageId];
-        logger.debug('[PluginsClient] userMessage.tokenCount', userMessage.tokenCount);
-      }
-      this.handleTokenCountMap(tokenCountMap);
-    }
-
-    this.result = {};
-    if (payload) {
-      this.currentMessages = payload;
-    }
-
-    if (!this.skipSaveUserMessage) {
-      userMessagePromise = this.saveMessageToDatabase(userMessage, saveOptions, user);
-      if (typeof opts?.getReqData === 'function') {
-        opts.getReqData({
-          userMessagePromise,
-        });
-      }
-    }
-
-    const balance = this.options.req?.app?.locals?.balance;
-    if (balance?.enabled) {
-      await checkBalance({
-        req: this.options.req,
-        res: this.options.res,
-        txData: {
-          user: this.user,
-          tokenType: 'prompt',
-          amount: promptTokens,
-          debug: this.options.debug,
-          model: this.modelOptions.model,
-          endpoint: EModelEndpoint.openAI,
-        },
-      });
-    }
-
-    const responseMessage = {
-      endpoint: EModelEndpoint.gptPlugins,
-      iconURL: this.options.iconURL,
-      messageId: responseMessageId,
-      conversationId,
-      parentMessageId: userMessage.messageId,
-      isCreatedByUser: false,
-      model: this.modelOptions.model,
-      sender: this.sender,
-      promptTokens,
-    };
-
-    await this.initialize({
-      user,
-      message,
-      onAgentAction,
-      onChainEnd,
-      signal: this.abortController.signal,
-      onProgress: opts.onProgress,
-    });
-
-    // const stream = async (text) => {
-    //   await this.generateTextStream.call(this, text, opts.onProgress, { delay: 1 });
-    // };
-    await this.executorCall(message, {
-      signal: this.abortController.signal,
-      // stream,
-      onToolStart,
-      onToolEnd,
-    });
-
-    // If message was aborted mid-generation
-    if (this.result?.errorMessage?.length > 0 && this.result?.errorMessage?.includes('cancel')) {
-      responseMessage.text = 'Cancelled.';
-      return await this.handleResponseMessage(responseMessage, saveOptions, user);
-    }
-
-    // If error occurred during generation (likely token_balance)
-    if (this.result?.errorMessage?.length > 0) {
-      responseMessage.error = true;
-      responseMessage.text = this.result.output;
-      return await this.handleResponseMessage(responseMessage, saveOptions, user);
-    }
-
-    if (this.agentOptions.skipCompletion && this.result.output && this.functionsAgent) {
-      const partialText = opts.getPartialText();
-      const trimmedPartial = opts.getPartialText().replaceAll(':::plugin:::\n', '');
-      responseMessage.text =
-        trimmedPartial.length === 0 ? `${partialText}${this.result.output}` : partialText;
-      addImages(this.result.intermediateSteps, responseMessage);
-      await this.generateTextStream(this.result.output, opts.onProgress, { delay: 5 });
-      return await this.handleResponseMessage(responseMessage, saveOptions, user);
-    }
-
-    if (this.agentOptions.skipCompletion && this.result.output) {
-      responseMessage.text = this.result.output;
-      addImages(this.result.intermediateSteps, responseMessage);
-      await this.generateTextStream(this.result.output, opts.onProgress, { delay: 5 });
-      return await this.handleResponseMessage(responseMessage, saveOptions, user);
-    }
-
-    logger.debug('[PluginsClient] Completion phase: this.result', this.result);
-
-    const promptPrefix = buildPromptPrefix({
-      result: this.result,
-      message,
-      functionsAgent: this.functionsAgent,
-    });
-
-    logger.debug('[PluginsClient]', { promptPrefix });
-
-    payload = await this.buildCompletionPrompt({
-      messages: this.currentMessages,
-      promptPrefix,
-    });
-
-    logger.debug('[PluginsClient] buildCompletionPrompt Payload', payload);
-    responseMessage.text = await this.sendCompletion(payload, opts);
-    return await this.handleResponseMessage(responseMessage, saveOptions, user);
-  }
-
-  async buildCompletionPrompt({ messages, promptPrefix: _promptPrefix }) {
-    logger.debug('[PluginsClient] buildCompletionPrompt messages', messages);
-
-    const orderedMessages = messages;
-    let promptPrefix = _promptPrefix.trim();
-    // If the prompt prefix doesn't end with the end token, add it.
-    if (!promptPrefix.endsWith(`${this.endToken}`)) {
-      promptPrefix = `${promptPrefix.trim()}${this.endToken}\n\n`;
-    }
-    promptPrefix = `${this.startToken}Instructions:\n${promptPrefix}`;
-    const promptSuffix = `${this.startToken}${this.chatGptLabel ?? 'Assistant'}:\n`;
-
-    const instructionsPayload = {
-      role: 'system',
-      content: promptPrefix,
-    };
-
-    const messagePayload = {
-      role: 'system',
-      content: promptSuffix,
-    };
-
-    if (this.isGpt3) {
-      instructionsPayload.role = 'user';
-      messagePayload.role = 'user';
-      instructionsPayload.content += `\n${promptSuffix}`;
-    }
-
-    // testing if this works with browser endpoint
-    if (!this.isGpt3 && this.options.reverseProxyUrl) {
-      instructionsPayload.role = 'user';
-    }
-
-    let currentTokenCount =
-      this.getTokenCountForMessage(instructionsPayload) +
-      this.getTokenCountForMessage(messagePayload);
-
-    let promptBody = '';
-    const maxTokenCount = this.maxPromptTokens;
-    // Iterate backwards through the messages, adding them to the prompt until we reach the max token count.
-    // Do this within a recursive async function so that it doesn't block the event loop for too long.
-    const buildPromptBody = async () => {
-      if (currentTokenCount < maxTokenCount && orderedMessages.length > 0) {
-        const message = orderedMessages.pop();
-        const isCreatedByUser = message.isCreatedByUser || message.role?.toLowerCase() === 'user';
-        const roleLabel = isCreatedByUser ? this.userLabel : this.chatGptLabel;
-        let messageString = `${this.startToken}${roleLabel}:\n${
-          message.text ?? message.content ?? ''
-        }${this.endToken}\n`;
-        let newPromptBody = `${messageString}${promptBody}`;
-
-        const tokenCountForMessage = this.getTokenCount(messageString);
-        const newTokenCount = currentTokenCount + tokenCountForMessage;
-        if (newTokenCount > maxTokenCount) {
-          if (promptBody) {
-            // This message would put us over the token limit, so don't add it.
-            return false;
-          }
-          // This is the first message, so we can't add it. Just throw an error.
-          throw new Error(
-            `Prompt is too long. Max token count is ${maxTokenCount}, but prompt is ${newTokenCount} tokens long.`,
-          );
-        }
-        promptBody = newPromptBody;
-        currentTokenCount = newTokenCount;
-        // wait for next tick to avoid blocking the event loop
-        await new Promise((resolve) => setTimeout(resolve, 0));
-        return buildPromptBody();
-      }
-      return true;
-    };
-
-    await buildPromptBody();
-    const prompt = promptBody;
-    messagePayload.content = prompt;
-    // Add 2 tokens for metadata after all messages have been counted.
-    currentTokenCount += 2;
-
-    if (this.isGpt3 && messagePayload.content.length > 0) {
-      const context = 'Chat History:\n';
-      messagePayload.content = `${context}${prompt}`;
-      currentTokenCount += this.getTokenCount(context);
-    }
-
-    // Use up to `this.maxContextTokens` tokens (prompt + response), but try to leave `this.maxTokens` tokens for the response.
-    this.modelOptions.max_tokens = Math.min(
-      this.maxContextTokens - currentTokenCount,
-      this.maxResponseTokens,
-    );
-
-    if (this.isGpt3) {
-      messagePayload.content += promptSuffix;
-      return [instructionsPayload, messagePayload];
-    }
-
-    const result = [messagePayload, instructionsPayload];
-
-    if (this.functionsAgent && !this.isGpt3) {
-      result[1].content = `${result[1].content}\n${this.startToken}${this.chatGptLabel}:\nSure thing! Here is the output you requested:\n`;
-    }
-
-    return result.filter((message) => message.content.length > 0);
-  }
-}
-
-module.exports = PluginsClient;

api/app/clients/index.js

@@ -1,15 +1,11 @@
-const ChatGPTClient = require('./ChatGPTClient');
 const OpenAIClient = require('./OpenAIClient');
-const PluginsClient = require('./PluginsClient');
 const GoogleClient = require('./GoogleClient');
 const TextStream = require('./TextStream');
 const AnthropicClient = require('./AnthropicClient');
 const toolUtils = require('./tools/util');
 
 module.exports = {
-  ChatGPTClient,
   OpenAIClient,
-  PluginsClient,
   GoogleClient,
   TextStream,
   AnthropicClient,

api/app/clients/prompts/createContextHandlers.js

@@ -1,6 +1,7 @@
 const axios = require('axios');
-const { isEnabled } = require('~/server/utils');
-const { logger } = require('~/config');
+const { isEnabled } = require('@librechat/api');
+const { logger } = require('@librechat/data-schemas');
+const { generateShortLivedToken } = require('~/server/services/AuthService');
 
 const footer = `Use the context as your learned knowledge to better answer the user.
 
@@ -18,7 +19,7 @@ function createContextHandlers(req, userMessageContent) {
   const queryPromises = [];
   const processedFiles = [];
   const processedIds = new Set();
-  const jwtToken = req.headers.authorization.split(' ')[1];
+  const jwtToken = generateShortLivedToken(req.user.id);
   const useFullContext = isEnabled(process.env.RAG_USE_FULL_CONTEXT);
 
   const query = async (file) => {
@@ -96,35 +97,35 @@ function createContextHandlers(req, userMessageContent) {
         resolvedQueries.length === 0
           ? '\n\tThe semantic search did not return any results.'
           : resolvedQueries
-            .map((queryResult, index) => {
-              const file = processedFiles[index];
-              let contextItems = queryResult.data;
+              .map((queryResult, index) => {
+                const file = processedFiles[index];
+                let contextItems = queryResult.data;
 
-              const generateContext = (currentContext) =>
-                `
+                const generateContext = (currentContext) =>
+                  `
           <file>
             <filename>${file.filename}</filename>
             <context>${currentContext}
             </context>
           </file>`;
 
-              if (useFullContext) {
-                return generateContext(`\n${contextItems}`);
-              }
+                if (useFullContext) {
+                  return generateContext(`\n${contextItems}`);
+                }
 
-              contextItems = queryResult.data
-                .map((item) => {
-                  const pageContent = item[0].page_content;
-                  return `
+                contextItems = queryResult.data
+                  .map((item) => {
+                    const pageContent = item[0].page_content;
+                    return `
             <contextItem>
               <![CDATA[${pageContent?.trim()}]]>
             </contextItem>`;
-                })
-                .join('');
+                  })
+                  .join('');
 
-              return generateContext(contextItems);
-            })
-            .join('');
+                return generateContext(contextItems);
+              })
+              .join('');
 
       if (useFullContext) {
         const prompt = `${header}

api/app/clients/specs/OpenAIClient.test.js

@@ -531,44 +531,6 @@ describe('OpenAIClient', () => {
     });
   });
 
-  describe('sendMessage/getCompletion/chatCompletion', () => {
-    afterEach(() => {
-      delete process.env.AZURE_OPENAI_DEFAULT_MODEL;
-      delete process.env.AZURE_USE_MODEL_AS_DEPLOYMENT_NAME;
-    });
-
-    it('should call getCompletion and fetchEventSource when using a text/instruct model', async () => {
-      const model = 'text-davinci-003';
-      const onProgress = jest.fn().mockImplementation(() => ({}));
-
-      const testClient = new OpenAIClient('test-api-key', {
-        ...defaultOptions,
-        modelOptions: { model },
-      });
-
-      const getCompletion = jest.spyOn(testClient, 'getCompletion');
-      await testClient.sendMessage('Hi mom!', { onProgress });
-
-      expect(getCompletion).toHaveBeenCalled();
-      expect(getCompletion.mock.calls.length).toBe(1);
-
-      expect(getCompletion.mock.calls[0][0]).toBe('||>User:\nHi mom!\n||>Assistant:\n');
-
-      expect(fetchEventSource).toHaveBeenCalled();
-      expect(fetchEventSource.mock.calls.length).toBe(1);
-
-      // Check if the first argument (url) is correct
-      const firstCallArgs = fetchEventSource.mock.calls[0];
-
-      const expectedURL = 'https://api.openai.com/v1/completions';
-      expect(firstCallArgs[0]).toBe(expectedURL);
-
-      const requestBody = JSON.parse(firstCallArgs[1].body);
-      expect(requestBody).toHaveProperty('model');
-      expect(requestBody.model).toBe(model);
-    });
-  });
-
   describe('checkVisionRequest functionality', () => {
     let client;
     const attachments = [{ type: 'image/png' }];

api/app/clients/specs/PluginsClient.test.js

@@ -1,314 +0,0 @@
-const crypto = require('crypto');
-const { Constants } = require('librechat-data-provider');
-const { HumanMessage, AIMessage } = require('@langchain/core/messages');
-const PluginsClient = require('../PluginsClient');
-
-jest.mock('~/db/connect');
-jest.mock('~/models/Conversation', () => {
-  return function () {
-    return {
-      save: jest.fn(),
-      deleteConvos: jest.fn(),
-    };
-  };
-});
-
-const defaultAzureOptions = {
-  azureOpenAIApiInstanceName: 'your-instance-name',
-  azureOpenAIApiDeploymentName: 'your-deployment-name',
-  azureOpenAIApiVersion: '2020-07-01-preview',
-};
-
-describe('PluginsClient', () => {
-  let TestAgent;
-  let options = {
-    tools: [],
-    modelOptions: {
-      model: 'gpt-3.5-turbo',
-      temperature: 0,
-      max_tokens: 2,
-    },
-    agentOptions: {
-      model: 'gpt-3.5-turbo',
-    },
-  };
-  let parentMessageId;
-  let conversationId;
-  const fakeMessages = [];
-  const userMessage = 'Hello, ChatGPT!';
-  const apiKey = 'fake-api-key';
-
-  beforeEach(() => {
-    TestAgent = new PluginsClient(apiKey, options);
-    TestAgent.loadHistory = jest
-      .fn()
-      .mockImplementation((conversationId, parentMessageId = null) => {
-        if (!conversationId) {
-          TestAgent.currentMessages = [];
-          return Promise.resolve([]);
-        }
-
-        const orderedMessages = TestAgent.constructor.getMessagesForConversation({
-          messages: fakeMessages,
-          parentMessageId,
-        });
-
-        const chatMessages = orderedMessages.map((msg) =>
-          msg?.isCreatedByUser || msg?.role?.toLowerCase() === 'user'
-            ? new HumanMessage(msg.text)
-            : new AIMessage(msg.text),
-        );
-
-        TestAgent.currentMessages = orderedMessages;
-        return Promise.resolve(chatMessages);
-      });
-    TestAgent.sendMessage = jest.fn().mockImplementation(async (message, opts = {}) => {
-      if (opts && typeof opts === 'object') {
-        TestAgent.setOptions(opts);
-      }
-      const conversationId = opts.conversationId || crypto.randomUUID();
-      const parentMessageId = opts.parentMessageId || Constants.NO_PARENT;
-      const userMessageId = opts.overrideParentMessageId || crypto.randomUUID();
-      this.pastMessages = await TestAgent.loadHistory(
-        conversationId,
-        TestAgent.options?.parentMessageId,
-      );
-
-      const userMessage = {
-        text: message,
-        sender: 'ChatGPT',
-        isCreatedByUser: true,
-        messageId: userMessageId,
-        parentMessageId,
-        conversationId,
-      };
-
-      const response = {
-        sender: 'ChatGPT',
-        text: 'Hello, User!',
-        isCreatedByUser: false,
-        messageId: crypto.randomUUID(),
-        parentMessageId: userMessage.messageId,
-        conversationId,
-      };
-
-      fakeMessages.push(userMessage);
-      fakeMessages.push(response);
-      return response;
-    });
-  });
-
-  test('initializes PluginsClient without crashing', () => {
-    expect(TestAgent).toBeInstanceOf(PluginsClient);
-  });
-
-  test('check setOptions function', () => {
-    expect(TestAgent.agentIsGpt3).toBe(true);
-  });
-
-  describe('sendMessage', () => {
-    test('sendMessage should return a response message', async () => {
-      const expectedResult = expect.objectContaining({
-        sender: 'ChatGPT',
-        text: expect.any(String),
-        isCreatedByUser: false,
-        messageId: expect.any(String),
-        parentMessageId: expect.any(String),
-        conversationId: expect.any(String),
-      });
-
-      const response = await TestAgent.sendMessage(userMessage);
-      parentMessageId = response.messageId;
-      conversationId = response.conversationId;
-      expect(response).toEqual(expectedResult);
-    });
-
-    test('sendMessage should work with provided conversationId and parentMessageId', async () => {
-      const userMessage = 'Second message in the conversation';
-      const opts = {
-        conversationId,
-        parentMessageId,
-      };
-
-      const expectedResult = expect.objectContaining({
-        sender: 'ChatGPT',
-        text: expect.any(String),
-        isCreatedByUser: false,
-        messageId: expect.any(String),
-        parentMessageId: expect.any(String),
-        conversationId: opts.conversationId,
-      });
-
-      const response = await TestAgent.sendMessage(userMessage, opts);
-      parentMessageId = response.messageId;
-      expect(response.conversationId).toEqual(conversationId);
-      expect(response).toEqual(expectedResult);
-    });
-
-    test('should return chat history', async () => {
-      const chatMessages = await TestAgent.loadHistory(conversationId, parentMessageId);
-      expect(TestAgent.currentMessages).toHaveLength(4);
-      expect(chatMessages[0].text).toEqual(userMessage);
-    });
-  });
-
-  describe('getFunctionModelName', () => {
-    let client;
-
-    beforeEach(() => {
-      client = new PluginsClient('dummy_api_key');
-    });
-
-    test('should return the input when it includes a dash followed by four digits', () => {
-      expect(client.getFunctionModelName('-1234')).toBe('-1234');
-      expect(client.getFunctionModelName('gpt-4-5678-preview')).toBe('gpt-4-5678-preview');
-    });
-
-    test('should return the input for all function-capable models (`0613` models and above)', () => {
-      expect(client.getFunctionModelName('gpt-4-0613')).toBe('gpt-4-0613');
-      expect(client.getFunctionModelName('gpt-4-32k-0613')).toBe('gpt-4-32k-0613');
-      expect(client.getFunctionModelName('gpt-3.5-turbo-0613')).toBe('gpt-3.5-turbo-0613');
-      expect(client.getFunctionModelName('gpt-3.5-turbo-16k-0613')).toBe('gpt-3.5-turbo-16k-0613');
-      expect(client.getFunctionModelName('gpt-3.5-turbo-1106')).toBe('gpt-3.5-turbo-1106');
-      expect(client.getFunctionModelName('gpt-4-1106-preview')).toBe('gpt-4-1106-preview');
-      expect(client.getFunctionModelName('gpt-4-1106')).toBe('gpt-4-1106');
-    });
-
-    test('should return the corresponding model if input is non-function capable (`0314` models)', () => {
-      expect(client.getFunctionModelName('gpt-4-0314')).toBe('gpt-4');
-      expect(client.getFunctionModelName('gpt-4-32k-0314')).toBe('gpt-4');
-      expect(client.getFunctionModelName('gpt-3.5-turbo-0314')).toBe('gpt-3.5-turbo');
-      expect(client.getFunctionModelName('gpt-3.5-turbo-16k-0314')).toBe('gpt-3.5-turbo');
-    });
-
-    test('should return "gpt-3.5-turbo" when the input includes "gpt-3.5-turbo"', () => {
-      expect(client.getFunctionModelName('test gpt-3.5-turbo model')).toBe('gpt-3.5-turbo');
-    });
-
-    test('should return "gpt-4" when the input includes "gpt-4"', () => {
-      expect(client.getFunctionModelName('testing gpt-4')).toBe('gpt-4');
-    });
-
-    test('should return "gpt-3.5-turbo" for input that does not meet any specific condition', () => {
-      expect(client.getFunctionModelName('random string')).toBe('gpt-3.5-turbo');
-      expect(client.getFunctionModelName('')).toBe('gpt-3.5-turbo');
-    });
-  });
-
-  describe('Azure OpenAI tests specific to Plugins', () => {
-    // TODO: add more tests for Azure OpenAI integration with Plugins
-    // let client;
-    // beforeEach(() => {
-    //   client = new PluginsClient('dummy_api_key');
-    // });
-
-    test('should not call getFunctionModelName when azure options are set', () => {
-      const spy = jest.spyOn(PluginsClient.prototype, 'getFunctionModelName');
-      const model = 'gpt-4-turbo';
-
-      // note, without the azure change in PR #1766, `getFunctionModelName` is called twice
-      const testClient = new PluginsClient('dummy_api_key', {
-        agentOptions: {
-          model,
-          agent: 'functions',
-        },
-        azure: defaultAzureOptions,
-      });
-
-      expect(spy).not.toHaveBeenCalled();
-      expect(testClient.agentOptions.model).toBe(model);
-
-      spy.mockRestore();
-    });
-  });
-
-  describe('sendMessage with filtered tools', () => {
-    let TestAgent;
-    const apiKey = 'fake-api-key';
-    const mockTools = [{ name: 'tool1' }, { name: 'tool2' }, { name: 'tool3' }, { name: 'tool4' }];
-
-    beforeEach(() => {
-      TestAgent = new PluginsClient(apiKey, {
-        tools: mockTools,
-        modelOptions: {
-          model: 'gpt-3.5-turbo',
-          temperature: 0,
-          max_tokens: 2,
-        },
-        agentOptions: {
-          model: 'gpt-3.5-turbo',
-        },
-      });
-
-      TestAgent.options.req = {
-        app: {
-          locals: {},
-        },
-      };
-
-      TestAgent.sendMessage = jest.fn().mockImplementation(async () => {
-        const { filteredTools = [], includedTools = [] } = TestAgent.options.req.app.locals;
-
-        if (includedTools.length > 0) {
-          const tools = TestAgent.options.tools.filter((plugin) =>
-            includedTools.includes(plugin.name),
-          );
-          TestAgent.options.tools = tools;
-        } else {
-          const tools = TestAgent.options.tools.filter(
-            (plugin) => !filteredTools.includes(plugin.name),
-          );
-          TestAgent.options.tools = tools;
-        }
-
-        return {
-          text: 'Mocked response',
-          tools: TestAgent.options.tools,
-        };
-      });
-    });
-
-    test('should filter out tools when filteredTools is provided', async () => {
-      TestAgent.options.req.app.locals.filteredTools = ['tool1', 'tool3'];
-      const response = await TestAgent.sendMessage('Test message');
-      expect(response.tools).toHaveLength(2);
-      expect(response.tools).toEqual(
-        expect.arrayContaining([
-          expect.objectContaining({ name: 'tool2' }),
-          expect.objectContaining({ name: 'tool4' }),
-        ]),
-      );
-    });
-
-    test('should only include specified tools when includedTools is provided', async () => {
-      TestAgent.options.req.app.locals.includedTools = ['tool2', 'tool4'];
-      const response = await TestAgent.sendMessage('Test message');
-      expect(response.tools).toHaveLength(2);
-      expect(response.tools).toEqual(
-        expect.arrayContaining([
-          expect.objectContaining({ name: 'tool2' }),
-          expect.objectContaining({ name: 'tool4' }),
-        ]),
-      );
-    });
-
-    test('should prioritize includedTools over filteredTools', async () => {
-      TestAgent.options.req.app.locals.filteredTools = ['tool1', 'tool3'];
-      TestAgent.options.req.app.locals.includedTools = ['tool1', 'tool2'];
-      const response = await TestAgent.sendMessage('Test message');
-      expect(response.tools).toHaveLength(2);
-      expect(response.tools).toEqual(
-        expect.arrayContaining([
-          expect.objectContaining({ name: 'tool1' }),
-          expect.objectContaining({ name: 'tool2' }),
-        ]),
-      );
-    });
-
-    test('should not modify tools when no filters are provided', async () => {
-      const response = await TestAgent.sendMessage('Test message');
-      expect(response.tools).toHaveLength(4);
-      expect(response.tools).toEqual(expect.arrayContaining(mockTools));
-    });
-  });
-});

api/app/clients/tools/structured/OpenAIImageTools.js

@@ -107,6 +107,12 @@ const getImageEditPromptDescription = () => {
   return process.env.IMAGE_EDIT_OAI_PROMPT_DESCRIPTION || DEFAULT_IMAGE_EDIT_PROMPT_DESCRIPTION;
 };
 
+function createAbortHandler() {
+  return function () {
+    logger.debug('[ImageGenOAI] Image generation aborted');
+  };
+}
+
 /**
  * Creates OpenAI Image tools (generation and editing)
  * @param {Object} fields - Configuration fields
@@ -201,10 +207,18 @@ function createOpenAIImageTools(fields = {}) {
       }
 
       let resp;
+      /** @type {AbortSignal} */
+      let derivedSignal = null;
+      /** @type {() => void} */
+      let abortHandler = null;
+
       try {
-        const derivedSignal = runnableConfig?.signal
-          ? AbortSignal.any([runnableConfig.signal])
-          : undefined;
+        if (runnableConfig?.signal) {
+          derivedSignal = AbortSignal.any([runnableConfig.signal]);
+          abortHandler = createAbortHandler();
+          derivedSignal.addEventListener('abort', abortHandler, { once: true });
+        }
+
         resp = await openai.images.generate(
           {
             model: 'gpt-image-1',
@@ -228,6 +242,10 @@ function createOpenAIImageTools(fields = {}) {
         logAxiosError({ error, message });
         return returnValue(`Something went wrong when trying to generate the image. The OpenAI API may be unavailable:
 Error Message: ${error.message}`);
+      } finally {
+        if (abortHandler && derivedSignal) {
+          derivedSignal.removeEventListener('abort', abortHandler);
+        }
       }
 
       if (!resp) {
@@ -409,10 +427,17 @@ Error Message: ${error.message}`);
         headers['Authorization'] = `Bearer ${apiKey}`;
       }
 
+      /** @type {AbortSignal} */
+      let derivedSignal = null;
+      /** @type {() => void} */
+      let abortHandler = null;
+
       try {
-        const derivedSignal = runnableConfig?.signal
-          ? AbortSignal.any([runnableConfig.signal])
-          : undefined;
+        if (runnableConfig?.signal) {
+          derivedSignal = AbortSignal.any([runnableConfig.signal]);
+          abortHandler = createAbortHandler();
+          derivedSignal.addEventListener('abort', abortHandler, { once: true });
+        }
 
         /** @type {import('axios').AxiosRequestConfig} */
         const axiosConfig = {
@@ -467,6 +492,10 @@ Error Message: ${error.message}`);
         logAxiosError({ error, message });
         return returnValue(`Something went wrong when trying to edit the image. The OpenAI API may be unavailable:
 Error Message: ${error.message || 'Unknown error'}`);
+      } finally {
+        if (abortHandler && derivedSignal) {
+          derivedSignal.removeEventListener('abort', abortHandler);
+        }
       }
     },
     {

api/app/clients/tools/util/fileSearch.js

@@ -1,9 +1,10 @@
 const { z } = require('zod');
 const axios = require('axios');
 const { tool } = require('@langchain/core/tools');
+const { logger } = require('@librechat/data-schemas');
 const { Tools, EToolResources } = require('librechat-data-provider');
+const { generateShortLivedToken } = require('~/server/services/AuthService');
 const { getFiles } = require('~/models/File');
-const { logger } = require('~/config');
 
 /**
  *
@@ -59,7 +60,7 @@ const createFileSearchTool = async ({ req, files, entity_id }) => {
       if (files.length === 0) {
         return 'No files to search. Instruct the user to add files for the search.';
       }
-      const jwtToken = req.headers.authorization.split(' ')[1];
+      const jwtToken = generateShortLivedToken(req.user.id);
       if (!jwtToken) {
         return 'There was an error authenticating the file search request.';
       }

api/cache/banViolation.js

@@ -1,7 +1,8 @@
 const { logger } = require('@librechat/data-schemas');
+const { isEnabled, math } = require('@librechat/api');
 const { ViolationTypes } = require('librechat-data-provider');
-const { isEnabled, math, removePorts } = require('~/server/utils');
 const { deleteAllUserSessions } = require('~/models');
+const { removePorts } = require('~/server/utils');
 const getLogStores = require('./getLogStores');
 
 const { BAN_VIOLATIONS, BAN_INTERVAL } = process.env ?? {};

api/cache/getLogStores.js

@@ -1,7 +1,7 @@
 const { Keyv } = require('keyv');
+const { isEnabled, math } = require('@librechat/api');
 const { CacheKeys, ViolationTypes, Time } = require('librechat-data-provider');
 const { logFile, violationFile } = require('./keyvFiles');
-const { isEnabled, math } = require('~/server/utils');
 const keyvRedis = require('./keyvRedis');
 const keyvMongo = require('./keyvMongo');
 

api/db/indexSync.js

@@ -1,8 +1,11 @@
 const mongoose = require('mongoose');
 const { MeiliSearch } = require('meilisearch');
 const { logger } = require('@librechat/data-schemas');
+const { FlowStateManager } = require('@librechat/api');
+const { CacheKeys } = require('librechat-data-provider');
 
 const { isEnabled } = require('~/server/utils');
+const { getLogStores } = require('~/cache');
 
 const Conversation = mongoose.models.Conversation;
 const Message = mongoose.models.Message;
@@ -28,43 +31,123 @@ class MeiliSearchClient {
   }
 }
 
+/**
+ * Performs the actual sync operations for messages and conversations
+ */
+async function performSync() {
+  const client = MeiliSearchClient.getInstance();
+
+  const { status } = await client.health();
+  if (status !== 'available') {
+    throw new Error('Meilisearch not available');
+  }
+
+  if (indexingDisabled === true) {
+    logger.info('[indexSync] Indexing is disabled, skipping...');
+    return { messagesSync: false, convosSync: false };
+  }
+
+  let messagesSync = false;
+  let convosSync = false;
+
+  // Check if we need to sync messages
+  const messageProgress = await Message.getSyncProgress();
+  if (!messageProgress.isComplete) {
+    logger.info(
+      `[indexSync] Messages need syncing: ${messageProgress.totalProcessed}/${messageProgress.totalDocuments} indexed`,
+    );
+
+    // Check if we should do a full sync or incremental
+    const messageCount = await Message.countDocuments();
+    const messagesIndexed = messageProgress.totalProcessed;
+    const syncThreshold = parseInt(process.env.MEILI_SYNC_THRESHOLD || '1000', 10);
+
+    if (messageCount - messagesIndexed > syncThreshold) {
+      logger.info('[indexSync] Starting full message sync due to large difference');
+      await Message.syncWithMeili();
+      messagesSync = true;
+    } else if (messageCount !== messagesIndexed) {
+      logger.warn('[indexSync] Messages out of sync, performing incremental sync');
+      await Message.syncWithMeili();
+      messagesSync = true;
+    }
+  } else {
+    logger.info(
+      `[indexSync] Messages are fully synced: ${messageProgress.totalProcessed}/${messageProgress.totalDocuments}`,
+    );
+  }
+
+  // Check if we need to sync conversations
+  const convoProgress = await Conversation.getSyncProgress();
+  if (!convoProgress.isComplete) {
+    logger.info(
+      `[indexSync] Conversations need syncing: ${convoProgress.totalProcessed}/${convoProgress.totalDocuments} indexed`,
+    );
+
+    const convoCount = await Conversation.countDocuments();
+    const convosIndexed = convoProgress.totalProcessed;
+    const syncThreshold = parseInt(process.env.MEILI_SYNC_THRESHOLD || '1000', 10);
+
+    if (convoCount - convosIndexed > syncThreshold) {
+      logger.info('[indexSync] Starting full conversation sync due to large difference');
+      await Conversation.syncWithMeili();
+      convosSync = true;
+    } else if (convoCount !== convosIndexed) {
+      logger.warn('[indexSync] Convos out of sync, performing incremental sync');
+      await Conversation.syncWithMeili();
+      convosSync = true;
+    }
+  } else {
+    logger.info(
+      `[indexSync] Conversations are fully synced: ${convoProgress.totalProcessed}/${convoProgress.totalDocuments}`,
+    );
+  }
+
+  return { messagesSync, convosSync };
+}
+
+/**
+ * Main index sync function that uses FlowStateManager to prevent concurrent execution
+ */
 async function indexSync() {
   if (!searchEnabled) {
     return;
   }
-  try {
-    const client = MeiliSearchClient.getInstance();
 
-    const { status } = await client.health();
-    if (status !== 'available') {
-      throw new Error('Meilisearch not available');
+  logger.info('[indexSync] Starting index synchronization check...');
+
+  try {
+    // Get or create FlowStateManager instance
+    const flowsCache = getLogStores(CacheKeys.FLOWS);
+    if (!flowsCache) {
+      logger.warn('[indexSync] Flows cache not available, falling back to direct sync');
+      return await performSync();
     }
 
-    if (indexingDisabled === true) {
-      logger.info('[indexSync] Indexing is disabled, skipping...');
+    const flowManager = new FlowStateManager(flowsCache, {
+      ttl: 60000 * 10, // 10 minutes TTL for sync operations
+    });
+
+    // Use a unique flow ID for the sync operation
+    const flowId = 'meili-index-sync';
+    const flowType = 'MEILI_SYNC';
+
+    // This will only execute the handler if no other instance is running the sync
+    const result = await flowManager.createFlowWithHandler(flowId, flowType, performSync);
+
+    if (result.messagesSync || result.convosSync) {
+      logger.info('[indexSync] Sync completed successfully');
+    } else {
+      logger.debug('[indexSync] No sync was needed');
+    }
+
+    return result;
+  } catch (err) {
+    if (err.message.includes('flow already exists')) {
+      logger.info('[indexSync] Sync already running on another instance');
       return;
     }
 
-    const messageCount = await Message.countDocuments();
-    const convoCount = await Conversation.countDocuments();
-    const messages = await client.index('messages').getStats();
-    const convos = await client.index('convos').getStats();
-    const messagesIndexed = messages.numberOfDocuments;
-    const convosIndexed = convos.numberOfDocuments;
-
-    logger.debug(`[indexSync] There are ${messageCount} messages and ${messagesIndexed} indexed`);
-    logger.debug(`[indexSync] There are ${convoCount} convos and ${convosIndexed} indexed`);
-
-    if (messageCount !== messagesIndexed) {
-      logger.debug('[indexSync] Messages out of sync, indexing');
-      Message.syncWithMeili();
-    }
-
-    if (convoCount !== convosIndexed) {
-      logger.debug('[indexSync] Convos out of sync, indexing');
-      Conversation.syncWithMeili();
-    }
-  } catch (err) {
     if (err.message.includes('not found')) {
       logger.debug('[indexSync] Creating indices...');
       currentTimeout = setTimeout(async () => {

api/models/Agent.js

@@ -70,6 +70,9 @@ const loadEphemeralAgent = async ({ req, agent_id, endpoint, model_parameters: _
   if (ephemeralAgent?.execute_code === true) {
     tools.push(Tools.execute_code);
   }
+  if (ephemeralAgent?.file_search === true) {
+    tools.push(Tools.file_search);
+  }
   if (ephemeralAgent?.web_search === true) {
     tools.push(Tools.web_search);
   }

api/models/Agent.spec.js

@@ -43,7 +43,7 @@ describe('models/Agent', () => {
       const mongoUri = mongoServer.getUri();
       Agent = mongoose.models.Agent || mongoose.model('Agent', agentSchema);
       await mongoose.connect(mongoUri);
-    });
+    }, 20000);
 
     afterAll(async () => {
       await mongoose.disconnect();
@@ -413,7 +413,7 @@ describe('models/Agent', () => {
       const mongoUri = mongoServer.getUri();
       Agent = mongoose.models.Agent || mongoose.model('Agent', agentSchema);
       await mongoose.connect(mongoUri);
-    });
+    }, 20000);
 
     afterAll(async () => {
       await mongoose.disconnect();
@@ -670,7 +670,7 @@ describe('models/Agent', () => {
       const mongoUri = mongoServer.getUri();
       Agent = mongoose.models.Agent || mongoose.model('Agent', agentSchema);
       await mongoose.connect(mongoUri);
-    });
+    }, 20000);
 
     afterAll(async () => {
       await mongoose.disconnect();
@@ -1332,7 +1332,7 @@ describe('models/Agent', () => {
       const mongoUri = mongoServer.getUri();
       Agent = mongoose.models.Agent || mongoose.model('Agent', agentSchema);
       await mongoose.connect(mongoUri);
-    });
+    }, 20000);
 
     afterAll(async () => {
       await mongoose.disconnect();
@@ -1514,7 +1514,7 @@ describe('models/Agent', () => {
       const mongoUri = mongoServer.getUri();
       Agent = mongoose.models.Agent || mongoose.model('Agent', agentSchema);
       await mongoose.connect(mongoUri);
-    });
+    }, 20000);
 
     afterAll(async () => {
       await mongoose.disconnect();
@@ -1798,7 +1798,7 @@ describe('models/Agent', () => {
       const mongoUri = mongoServer.getUri();
       Agent = mongoose.models.Agent || mongoose.model('Agent', agentSchema);
       await mongoose.connect(mongoUri);
-    });
+    }, 20000);
 
     afterAll(async () => {
       await mongoose.disconnect();
@@ -2350,7 +2350,7 @@ describe('models/Agent', () => {
       const mongoUri = mongoServer.getUri();
       Agent = mongoose.models.Agent || mongoose.model('Agent', agentSchema);
       await mongoose.connect(mongoUri);
-    });
+    }, 20000);
 
     afterAll(async () => {
       await mongoose.disconnect();

api/models/Conversation.js

@@ -1,4 +1,6 @@
 const { logger } = require('@librechat/data-schemas');
+const { createTempChatExpirationDate } = require('@librechat/api');
+const getCustomConfig = require('~/server/services/Config/loadCustomConfig');
 const { getMessages, deleteMessages } = require('./Message');
 const { Conversation } = require('~/db/models');
 
@@ -98,10 +100,15 @@ module.exports = {
         update.conversationId = newConversationId;
       }
 
-      if (req.body.isTemporary) {
-        const expiredAt = new Date();
-        expiredAt.setDate(expiredAt.getDate() + 30);
-        update.expiredAt = expiredAt;
+      if (req?.body?.isTemporary) {
+        try {
+          const customConfig = await getCustomConfig();
+          update.expiredAt = createTempChatExpirationDate(customConfig);
+        } catch (err) {
+          logger.error('Error creating temporary chat expiration date:', err);
+          logger.info(`---\`saveConvo\` context: ${metadata?.context}`);
+          update.expiredAt = null;
+        }
       } else {
         update.expiredAt = null;
       }

api/models/File.js

@@ -1,5 +1,5 @@
 const { logger } = require('@librechat/data-schemas');
-const { EToolResources } = require('librechat-data-provider');
+const { EToolResources, FileContext } = require('librechat-data-provider');
 const { File } = require('~/db/models');
 
 /**
@@ -32,19 +32,19 @@ const getFiles = async (filter, _sortOptions, selectFields = { text: 0 }) => {
  * @returns {Promise<Array<MongoFile>>} Files that match the criteria
  */
 const getToolFilesByIds = async (fileIds, toolResourceSet) => {
-  if (!fileIds || !fileIds.length) {
+  if (!fileIds || !fileIds.length || !toolResourceSet?.size) {
     return [];
   }
 
   try {
     const filter = {
       file_id: { $in: fileIds },
+      $or: [],
     };
 
-    if (toolResourceSet.size) {
-      filter.$or = [];
+    if (toolResourceSet.has(EToolResources.ocr)) {
+      filter.$or.push({ text: { $exists: true, $ne: null }, context: FileContext.agents });
     }
-
     if (toolResourceSet.has(EToolResources.file_search)) {
       filter.$or.push({ embedded: true });
     }

api/models/Message.js

@@ -1,5 +1,7 @@
 const { z } = require('zod');
 const { logger } = require('@librechat/data-schemas');
+const { createTempChatExpirationDate } = require('@librechat/api');
+const getCustomConfig = require('~/server/services/Config/loadCustomConfig');
 const { Message } = require('~/db/models');
 
 const idSchema = z.string().uuid();
@@ -54,9 +56,14 @@ async function saveMessage(req, params, metadata) {
     };
 
     if (req?.body?.isTemporary) {
-      const expiredAt = new Date();
-      expiredAt.setDate(expiredAt.getDate() + 30);
-      update.expiredAt = expiredAt;
+      try {
+        const customConfig = await getCustomConfig();
+        update.expiredAt = createTempChatExpirationDate(customConfig);
+      } catch (err) {
+        logger.error('Error creating temporary chat expiration date:', err);
+        logger.info(`---\`saveMessage\` context: ${metadata?.context}`);
+        update.expiredAt = null;
+      }
     } else {
       update.expiredAt = null;
     }

api/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@librechat/backend",
-  "version": "v0.7.8",
+  "version": "v0.7.9-rc1",
   "description": "",
   "scripts": {
     "start": "echo 'please run this from the root directory'",
@@ -48,14 +48,13 @@
     "@langchain/google-genai": "^0.2.13",
     "@langchain/google-vertexai": "^0.2.13",
     "@langchain/textsplitters": "^0.1.0",
-    "@librechat/agents": "^2.4.41",
+    "@librechat/agents": "^2.4.50",
     "@librechat/api": "*",
     "@librechat/data-schemas": "*",
     "@node-saml/passport-saml": "^5.0.0",
     "@waylaidwanderer/fetch-event-source": "^3.0.1",
     "axios": "^1.8.2",
     "bcryptjs": "^2.4.3",
-    "cohere-ai": "^7.9.1",
     "compression": "^1.7.4",
     "connect-redis": "^7.1.0",
     "cookie": "^0.7.2",

api/server/cleanup.js

@@ -169,9 +169,6 @@ function disposeClient(client) {
       client.isGenerativeModel = null;
     }
     // Properties specific to OpenAIClient
-    if (client.ChatGPTClient) {
-      client.ChatGPTClient = null;
-    }
     if (client.completionsUrl) {
       client.completionsUrl = null;
     }

api/server/controllers/AskController.js

@@ -1,282 +0,0 @@
-const { getResponseSender, Constants } = require('librechat-data-provider');
-const {
-  handleAbortError,
-  createAbortController,
-  cleanupAbortController,
-} = require('~/server/middleware');
-const {
-  disposeClient,
-  processReqData,
-  clientRegistry,
-  requestDataMap,
-} = require('~/server/cleanup');
-const { sendMessage, createOnProgress } = require('~/server/utils');
-const { saveMessage } = require('~/models');
-const { logger } = require('~/config');
-
-const AskController = async (req, res, next, initializeClient, addTitle) => {
-  let {
-    text,
-    endpointOption,
-    conversationId,
-    modelDisplayLabel,
-    parentMessageId = null,
-    overrideParentMessageId = null,
-  } = req.body;
-
-  let client = null;
-  let abortKey = null;
-  let cleanupHandlers = [];
-  let clientRef = null;
-
-  logger.debug('[AskController]', {
-    text,
-    conversationId,
-    ...endpointOption,
-    modelsConfig: endpointOption?.modelsConfig ? 'exists' : '',
-  });
-
-  let userMessage = null;
-  let userMessagePromise = null;
-  let promptTokens = null;
-  let userMessageId = null;
-  let responseMessageId = null;
-  let getAbortData = null;
-
-  const sender = getResponseSender({
-    ...endpointOption,
-    model: endpointOption.modelOptions.model,
-    modelDisplayLabel,
-  });
-  const initialConversationId = conversationId;
-  const newConvo = !initialConversationId;
-  const userId = req.user.id;
-
-  let reqDataContext = {
-    userMessage,
-    userMessagePromise,
-    responseMessageId,
-    promptTokens,
-    conversationId,
-    userMessageId,
-  };
-
-  const updateReqData = (data = {}) => {
-    reqDataContext = processReqData(data, reqDataContext);
-    abortKey = reqDataContext.abortKey;
-    userMessage = reqDataContext.userMessage;
-    userMessagePromise = reqDataContext.userMessagePromise;
-    responseMessageId = reqDataContext.responseMessageId;
-    promptTokens = reqDataContext.promptTokens;
-    conversationId = reqDataContext.conversationId;
-    userMessageId = reqDataContext.userMessageId;
-  };
-
-  let { onProgress: progressCallback, getPartialText } = createOnProgress();
-
-  const performCleanup = () => {
-    logger.debug('[AskController] Performing cleanup');
-    if (Array.isArray(cleanupHandlers)) {
-      for (const handler of cleanupHandlers) {
-        try {
-          if (typeof handler === 'function') {
-            handler();
-          }
-        } catch (e) {
-          // Ignore
-        }
-      }
-    }
-
-    if (abortKey) {
-      logger.debug('[AskController] Cleaning up abort controller');
-      cleanupAbortController(abortKey);
-      abortKey = null;
-    }
-
-    if (client) {
-      disposeClient(client);
-      client = null;
-    }
-
-    reqDataContext = null;
-    userMessage = null;
-    userMessagePromise = null;
-    promptTokens = null;
-    getAbortData = null;
-    progressCallback = null;
-    endpointOption = null;
-    cleanupHandlers = null;
-    addTitle = null;
-
-    if (requestDataMap.has(req)) {
-      requestDataMap.delete(req);
-    }
-    logger.debug('[AskController] Cleanup completed');
-  };
-
-  try {
-    ({ client } = await initializeClient({ req, res, endpointOption }));
-    if (clientRegistry && client) {
-      clientRegistry.register(client, { userId }, client);
-    }
-
-    if (client) {
-      requestDataMap.set(req, { client });
-    }
-
-    clientRef = new WeakRef(client);
-
-    getAbortData = () => {
-      const currentClient = clientRef?.deref();
-      const currentText =
-        currentClient?.getStreamText != null ? currentClient.getStreamText() : getPartialText();
-
-      return {
-        sender,
-        conversationId,
-        messageId: reqDataContext.responseMessageId,
-        parentMessageId: overrideParentMessageId ?? userMessageId,
-        text: currentText,
-        userMessage: userMessage,
-        userMessagePromise: userMessagePromise,
-        promptTokens: reqDataContext.promptTokens,
-      };
-    };
-
-    const { onStart, abortController } = createAbortController(
-      req,
-      res,
-      getAbortData,
-      updateReqData,
-    );
-
-    const closeHandler = () => {
-      logger.debug('[AskController] Request closed');
-      if (!abortController || abortController.signal.aborted || abortController.requestCompleted) {
-        return;
-      }
-      abortController.abort();
-      logger.debug('[AskController] Request aborted on close');
-    };
-
-    res.on('close', closeHandler);
-    cleanupHandlers.push(() => {
-      try {
-        res.removeListener('close', closeHandler);
-      } catch (e) {
-        // Ignore
-      }
-    });
-
-    const messageOptions = {
-      user: userId,
-      parentMessageId,
-      conversationId: reqDataContext.conversationId,
-      overrideParentMessageId,
-      getReqData: updateReqData,
-      onStart,
-      abortController,
-      progressCallback,
-      progressOptions: {
-        res,
-      },
-    };
-
-    /** @type {TMessage} */
-    let response = await client.sendMessage(text, messageOptions);
-    response.endpoint = endpointOption.endpoint;
-
-    const databasePromise = response.databasePromise;
-    delete response.databasePromise;
-
-    const { conversation: convoData = {} } = await databasePromise;
-    const conversation = { ...convoData };
-    conversation.title =
-      conversation && !conversation.title ? null : conversation?.title || 'New Chat';
-
-    const latestUserMessage = reqDataContext.userMessage;
-
-    if (client?.options?.attachments && latestUserMessage) {
-      latestUserMessage.files = client.options.attachments;
-      if (endpointOption?.modelOptions?.model) {
-        conversation.model = endpointOption.modelOptions.model;
-      }
-      delete latestUserMessage.image_urls;
-    }
-
-    if (!abortController.signal.aborted) {
-      const finalResponseMessage = { ...response };
-
-      sendMessage(res, {
-        final: true,
-        conversation,
-        title: conversation.title,
-        requestMessage: latestUserMessage,
-        responseMessage: finalResponseMessage,
-      });
-      res.end();
-
-      if (client?.savedMessageIds && !client.savedMessageIds.has(response.messageId)) {
-        await saveMessage(
-          req,
-          { ...finalResponseMessage, user: userId },
-          { context: 'api/server/controllers/AskController.js - response end' },
-        );
-      }
-    }
-
-    if (!client?.skipSaveUserMessage && latestUserMessage) {
-      await saveMessage(req, latestUserMessage, {
-        context: "api/server/controllers/AskController.js - don't skip saving user message",
-      });
-    }
-
-    if (typeof addTitle === 'function' && parentMessageId === Constants.NO_PARENT && newConvo) {
-      addTitle(req, {
-        text,
-        response: { ...response },
-        client,
-      })
-        .then(() => {
-          logger.debug('[AskController] Title generation started');
-        })
-        .catch((err) => {
-          logger.error('[AskController] Error in title generation', err);
-        })
-        .finally(() => {
-          logger.debug('[AskController] Title generation completed');
-          performCleanup();
-        });
-    } else {
-      performCleanup();
-    }
-  } catch (error) {
-    logger.error('[AskController] Error handling request', error);
-    let partialText = '';
-    try {
-      const currentClient = clientRef?.deref();
-      partialText =
-        currentClient?.getStreamText != null ? currentClient.getStreamText() : getPartialText();
-    } catch (getTextError) {
-      logger.error('[AskController] Error calling getText() during error handling', getTextError);
-    }
-
-    handleAbortError(res, req, error, {
-      sender,
-      partialText,
-      conversationId: reqDataContext.conversationId,
-      messageId: reqDataContext.responseMessageId,
-      parentMessageId: overrideParentMessageId ?? reqDataContext.userMessageId ?? parentMessageId,
-      userMessageId: reqDataContext.userMessageId,
-    })
-      .catch((err) => {
-        logger.error('[AskController] Error in `handleAbortError` during catch block', err);
-      })
-      .finally(() => {
-        performCleanup();
-      });
-  }
-};
-
-module.exports = AskController;

api/server/controllers/AuthController.js

@@ -1,17 +1,17 @@
 const cookies = require('cookie');
 const jwt = require('jsonwebtoken');
 const openIdClient = require('openid-client');
+const { isEnabled } = require('@librechat/api');
 const { logger } = require('@librechat/data-schemas');
 const {
-  registerUser,
-  resetPassword,
-  setAuthTokens,
   requestPasswordReset,
   setOpenIDAuthTokens,
+  resetPassword,
+  setAuthTokens,
+  registerUser,
 } = require('~/server/services/AuthService');
 const { findUser, getUserById, deleteAllUserSessions, findSession } = require('~/models');
 const { getOpenIdConfig } = require('~/strategies');
-const { isEnabled } = require('~/server/utils');
 
 const registrationController = async (req, res) => {
   try {

api/server/controllers/EditController.js

@@ -1,3 +1,5 @@
+const { sendEvent } = require('@librechat/api');
+const { logger } = require('@librechat/data-schemas');
 const { getResponseSender } = require('librechat-data-provider');
 const {
   handleAbortError,
@@ -10,9 +12,8 @@ const {
   clientRegistry,
   requestDataMap,
 } = require('~/server/cleanup');
-const { sendMessage, createOnProgress } = require('~/server/utils');
+const { createOnProgress } = require('~/server/utils');
 const { saveMessage } = require('~/models');
-const { logger } = require('~/config');
 
 const EditController = async (req, res, next, initializeClient) => {
   let {
@@ -84,7 +85,7 @@ const EditController = async (req, res, next, initializeClient) => {
     }
 
     if (abortKey) {
-      logger.debug('[AskController] Cleaning up abort controller');
+      logger.debug('[EditController] Cleaning up abort controller');
       cleanupAbortController(abortKey);
       abortKey = null;
     }
@@ -198,7 +199,7 @@ const EditController = async (req, res, next, initializeClient) => {
       const finalUserMessage = reqDataContext.userMessage;
       const finalResponseMessage = { ...response };
 
-      sendMessage(res, {
+      sendEvent(res, {
         final: true,
         conversation,
         title: conversation.title,

api/server/controllers/agents/__tests__/v1.spec.js

@@ -0,0 +1,195 @@
+const { duplicateAgent } = require('../v1');
+const { getAgent, createAgent } = require('~/models/Agent');
+const { getActions } = require('~/models/Action');
+const { nanoid } = require('nanoid');
+
+jest.mock('~/models/Agent');
+jest.mock('~/models/Action');
+jest.mock('nanoid');
+
+describe('duplicateAgent', () => {
+  let req, res;
+
+  beforeEach(() => {
+    req = {
+      params: { id: 'agent_123' },
+      user: { id: 'user_456' },
+    };
+    res = {
+      status: jest.fn().mockReturnThis(),
+      json: jest.fn(),
+    };
+    jest.clearAllMocks();
+  });
+
+  it('should duplicate an agent successfully', async () => {
+    const mockAgent = {
+      id: 'agent_123',
+      name: 'Test Agent',
+      description: 'Test Description',
+      instructions: 'Test Instructions',
+      provider: 'openai',
+      model: 'gpt-4',
+      tools: ['file_search'],
+      actions: [],
+      author: 'user_789',
+      versions: [{ name: 'Test Agent', version: 1 }],
+      __v: 0,
+    };
+
+    const mockNewAgent = {
+      id: 'agent_new_123',
+      name: 'Test Agent (1/2/23, 12:34)',
+      description: 'Test Description',
+      instructions: 'Test Instructions',
+      provider: 'openai',
+      model: 'gpt-4',
+      tools: ['file_search'],
+      actions: [],
+      author: 'user_456',
+      versions: [
+        {
+          name: 'Test Agent (1/2/23, 12:34)',
+          description: 'Test Description',
+          instructions: 'Test Instructions',
+          provider: 'openai',
+          model: 'gpt-4',
+          tools: ['file_search'],
+          actions: [],
+          createdAt: new Date(),
+          updatedAt: new Date(),
+        },
+      ],
+    };
+
+    getAgent.mockResolvedValue(mockAgent);
+    getActions.mockResolvedValue([]);
+    nanoid.mockReturnValue('new_123');
+    createAgent.mockResolvedValue(mockNewAgent);
+
+    await duplicateAgent(req, res);
+
+    expect(getAgent).toHaveBeenCalledWith({ id: 'agent_123' });
+    expect(getActions).toHaveBeenCalledWith({ agent_id: 'agent_123' }, true);
+    expect(createAgent).toHaveBeenCalledWith(
+      expect.objectContaining({
+        id: 'agent_new_123',
+        author: 'user_456',
+        name: expect.stringContaining('Test Agent ('),
+        description: 'Test Description',
+        instructions: 'Test Instructions',
+        provider: 'openai',
+        model: 'gpt-4',
+        tools: ['file_search'],
+        actions: [],
+      }),
+    );
+
+    expect(createAgent).toHaveBeenCalledWith(
+      expect.not.objectContaining({
+        versions: expect.anything(),
+        __v: expect.anything(),
+      }),
+    );
+
+    expect(res.status).toHaveBeenCalledWith(201);
+    expect(res.json).toHaveBeenCalledWith({
+      agent: mockNewAgent,
+      actions: [],
+    });
+  });
+
+  it('should ensure duplicated agent has clean versions array without nested fields', async () => {
+    const mockAgent = {
+      id: 'agent_123',
+      name: 'Test Agent',
+      description: 'Test Description',
+      versions: [
+        {
+          name: 'Test Agent',
+          versions: [{ name: 'Nested' }],
+          __v: 1,
+        },
+      ],
+      __v: 2,
+    };
+
+    const mockNewAgent = {
+      id: 'agent_new_123',
+      name: 'Test Agent (1/2/23, 12:34)',
+      description: 'Test Description',
+      versions: [
+        {
+          name: 'Test Agent (1/2/23, 12:34)',
+          description: 'Test Description',
+          createdAt: new Date(),
+          updatedAt: new Date(),
+        },
+      ],
+    };
+
+    getAgent.mockResolvedValue(mockAgent);
+    getActions.mockResolvedValue([]);
+    nanoid.mockReturnValue('new_123');
+    createAgent.mockResolvedValue(mockNewAgent);
+
+    await duplicateAgent(req, res);
+
+    expect(mockNewAgent.versions).toHaveLength(1);
+
+    const firstVersion = mockNewAgent.versions[0];
+    expect(firstVersion).not.toHaveProperty('versions');
+    expect(firstVersion).not.toHaveProperty('__v');
+
+    expect(mockNewAgent).not.toHaveProperty('__v');
+
+    expect(res.status).toHaveBeenCalledWith(201);
+  });
+
+  it('should return 404 if agent not found', async () => {
+    getAgent.mockResolvedValue(null);
+
+    await duplicateAgent(req, res);
+
+    expect(res.status).toHaveBeenCalledWith(404);
+    expect(res.json).toHaveBeenCalledWith({
+      error: 'Agent not found',
+      status: 'error',
+    });
+  });
+
+  it('should handle tool_resources.ocr correctly', async () => {
+    const mockAgent = {
+      id: 'agent_123',
+      name: 'Test Agent',
+      tool_resources: {
+        ocr: { enabled: true, config: 'test' },
+        other: { should: 'not be copied' },
+      },
+    };
+
+    getAgent.mockResolvedValue(mockAgent);
+    getActions.mockResolvedValue([]);
+    nanoid.mockReturnValue('new_123');
+    createAgent.mockResolvedValue({ id: 'agent_new_123' });
+
+    await duplicateAgent(req, res);
+
+    expect(createAgent).toHaveBeenCalledWith(
+      expect.objectContaining({
+        tool_resources: {
+          ocr: { enabled: true, config: 'test' },
+        },
+      }),
+    );
+  });
+
+  it('should handle errors gracefully', async () => {
+    getAgent.mockRejectedValue(new Error('Database error'));
+
+    await duplicateAgent(req, res);
+
+    expect(res.status).toHaveBeenCalledWith(500);
+    expect(res.json).toHaveBeenCalledWith({ error: 'Database error' });
+  });
+});

api/server/controllers/agents/client.js

@@ -4,11 +4,13 @@ const {
   sendEvent,
   createRun,
   Tokenizer,
+  checkAccess,
   memoryInstructions,
   createMemoryProcessor,
 } = require('@librechat/api');
 const {
   Callback,
+  Providers,
   GraphEvents,
   formatMessage,
   formatAgentMessages,
@@ -31,22 +33,29 @@ const {
 } = require('librechat-data-provider');
 const { DynamicStructuredTool } = require('@langchain/core/tools');
 const { getBufferString, HumanMessage } = require('@langchain/core/messages');
-const {
-  getCustomEndpointConfig,
-  createGetMCPAuthMap,
-  checkCapability,
-} = require('~/server/services/Config');
+const { createGetMCPAuthMap, checkCapability } = require('~/server/services/Config');
 const { addCacheControl, createContextHandlers } = require('~/app/clients/prompts');
 const { initializeAgent } = require('~/server/services/Endpoints/agents/agent');
 const { spendTokens, spendStructuredTokens } = require('~/models/spendTokens');
 const { getFormattedMemories, deleteMemory, setMemory } = require('~/models');
 const { encodeAndFormat } = require('~/server/services/Files/images/encode');
-const initOpenAI = require('~/server/services/Endpoints/openAI/initialize');
-const { checkAccess } = require('~/server/middleware/roles/access');
+const { getProviderConfig } = require('~/server/services/Endpoints');
 const BaseClient = require('~/app/clients/BaseClient');
+const { getRoleByName } = require('~/models/Role');
 const { loadAgent } = require('~/models/Agent');
 const { getMCPManager } = require('~/config');
 
+const omitTitleOptions = new Set([
+  'stream',
+  'thinking',
+  'streaming',
+  'clientOptions',
+  'thinkingConfig',
+  'thinkingBudget',
+  'includeThoughts',
+  'maxOutputTokens',
+]);
+
 /**
  * @param {ServerRequest} req
  * @param {Agent} agent
@@ -393,7 +402,12 @@ class AgentClient extends BaseClient {
     if (user.personalization?.memories === false) {
       return;
     }
-    const hasAccess = await checkAccess(user, PermissionTypes.MEMORIES, [Permissions.USE]);
+    const hasAccess = await checkAccess({
+      user,
+      permissionType: PermissionTypes.MEMORIES,
+      permissions: [Permissions.USE],
+      getRoleByName,
+    });
 
     if (!hasAccess) {
       logger.debug(
@@ -511,7 +525,10 @@ class AgentClient extends BaseClient {
           messagesToProcess = [...messages.slice(-messageWindowSize)];
         }
       }
-      return await this.processMemory(messagesToProcess);
+
+      const bufferString = getBufferString(messagesToProcess);
+      const bufferMessage = new HumanMessage(`# Current Chat:\n\n${bufferString}`);
+      return await this.processMemory([bufferMessage]);
     } catch (error) {
       logger.error('Memory Agent failed to process memory', error);
     }
@@ -677,7 +694,7 @@ class AgentClient extends BaseClient {
           hide_sequential_outputs: this.options.agent.hide_sequential_outputs,
           user: this.options.req.user,
         },
-        recursionLimit: agentsEConfig?.recursionLimit,
+        recursionLimit: agentsEConfig?.recursionLimit ?? 25,
         signal: abortController.signal,
         streamMode: 'values',
         version: 'v2',
@@ -983,23 +1000,26 @@ class AgentClient extends BaseClient {
       throw new Error('Run not initialized');
     }
     const { handleLLMEnd, collected: collectedMetadata } = createMetadataAggregator();
-    const endpoint = this.options.agent.endpoint;
-    const { req, res } = this.options;
+    const { req, res, agent } = this.options;
+    const endpoint = agent.endpoint;
+
     /** @type {import('@librechat/agents').ClientOptions} */
     let clientOptions = {
       maxTokens: 75,
+      model: agent.model_parameters.model,
     };
-    let endpointConfig = req.app.locals[endpoint];
+
+    const { getOptions, overrideProvider, customEndpointConfig } =
+      await getProviderConfig(endpoint);
+
+    /** @type {TEndpoint | undefined} */
+    const endpointConfig = req.app.locals[endpoint] ?? customEndpointConfig;
     if (!endpointConfig) {
-      try {
-        endpointConfig = await getCustomEndpointConfig(endpoint);
-      } catch (err) {
-        logger.error(
-          '[api/server/controllers/agents/client.js #titleConvo] Error getting custom endpoint config',
-          err,
-        );
-      }
+      logger.warn(
+        '[api/server/controllers/agents/client.js #titleConvo] Error getting endpoint config',
+      );
     }
+
     if (
       endpointConfig &&
       endpointConfig.titleModel &&
@@ -1007,30 +1027,50 @@ class AgentClient extends BaseClient {
     ) {
       clientOptions.model = endpointConfig.titleModel;
     }
+
+    const options = await getOptions({
+      req,
+      res,
+      optionsOnly: true,
+      overrideEndpoint: endpoint,
+      overrideModel: clientOptions.model,
+      endpointOption: { model_parameters: clientOptions },
+    });
+
+    let provider = options.provider ?? overrideProvider ?? agent.provider;
     if (
       endpoint === EModelEndpoint.azureOpenAI &&
-      clientOptions.model &&
-      this.options.agent.model_parameters.model !== clientOptions.model
+      options.llmConfig?.azureOpenAIApiInstanceName == null
     ) {
-      clientOptions =
-        (
-          await initOpenAI({
-            req,
-            res,
-            optionsOnly: true,
-            overrideModel: clientOptions.model,
-            overrideEndpoint: endpoint,
-            endpointOption: {
-              model_parameters: clientOptions,
-            },
-          })
-        )?.llmConfig ?? clientOptions;
+      provider = Providers.OPENAI;
     }
-    if (/\b(o\d)\b/i.test(clientOptions.model) && clientOptions.maxTokens != null) {
+
+    /** @type {import('@librechat/agents').ClientOptions} */
+    clientOptions = { ...options.llmConfig };
+    if (options.configOptions) {
+      clientOptions.configuration = options.configOptions;
+    }
+
+    // Ensure maxTokens is set for non-o1 models
+    if (!/\b(o\d)\b/i.test(clientOptions.model) && !clientOptions.maxTokens) {
+      clientOptions.maxTokens = 75;
+    } else if (/\b(o\d)\b/i.test(clientOptions.model) && clientOptions.maxTokens != null) {
       delete clientOptions.maxTokens;
     }
+
+    clientOptions = Object.assign(
+      Object.fromEntries(
+        Object.entries(clientOptions).filter(([key]) => !omitTitleOptions.has(key)),
+      ),
+    );
+
+    if (provider === Providers.GOOGLE) {
+      clientOptions.json = true;
+    }
+
     try {
       const titleResult = await this.run.generateTitle({
+        provider,
         inputText: text,
         contentParts: this.contentParts,
         clientOptions,
@@ -1048,8 +1088,10 @@ class AgentClient extends BaseClient {
         let input_tokens, output_tokens;
 
         if (item.usage) {
-          input_tokens = item.usage.input_tokens || item.usage.inputTokens;
-          output_tokens = item.usage.output_tokens || item.usage.outputTokens;
+          input_tokens =
+            item.usage.prompt_tokens || item.usage.input_tokens || item.usage.inputTokens;
+          output_tokens =
+            item.usage.completion_tokens || item.usage.output_tokens || item.usage.outputTokens;
         } else if (item.tokenUsage) {
           input_tokens = item.tokenUsage.promptTokens;
           output_tokens = item.tokenUsage.completionTokens;

api/server/controllers/agents/errors.js

@@ -1,10 +1,10 @@
 // errorHandler.js
-const { logger } = require('~/config');
-const getLogStores = require('~/cache/getLogStores');
+const { logger } = require('@librechat/data-schemas');
 const { CacheKeys, ViolationTypes } = require('librechat-data-provider');
+const { sendResponse } = require('~/server/middleware/error');
 const { recordUsage } = require('~/server/services/Threads');
 const { getConvo } = require('~/models/Conversation');
-const { sendResponse } = require('~/server/utils');
+const getLogStores = require('~/cache/getLogStores');
 
 /**
  * @typedef {Object} ErrorHandlerContext
@@ -75,7 +75,7 @@ const createErrorHandler = ({ req, res, getContext, originPath = '/assistants/ch
     } else if (/Files.*are invalid/.test(error.message)) {
       const errorMessage = `Files are invalid, or may not have uploaded yet.${
         endpoint === 'azureAssistants'
-          ? ' If using Azure OpenAI, files are only available in the region of the assistant\'s model at the time of upload.'
+          ? " If using Azure OpenAI, files are only available in the region of the assistant's model at the time of upload."
           : ''
       }`;
       return sendResponse(req, res, messageData, errorMessage);

api/server/controllers/agents/llm.js

@@ -1,106 +0,0 @@
-const { HttpsProxyAgent } = require('https-proxy-agent');
-const { resolveHeaders } = require('librechat-data-provider');
-const { createLLM } = require('~/app/clients/llm');
-
-/**
- * Initializes and returns a Language Learning Model (LLM) instance.
- *
- * @param {Object} options - Configuration options for the LLM.
- * @param {string} options.model - The model identifier.
- * @param {string} options.modelName - The specific name of the model.
- * @param {number} options.temperature - The temperature setting for the model.
- * @param {number} options.presence_penalty - The presence penalty for the model.
- * @param {number} options.frequency_penalty - The frequency penalty for the model.
- * @param {number} options.max_tokens - The maximum number of tokens for the model output.
- * @param {boolean} options.streaming - Whether to use streaming for the model output.
- * @param {Object} options.context - The context for the conversation.
- * @param {number} options.tokenBuffer - The token buffer size.
- * @param {number} options.initialMessageCount - The initial message count.
- * @param {string} options.conversationId - The ID of the conversation.
- * @param {string} options.user - The user identifier.
- * @param {string} options.langchainProxy - The langchain proxy URL.
- * @param {boolean} options.useOpenRouter - Whether to use OpenRouter.
- * @param {Object} options.options - Additional options.
- * @param {Object} options.options.headers - Custom headers for the request.
- * @param {string} options.options.proxy - Proxy URL.
- * @param {Object} options.options.req - The request object.
- * @param {Object} options.options.res - The response object.
- * @param {boolean} options.options.debug - Whether to enable debug mode.
- * @param {string} options.apiKey - The API key for authentication.
- * @param {Object} options.azure - Azure-specific configuration.
- * @param {Object} options.abortController - The AbortController instance.
- * @returns {Object} The initialized LLM instance.
- */
-function initializeLLM(options) {
-  const {
-    model,
-    modelName,
-    temperature,
-    presence_penalty,
-    frequency_penalty,
-    max_tokens,
-    streaming,
-    user,
-    langchainProxy,
-    useOpenRouter,
-    options: { headers, proxy },
-    apiKey,
-    azure,
-  } = options;
-
-  const modelOptions = {
-    modelName: modelName || model,
-    temperature,
-    presence_penalty,
-    frequency_penalty,
-    user,
-  };
-
-  if (max_tokens) {
-    modelOptions.max_tokens = max_tokens;
-  }
-
-  const configOptions = {};
-
-  if (langchainProxy) {
-    configOptions.basePath = langchainProxy;
-  }
-
-  if (useOpenRouter) {
-    configOptions.basePath = 'https://openrouter.ai/api/v1';
-    configOptions.baseOptions = {
-      headers: {
-        'HTTP-Referer': 'https://librechat.ai',
-        'X-Title': 'LibreChat',
-      },
-    };
-  }
-
-  if (headers && typeof headers === 'object' && !Array.isArray(headers)) {
-    configOptions.baseOptions = {
-      headers: resolveHeaders({
-        ...headers,
-        ...configOptions?.baseOptions?.headers,
-      }),
-    };
-  }
-
-  if (proxy) {
-    configOptions.httpAgent = new HttpsProxyAgent(proxy);
-    configOptions.httpsAgent = new HttpsProxyAgent(proxy);
-  }
-
-  const llm = createLLM({
-    modelOptions,
-    configOptions,
-    openAIApiKey: apiKey,
-    azure,
-    streaming,
-  });
-
-  return llm;
-}
-
-module.exports = {
-  initializeLLM,
-};

api/server/controllers/agents/request.js

@@ -1,3 +1,5 @@
+const { sendEvent } = require('@librechat/api');
+const { logger } = require('@librechat/data-schemas');
 const { Constants } = require('librechat-data-provider');
 const {
   handleAbortError,
@@ -5,17 +7,18 @@ const {
   cleanupAbortController,
 } = require('~/server/middleware');
 const { disposeClient, clientRegistry, requestDataMap } = require('~/server/cleanup');
-const { sendMessage } = require('~/server/utils');
 const { saveMessage } = require('~/models');
-const { logger } = require('~/config');
 
 const AgentController = async (req, res, next, initializeClient, addTitle) => {
   let {
     text,
     endpointOption,
     conversationId,
+    isContinued = false,
+    editedContent = null,
     parentMessageId = null,
     overrideParentMessageId = null,
+    responseMessageId: editedResponseMessageId = null,
   } = req.body;
 
   let sender;
@@ -67,7 +70,7 @@ const AgentController = async (req, res, next, initializeClient, addTitle) => {
             handler();
           }
         } catch (e) {
-          // Ignore cleanup errors
+          logger.error('[AgentController] Error in cleanup handler', e);
         }
       }
     }
@@ -155,7 +158,7 @@ const AgentController = async (req, res, next, initializeClient, addTitle) => {
       try {
         res.removeListener('close', closeHandler);
       } catch (e) {
-        // Ignore
+        logger.error('[AgentController] Error removing close listener', e);
       }
     });
 
@@ -163,10 +166,14 @@ const AgentController = async (req, res, next, initializeClient, addTitle) => {
       user: userId,
       onStart,
       getReqData,
+      isContinued,
+      editedContent,
       conversationId,
       parentMessageId,
       abortController,
       overrideParentMessageId,
+      isEdited: !!editedContent,
+      responseMessageId: editedResponseMessageId,
       progressOptions: {
         res,
       },
@@ -206,7 +213,7 @@ const AgentController = async (req, res, next, initializeClient, addTitle) => {
       // Create a new response object with minimal copies
       const finalResponse = { ...response };
 
-      sendMessage(res, {
+      sendEvent(res, {
         final: true,
         conversation,
         title: conversation.title,

api/server/controllers/agents/v1.js

@@ -242,6 +242,8 @@ const duplicateAgentHandler = async (req, res) => {
       createdAt: _createdAt,
       updatedAt: _updatedAt,
       tool_resources: _tool_resources = {},
+      versions: _versions,
+      __v: _v,
       ...cloneData
     } = agent;
     cloneData.name = `${agent.name} (${new Date().toLocaleString('en-US', {

api/server/controllers/assistants/chatV1.js

@@ -1,4 +1,7 @@
 const { v4 } = require('uuid');
+const { sleep } = require('@librechat/agents');
+const { sendEvent } = require('@librechat/api');
+const { logger } = require('@librechat/data-schemas');
 const {
   Time,
   Constants,
@@ -19,20 +22,20 @@ const {
   addThreadMetadata,
   saveAssistantMessage,
 } = require('~/server/services/Threads');
-const { sendResponse, sendMessage, sleep, countTokens } = require('~/server/utils');
 const { runAssistant, createOnTextProgress } = require('~/server/services/AssistantService');
 const validateAuthor = require('~/server/middleware/assistants/validateAuthor');
 const { formatMessage, createVisionPrompt } = require('~/app/clients/prompts');
 const { createRun, StreamRunManager } = require('~/server/services/Runs');
 const { addTitle } = require('~/server/services/Endpoints/assistants');
 const { createRunBody } = require('~/server/services/createRunBody');
+const { sendResponse } = require('~/server/middleware/error');
 const { getTransactions } = require('~/models/Transaction');
 const { checkBalance } = require('~/models/balanceMethods');
 const { getConvo } = require('~/models/Conversation');
 const getLogStores = require('~/cache/getLogStores');
+const { countTokens } = require('~/server/utils');
 const { getModelMaxTokens } = require('~/utils');
 const { getOpenAIClient } = require('./helpers');
-const { logger } = require('~/config');
 
 /**
  * @route POST /
@@ -471,7 +474,7 @@ const chatV1 = async (req, res) => {
     await Promise.all(promises);
 
     const sendInitialResponse = () => {
-      sendMessage(res, {
+      sendEvent(res, {
         sync: true,
         conversationId,
         // messages: previousMessages,
@@ -587,7 +590,7 @@ const chatV1 = async (req, res) => {
       iconURL: endpointOption.iconURL,
     };
 
-    sendMessage(res, {
+    sendEvent(res, {
       final: true,
       conversation,
       requestMessage: {

api/server/controllers/assistants/chatV2.js

@@ -1,4 +1,7 @@
 const { v4 } = require('uuid');
+const { sleep } = require('@librechat/agents');
+const { sendEvent } = require('@librechat/api');
+const { logger } = require('@librechat/data-schemas');
 const {
   Time,
   Constants,
@@ -22,15 +25,14 @@ const { createErrorHandler } = require('~/server/controllers/assistants/errors')
 const validateAuthor = require('~/server/middleware/assistants/validateAuthor');
 const { createRun, StreamRunManager } = require('~/server/services/Runs');
 const { addTitle } = require('~/server/services/Endpoints/assistants');
-const { sendMessage, sleep, countTokens } = require('~/server/utils');
 const { createRunBody } = require('~/server/services/createRunBody');
 const { getTransactions } = require('~/models/Transaction');
 const { checkBalance } = require('~/models/balanceMethods');
 const { getConvo } = require('~/models/Conversation');
 const getLogStores = require('~/cache/getLogStores');
+const { countTokens } = require('~/server/utils');
 const { getModelMaxTokens } = require('~/utils');
 const { getOpenAIClient } = require('./helpers');
-const { logger } = require('~/config');
 
 /**
  * @route POST /
@@ -309,7 +311,7 @@ const chatV2 = async (req, res) => {
     await Promise.all(promises);
 
     const sendInitialResponse = () => {
-      sendMessage(res, {
+      sendEvent(res, {
         sync: true,
         conversationId,
         // messages: previousMessages,
@@ -432,7 +434,7 @@ const chatV2 = async (req, res) => {
       iconURL: endpointOption.iconURL,
     };
 
-    sendMessage(res, {
+    sendEvent(res, {
       final: true,
       conversation,
       requestMessage: {

api/server/controllers/assistants/errors.js

@@ -1,10 +1,10 @@
 // errorHandler.js
-const { sendResponse } = require('~/server/utils');
-const { logger } = require('~/config');
-const getLogStores = require('~/cache/getLogStores');
+const { logger } = require('@librechat/data-schemas');
 const { CacheKeys, ViolationTypes, ContentTypes } = require('librechat-data-provider');
-const { getConvo } = require('~/models/Conversation');
 const { recordUsage, checkMessageGaps } = require('~/server/services/Threads');
+const { sendResponse } = require('~/server/middleware/error');
+const { getConvo } = require('~/models/Conversation');
+const getLogStores = require('~/cache/getLogStores');
 
 /**
  * @typedef {Object} ErrorHandlerContext
@@ -78,7 +78,7 @@ const createErrorHandler = ({ req, res, getContext, originPath = '/assistants/ch
     } else if (/Files.*are invalid/.test(error.message)) {
       const errorMessage = `Files are invalid, or may not have uploaded yet.${
         endpoint === 'azureAssistants'
-          ? ' If using Azure OpenAI, files are only available in the region of the assistant\'s model at the time of upload.'
+          ? " If using Azure OpenAI, files are only available in the region of the assistant's model at the time of upload."
           : ''
       }`;
       return sendResponse(req, res, messageData, errorMessage);

api/server/controllers/tools.js

@@ -1,5 +1,7 @@
 const { nanoid } = require('nanoid');
 const { EnvVar } = require('@librechat/agents');
+const { checkAccess } = require('@librechat/api');
+const { logger } = require('@librechat/data-schemas');
 const {
   Tools,
   AuthType,
@@ -13,9 +15,8 @@ const { processCodeOutput } = require('~/server/services/Files/Code/process');
 const { createToolCall, getToolCallsByConvo } = require('~/models/ToolCall');
 const { loadAuthValues } = require('~/server/services/Tools/credentials');
 const { loadTools } = require('~/app/clients/tools/util');
-const { checkAccess } = require('~/server/middleware');
+const { getRoleByName } = require('~/models/Role');
 const { getMessage } = require('~/models/Message');
-const { logger } = require('~/config');
 
 const fieldsMap = {
   [Tools.execute_code]: [EnvVar.CODE_API_KEY],
@@ -79,6 +80,7 @@ const verifyToolAuth = async (req, res) => {
         throwError: false,
       });
     } catch (error) {
+      logger.error('Error loading auth values', error);
       res.status(200).json({ authenticated: false, message: AuthType.USER_PROVIDED });
       return;
     }
@@ -132,7 +134,12 @@ const callTool = async (req, res) => {
     logger.debug(`[${toolId}/call] User: ${req.user.id}`);
     let hasAccess = true;
     if (toolAccessPermType[toolId]) {
-      hasAccess = await checkAccess(req.user, toolAccessPermType[toolId], [Permissions.USE]);
+      hasAccess = await checkAccess({
+        user: req.user,
+        permissionType: toolAccessPermType[toolId],
+        permissions: [Permissions.USE],
+        getRoleByName,
+      });
     }
     if (!hasAccess) {
       logger.warn(

api/server/index.js

@@ -39,7 +39,9 @@ const startServer = async () => {
   await connectDb();
 
   logger.info('Connected to MongoDB');
-  await indexSync();
+  indexSync().catch((err) => {
+    logger.error('[indexSync] Background sync failed:', err);
+  });
 
   app.disable('x-powered-by');
   app.set('trust proxy', trusted_proxy);
@@ -95,7 +97,6 @@ const startServer = async () => {
   app.use('/api/actions', routes.actions);
   app.use('/api/keys', routes.keys);
   app.use('/api/user', routes.user);
-  app.use('/api/ask', routes.ask);
   app.use('/api/search', routes.search);
   app.use('/api/edit', routes.edit);
   app.use('/api/messages', routes.messages);
@@ -116,7 +117,6 @@ const startServer = async () => {
   app.use('/api/roles', routes.roles);
   app.use('/api/agents', routes.agents);
   app.use('/api/banner', routes.banner);
-  app.use('/api/bedrock', routes.bedrock);
   app.use('/api/memories', routes.memories);
   app.use('/api/tags', routes.tags);
   app.use('/api/mcp', routes.mcp);

api/server/middleware/abortMiddleware.js

@@ -1,13 +1,13 @@
-// abortMiddleware.js
+const { logger } = require('@librechat/data-schemas');
+const { countTokens, isEnabled, sendEvent } = require('@librechat/api');
 const { isAssistantsEndpoint, ErrorTypes } = require('librechat-data-provider');
-const { sendMessage, sendError, countTokens, isEnabled } = require('~/server/utils');
 const { truncateText, smartTruncateText } = require('~/app/clients/prompts');
 const clearPendingReq = require('~/cache/clearPendingReq');
+const { sendError } = require('~/server/middleware/error');
 const { spendTokens } = require('~/models/spendTokens');
 const abortControllers = require('./abortControllers');
 const { saveMessage, getConvo } = require('~/models');
 const { abortRun } = require('./abortRun');
-const { logger } = require('~/config');
 
 const abortDataMap = new WeakMap();
 
@@ -101,7 +101,7 @@ async function abortMessage(req, res) {
   cleanupAbortController(abortKey);
 
   if (res.headersSent && finalEvent) {
-    return sendMessage(res, finalEvent);
+    return sendEvent(res, finalEvent);
   }
 
   res.setHeader('Content-Type', 'application/json');
@@ -174,7 +174,7 @@ const createAbortController = (req, res, getAbortData, getReqData) => {
    * @param {string} responseMessageId
    */
   const onStart = (userMessage, responseMessageId) => {
-    sendMessage(res, { message: userMessage, created: true });
+    sendEvent(res, { message: userMessage, created: true });
 
     const abortKey = userMessage?.conversationId ?? req.user.id;
     getReqData({ abortKey });

api/server/middleware/abortRun.js

@@ -1,11 +1,11 @@
+const { sendEvent } = require('@librechat/api');
+const { logger } = require('@librechat/data-schemas');
 const { CacheKeys, RunStatus, isUUID } = require('librechat-data-provider');
 const { initializeClient } = require('~/server/services/Endpoints/assistants');
 const { checkMessageGaps, recordUsage } = require('~/server/services/Threads');
 const { deleteMessages } = require('~/models/Message');
 const { getConvo } = require('~/models/Conversation');
 const getLogStores = require('~/cache/getLogStores');
-const { sendMessage } = require('~/server/utils');
-const { logger } = require('~/config');
 
 const three_minutes = 1000 * 60 * 3;
 
@@ -34,7 +34,7 @@ async function abortRun(req, res) {
   const [thread_id, run_id] = runValues.split(':');
 
   if (!run_id) {
-    logger.warn('[abortRun] Couldn\'t find run for cancel request', { thread_id });
+    logger.warn("[abortRun] Couldn't find run for cancel request", { thread_id });
     return res.status(204).send({ message: 'Run not found' });
   } else if (run_id === 'cancelled') {
     logger.warn('[abortRun] Run already cancelled', { thread_id });
@@ -93,7 +93,7 @@ async function abortRun(req, res) {
   };
 
   if (res.headersSent && finalEvent) {
-    return sendMessage(res, finalEvent);
+    return sendEvent(res, finalEvent);
   }
 
   res.json(finalEvent);

api/server/middleware/buildEndpointOption.js

@@ -1,13 +1,12 @@
+const { logger } = require('@librechat/data-schemas');
 const {
-  parseCompactConvo,
+  EndpointURLs,
   EModelEndpoint,
   isAgentsEndpoint,
-  EndpointURLs,
+  parseCompactConvo,
 } = require('librechat-data-provider');
 const azureAssistants = require('~/server/services/Endpoints/azureAssistants');
-const { getModelsConfig } = require('~/server/controllers/ModelController');
 const assistants = require('~/server/services/Endpoints/assistants');
-const gptPlugins = require('~/server/services/Endpoints/gptPlugins');
 const { processFiles } = require('~/server/services/Files/process');
 const anthropic = require('~/server/services/Endpoints/anthropic');
 const bedrock = require('~/server/services/Endpoints/bedrock');
@@ -25,7 +24,6 @@ const buildFunction = {
   [EModelEndpoint.bedrock]: bedrock.buildOptions,
   [EModelEndpoint.azureOpenAI]: openAI.buildOptions,
   [EModelEndpoint.anthropic]: anthropic.buildOptions,
-  [EModelEndpoint.gptPlugins]: gptPlugins.buildOptions,
   [EModelEndpoint.assistants]: assistants.buildOptions,
   [EModelEndpoint.azureAssistants]: azureAssistants.buildOptions,
 };
@@ -36,6 +34,9 @@ async function buildEndpointOption(req, res, next) {
   try {
     parsedBody = parseCompactConvo({ endpoint, endpointType, conversation: req.body });
   } catch (error) {
+    logger.warn(
+      `Error parsing conversation for endpoint ${endpoint}${error?.message ? `: ${error.message}` : ''}`,
+    );
     return handleError(res, { text: 'Error parsing conversation' });
   }
 
@@ -57,15 +58,6 @@ async function buildEndpointOption(req, res, next) {
       return handleError(res, { text: 'Model spec mismatch' });
     }
 
-    if (
-      currentModelSpec.preset.endpoint !== EModelEndpoint.gptPlugins &&
-      currentModelSpec.preset.tools
-    ) {
-      return handleError(res, {
-        text: `Only the "${EModelEndpoint.gptPlugins}" endpoint can have tools defined in the preset`,
-      });
-    }
-
     try {
       currentModelSpec.preset.spec = spec;
       if (currentModelSpec.iconURL != null && currentModelSpec.iconURL !== '') {
@@ -77,6 +69,7 @@ async function buildEndpointOption(req, res, next) {
         conversation: currentModelSpec.preset,
       });
     } catch (error) {
+      logger.error(`Error parsing model spec for endpoint ${endpoint}`, error);
       return handleError(res, { text: 'Error parsing model spec' });
     }
   }
@@ -84,20 +77,23 @@ async function buildEndpointOption(req, res, next) {
   try {
     const isAgents =
       isAgentsEndpoint(endpoint) || req.baseUrl.startsWith(EndpointURLs[EModelEndpoint.agents]);
-    const endpointFn = buildFunction[isAgents ? EModelEndpoint.agents : (endpointType ?? endpoint)];
-    const builder = isAgents ? (...args) => endpointFn(req, ...args) : endpointFn;
+    const builder = isAgents
+      ? (...args) => buildFunction[EModelEndpoint.agents](req, ...args)
+      : buildFunction[endpointType ?? endpoint];
 
     // TODO: use object params
     req.body.endpointOption = await builder(endpoint, parsedBody, endpointType);
 
-    // TODO: use `getModelsConfig` only when necessary
-    const modelsConfig = await getModelsConfig(req);
-    req.body.endpointOption.modelsConfig = modelsConfig;
     if (req.body.files && !isAgents) {
       req.body.endpointOption.attachments = processFiles(req.body.files);
     }
+
     next();
   } catch (error) {
+    logger.error(
+      `Error building endpoint option for endpoint ${endpoint} with type ${endpointType}`,
+      error,
+    );
     return handleError(res, { text: 'Error building endpoint option' });
   }
 }

api/server/middleware/denyRequest.js

@@ -1,6 +1,7 @@
 const crypto = require('crypto');
+const { sendEvent } = require('@librechat/api');
 const { getResponseSender, Constants } = require('librechat-data-provider');
-const { sendMessage, sendError } = require('~/server/utils');
+const { sendError } = require('~/server/middleware/error');
 const { saveMessage } = require('~/models');
 
 /**
@@ -36,7 +37,7 @@ const denyRequest = async (req, res, errorMessage) => {
     isCreatedByUser: true,
     text,
   };
-  sendMessage(res, { message: userMessage, created: true });
+  sendEvent(res, { message: userMessage, created: true });
 
   const shouldSaveMessage = _convoId && parentMessageId && parentMessageId !== Constants.NO_PARENT;
 

api/server/middleware/error.js

@@ -1,31 +1,9 @@
 const crypto = require('crypto');
+const { logger } = require('@librechat/data-schemas');
 const { parseConvo } = require('librechat-data-provider');
+const { sendEvent, handleError } = require('@librechat/api');
 const { saveMessage, getMessages } = require('~/models/Message');
 const { getConvo } = require('~/models/Conversation');
-const { logger } = require('~/config');
-
-/**
- * Sends error data in Server Sent Events format and ends the response.
- * @param {object} res - The server response.
- * @param {string} message - The error message.
- */
-const handleError = (res, message) => {
-  res.write(`event: error\ndata: ${JSON.stringify(message)}\n\n`);
-  res.end();
-};
-
-/**
- * Sends message data in Server Sent Events format.
- * @param {Express.Response} res - - The server response.
- * @param {string | Object} message - The message to be sent.
- * @param {'message' | 'error' | 'cancel'} event - [Optional] The type of event. Default is 'message'.
- */
-const sendMessage = (res, message, event = 'message') => {
-  if (typeof message === 'string' && message.length === 0) {
-    return;
-  }
-  res.write(`event: ${event}\ndata: ${JSON.stringify(message)}\n\n`);
-};
 
 /**
  * Processes an error with provided options, saves the error message and sends a corresponding SSE response
@@ -91,7 +69,7 @@ const sendError = async (req, res, options, callback) => {
       convo = parseConvo(errorMessage);
     }
 
-    return sendMessage(res, {
+    return sendEvent(res, {
       final: true,
       requestMessage: query?.[0] ? query[0] : requestMessage,
       responseMessage: errorMessage,
@@ -120,12 +98,10 @@ const sendResponse = (req, res, data, errorMessage) => {
   if (errorMessage) {
     return sendError(req, res, { ...data, text: errorMessage });
   }
-  return sendMessage(res, data);
+  return sendEvent(res, data);
 };
 
 module.exports = {
-  sendResponse,
-  handleError,
-  sendMessage,
   sendError,
+  sendResponse,
 };

api/server/middleware/roles/access.js

@@ -1,78 +0,0 @@
-const { getRoleByName } = require('~/models/Role');
-const { logger } = require('~/config');
-
-/**
- * Core function to check if a user has one or more required permissions
- *
- * @param {object} user - The user object
- * @param {PermissionTypes} permissionType - The type of permission to check
- * @param {Permissions[]} permissions - The list of specific permissions to check
- * @param {Record<Permissions, string[]>} [bodyProps] - An optional object where keys are permissions and values are arrays of properties to check
- * @param {object} [checkObject] - The object to check properties against
- * @returns {Promise<boolean>} Whether the user has the required permissions
- */
-const checkAccess = async (user, permissionType, permissions, bodyProps = {}, checkObject = {}) => {
-  if (!user) {
-    return false;
-  }
-
-  const role = await getRoleByName(user.role);
-  if (role && role.permissions && role.permissions[permissionType]) {
-    const hasAnyPermission = permissions.some((permission) => {
-      if (role.permissions[permissionType][permission]) {
-        return true;
-      }
-
-      if (bodyProps[permission] && checkObject) {
-        return bodyProps[permission].some((prop) =>
-          Object.prototype.hasOwnProperty.call(checkObject, prop),
-        );
-      }
-
-      return false;
-    });
-
-    return hasAnyPermission;
-  }
-
-  return false;
-};
-
-/**
- * Middleware to check if a user has one or more required permissions, optionally based on `req.body` properties.
- *
- * @param {PermissionTypes} permissionType - The type of permission to check.
- * @param {Permissions[]} permissions - The list of specific permissions to check.
- * @param {Record<Permissions, string[]>} [bodyProps] - An optional object where keys are permissions and values are arrays of `req.body` properties to check.
- * @returns {(req: ServerRequest, res: ServerResponse, next: NextFunction) => Promise<void>} Express middleware function.
- */
-const generateCheckAccess = (permissionType, permissions, bodyProps = {}) => {
-  return async (req, res, next) => {
-    try {
-      const hasAccess = await checkAccess(
-        req.user,
-        permissionType,
-        permissions,
-        bodyProps,
-        req.body,
-      );
-
-      if (hasAccess) {
-        return next();
-      }
-
-      logger.warn(
-        `[${permissionType}] Forbidden: Insufficient permissions for User ${req.user.id}: ${permissions.join(', ')}`,
-      );
-      return res.status(403).json({ message: 'Forbidden: Insufficient permissions' });
-    } catch (error) {
-      logger.error(error);
-      return res.status(500).json({ message: `Server error: ${error.message}` });
-    }
-  };
-};
-
-module.exports = {
-  checkAccess,
-  generateCheckAccess,
-};

api/server/middleware/roles/index.js

@@ -1,8 +1,5 @@
 const checkAdmin = require('./admin');
-const { checkAccess, generateCheckAccess } = require('./access');
 
 module.exports = {
   checkAdmin,
-  checkAccess,
-  generateCheckAccess,
 };

api/server/routes/agents/actions.js

@@ -1,14 +1,28 @@
 const express = require('express');
 const { nanoid } = require('nanoid');
-const { actionDelimiter, SystemRoles, removeNullishValues } = require('librechat-data-provider');
+const { logger } = require('@librechat/data-schemas');
+const { generateCheckAccess } = require('@librechat/api');
+const {
+  SystemRoles,
+  Permissions,
+  PermissionTypes,
+  actionDelimiter,
+  removeNullishValues,
+} = require('librechat-data-provider');
 const { encryptMetadata, domainParser } = require('~/server/services/ActionService');
 const { updateAction, getActions, deleteAction } = require('~/models/Action');
 const { isActionDomainAllowed } = require('~/server/services/domains');
 const { getAgent, updateAgent } = require('~/models/Agent');
-const { logger } = require('~/config');
+const { getRoleByName } = require('~/models/Role');
 
 const router = express.Router();
 
+const checkAgentCreate = generateCheckAccess({
+  permissionType: PermissionTypes.AGENTS,
+  permissions: [Permissions.USE, Permissions.CREATE],
+  getRoleByName,
+});
+
 // If the user has ADMIN role
 // then action edition is possible even if not owner of the assistant
 const isAdmin = (req) => {
@@ -41,7 +55,7 @@ router.get('/', async (req, res) => {
  * @param {ActionMetadata} req.body.metadata - Metadata for the action.
  * @returns {Object} 200 - success response - application/json
  */
-router.post('/:agent_id', async (req, res) => {
+router.post('/:agent_id', checkAgentCreate, async (req, res) => {
   try {
     const { agent_id } = req.params;
 
@@ -149,7 +163,7 @@ router.post('/:agent_id', async (req, res) => {
  * @param {string} req.params.action_id - The ID of the action to delete.
  * @returns {Object} 200 - success response - application/json
  */
-router.delete('/:agent_id/:action_id', async (req, res) => {
+router.delete('/:agent_id/:action_id', checkAgentCreate, async (req, res) => {
   try {
     const { agent_id, action_id } = req.params;
     const admin = isAdmin(req);

api/server/routes/agents/chat.js

@@ -1,22 +1,28 @@
 const express = require('express');
+const { generateCheckAccess, skipAgentCheck } = require('@librechat/api');
 const { PermissionTypes, Permissions } = require('librechat-data-provider');
 const {
   setHeaders,
   moderateText,
   // validateModel,
-  generateCheckAccess,
   validateConvoAccess,
   buildEndpointOption,
 } = require('~/server/middleware');
 const { initializeClient } = require('~/server/services/Endpoints/agents');
 const AgentController = require('~/server/controllers/agents/request');
 const addTitle = require('~/server/services/Endpoints/agents/title');
+const { getRoleByName } = require('~/models/Role');
 
 const router = express.Router();
 
 router.use(moderateText);
 
-const checkAgentAccess = generateCheckAccess(PermissionTypes.AGENTS, [Permissions.USE]);
+const checkAgentAccess = generateCheckAccess({
+  permissionType: PermissionTypes.AGENTS,
+  permissions: [Permissions.USE],
+  skipCheck: skipAgentCheck,
+  getRoleByName,
+});
 
 router.use(checkAgentAccess);
 router.use(validateConvoAccess);

api/server/routes/agents/v1.js

@@ -1,29 +1,36 @@
 const express = require('express');
+const { generateCheckAccess } = require('@librechat/api');
 const { PermissionTypes, Permissions } = require('librechat-data-provider');
-const { requireJwtAuth, generateCheckAccess } = require('~/server/middleware');
+const { requireJwtAuth } = require('~/server/middleware');
 const v1 = require('~/server/controllers/agents/v1');
+const { getRoleByName } = require('~/models/Role');
 const actions = require('./actions');
 const tools = require('./tools');
 
 const router = express.Router();
 const avatar = express.Router();
 
-const checkAgentAccess = generateCheckAccess(PermissionTypes.AGENTS, [Permissions.USE]);
-const checkAgentCreate = generateCheckAccess(PermissionTypes.AGENTS, [
-  Permissions.USE,
-  Permissions.CREATE,
-]);
+const checkAgentAccess = generateCheckAccess({
+  permissionType: PermissionTypes.AGENTS,
+  permissions: [Permissions.USE],
+  getRoleByName,
+});
+const checkAgentCreate = generateCheckAccess({
+  permissionType: PermissionTypes.AGENTS,
+  permissions: [Permissions.USE, Permissions.CREATE],
+  getRoleByName,
+});
 
-const checkGlobalAgentShare = generateCheckAccess(
-  PermissionTypes.AGENTS,
-  [Permissions.USE, Permissions.CREATE],
-  {
+const checkGlobalAgentShare = generateCheckAccess({
+  permissionType: PermissionTypes.AGENTS,
+  permissions: [Permissions.USE, Permissions.CREATE],
+  bodyProps: {
     [Permissions.SHARED_GLOBAL]: ['projectIds', 'removeProjectIds'],
   },
-);
+  getRoleByName,
+});
 
 router.use(requireJwtAuth);
-router.use(checkAgentAccess);
 
 /**
  * Agent actions route.

api/server/routes/ask/addToCache.js

@@ -1,63 +0,0 @@
-const { Keyv } = require('keyv');
-const { KeyvFile } = require('keyv-file');
-const { logger } = require('~/config');
-
-const addToCache = async ({ endpoint, endpointOption, userMessage, responseMessage }) => {
-  try {
-    const conversationsCache = new Keyv({
-      store: new KeyvFile({ filename: './data/cache.json' }),
-      namespace: 'chatgpt', // should be 'bing' for bing/sydney
-    });
-
-    const {
-      conversationId,
-      messageId: userMessageId,
-      parentMessageId: userParentMessageId,
-      text: userText,
-    } = userMessage;
-    const {
-      messageId: responseMessageId,
-      parentMessageId: responseParentMessageId,
-      text: responseText,
-    } = responseMessage;
-
-    let conversation = await conversationsCache.get(conversationId);
-    // used to generate a title for the conversation if none exists
-    // let isNewConversation = false;
-    if (!conversation) {
-      conversation = {
-        messages: [],
-        createdAt: Date.now(),
-      };
-      // isNewConversation = true;
-    }
-
-    const roles = (options) => {
-      if (endpoint === 'openAI') {
-        return options?.chatGptLabel || 'ChatGPT';
-      }
-    };
-
-    let _userMessage = {
-      id: userMessageId,
-      parentMessageId: userParentMessageId,
-      role: 'User',
-      message: userText,
-    };
-
-    let _responseMessage = {
-      id: responseMessageId,
-      parentMessageId: responseParentMessageId,
-      role: roles(endpointOption),
-      message: responseText,
-    };
-
-    conversation.messages.push(_userMessage, _responseMessage);
-
-    await conversationsCache.set(conversationId, conversation);
-  } catch (error) {
-    logger.error('[addToCache] Error adding conversation to cache', error);
-  }
-};
-
-module.exports = addToCache;

api/server/routes/ask/anthropic.js

@@ -1,25 +0,0 @@
-const express = require('express');
-const AskController = require('~/server/controllers/AskController');
-const { addTitle, initializeClient } = require('~/server/services/Endpoints/anthropic');
-const {
-  setHeaders,
-  handleAbort,
-  validateModel,
-  validateEndpoint,
-  buildEndpointOption,
-} = require('~/server/middleware');
-
-const router = express.Router();
-
-router.post(
-  '/',
-  validateEndpoint,
-  validateModel,
-  buildEndpointOption,
-  setHeaders,
-  async (req, res, next) => {
-    await AskController(req, res, next, initializeClient, addTitle);
-  },
-);
-
-module.exports = router;

api/server/routes/ask/custom.js

@@ -1,25 +0,0 @@
-const express = require('express');
-const AskController = require('~/server/controllers/AskController');
-const { initializeClient } = require('~/server/services/Endpoints/custom');
-const { addTitle } = require('~/server/services/Endpoints/openAI');
-const {
-  setHeaders,
-  validateModel,
-  validateEndpoint,
-  buildEndpointOption,
-} = require('~/server/middleware');
-
-const router = express.Router();
-
-router.post(
-  '/',
-  validateEndpoint,
-  validateModel,
-  buildEndpointOption,
-  setHeaders,
-  async (req, res, next) => {
-    await AskController(req, res, next, initializeClient, addTitle);
-  },
-);
-
-module.exports = router;

api/server/routes/ask/google.js

@@ -1,24 +0,0 @@
-const express = require('express');
-const AskController = require('~/server/controllers/AskController');
-const { initializeClient, addTitle } = require('~/server/services/Endpoints/google');
-const {
-  setHeaders,
-  validateModel,
-  validateEndpoint,
-  buildEndpointOption,
-} = require('~/server/middleware');
-
-const router = express.Router();
-
-router.post(
-  '/',
-  validateEndpoint,
-  validateModel,
-  buildEndpointOption,
-  setHeaders,
-  async (req, res, next) => {
-    await AskController(req, res, next, initializeClient, addTitle);
-  },
-);
-
-module.exports = router;

api/server/routes/ask/gptPlugins.js

@@ -1,241 +0,0 @@
-const express = require('express');
-const { getResponseSender, Constants } = require('librechat-data-provider');
-const { initializeClient } = require('~/server/services/Endpoints/gptPlugins');
-const { sendMessage, createOnProgress } = require('~/server/utils');
-const { addTitle } = require('~/server/services/Endpoints/openAI');
-const { saveMessage, updateMessage } = require('~/models');
-const {
-  handleAbort,
-  createAbortController,
-  handleAbortError,
-  setHeaders,
-  validateModel,
-  validateEndpoint,
-  buildEndpointOption,
-  moderateText,
-} = require('~/server/middleware');
-const { validateTools } = require('~/app');
-const { logger } = require('~/config');
-
-const router = express.Router();
-
-router.use(moderateText);
-
-router.post(
-  '/',
-  validateEndpoint,
-  validateModel,
-  buildEndpointOption,
-  setHeaders,
-  async (req, res) => {
-    let {
-      text,
-      endpointOption,
-      conversationId,
-      parentMessageId = null,
-      overrideParentMessageId = null,
-    } = req.body;
-
-    logger.debug('[/ask/gptPlugins]', { text, conversationId, ...endpointOption });
-
-    let userMessage;
-    let userMessagePromise;
-    let promptTokens;
-    let userMessageId;
-    let responseMessageId;
-    const sender = getResponseSender({
-      ...endpointOption,
-      model: endpointOption.modelOptions.model,
-    });
-    const newConvo = !conversationId;
-    const user = req.user.id;
-
-    const plugins = [];
-
-    const getReqData = (data = {}) => {
-      for (let key in data) {
-        if (key === 'userMessage') {
-          userMessage = data[key];
-          userMessageId = data[key].messageId;
-        } else if (key === 'userMessagePromise') {
-          userMessagePromise = data[key];
-        } else if (key === 'responseMessageId') {
-          responseMessageId = data[key];
-        } else if (key === 'promptTokens') {
-          promptTokens = data[key];
-        } else if (!conversationId && key === 'conversationId') {
-          conversationId = data[key];
-        }
-      }
-    };
-
-    let streaming = null;
-    let timer = null;
-
-    const {
-      onProgress: progressCallback,
-      sendIntermediateMessage,
-      getPartialText,
-    } = createOnProgress({
-      onProgress: () => {
-        if (timer) {
-          clearTimeout(timer);
-        }
-
-        streaming = new Promise((resolve) => {
-          timer = setTimeout(() => {
-            resolve();
-          }, 250);
-        });
-      },
-    });
-
-    const pluginMap = new Map();
-    const onAgentAction = async (action, runId) => {
-      pluginMap.set(runId, action.tool);
-      sendIntermediateMessage(res, {
-        plugins,
-        parentMessageId: userMessage.messageId,
-        messageId: responseMessageId,
-      });
-    };
-
-    const onToolStart = async (tool, input, runId, parentRunId) => {
-      const pluginName = pluginMap.get(parentRunId);
-      const latestPlugin = {
-        runId,
-        loading: true,
-        inputs: [input],
-        latest: pluginName,
-        outputs: null,
-      };
-
-      if (streaming) {
-        await streaming;
-      }
-      const extraTokens = ':::plugin:::\n';
-      plugins.push(latestPlugin);
-      sendIntermediateMessage(
-        res,
-        { plugins, parentMessageId: userMessage.messageId, messageId: responseMessageId },
-        extraTokens,
-      );
-    };
-
-    const onToolEnd = async (output, runId) => {
-      if (streaming) {
-        await streaming;
-      }
-
-      const pluginIndex = plugins.findIndex((plugin) => plugin.runId === runId);
-
-      if (pluginIndex !== -1) {
-        plugins[pluginIndex].loading = false;
-        plugins[pluginIndex].outputs = output;
-      }
-    };
-
-    const getAbortData = () => ({
-      sender,
-      conversationId,
-      userMessagePromise,
-      messageId: responseMessageId,
-      parentMessageId: overrideParentMessageId ?? userMessageId,
-      text: getPartialText(),
-      plugins: plugins.map((p) => ({ ...p, loading: false })),
-      userMessage,
-      promptTokens,
-    });
-    const { abortController, onStart } = createAbortController(req, res, getAbortData, getReqData);
-
-    try {
-      endpointOption.tools = await validateTools(user, endpointOption.tools);
-      const { client } = await initializeClient({ req, res, endpointOption });
-
-      const onChainEnd = () => {
-        if (!client.skipSaveUserMessage) {
-          saveMessage(
-            req,
-            { ...userMessage, user },
-            { context: 'api/server/routes/ask/gptPlugins.js - onChainEnd' },
-          );
-        }
-        sendIntermediateMessage(res, {
-          plugins,
-          parentMessageId: userMessage.messageId,
-          messageId: responseMessageId,
-        });
-      };
-
-      let response = await client.sendMessage(text, {
-        user,
-        conversationId,
-        parentMessageId,
-        overrideParentMessageId,
-        getReqData,
-        onAgentAction,
-        onChainEnd,
-        onToolStart,
-        onToolEnd,
-        onStart,
-        getPartialText,
-        ...endpointOption,
-        progressCallback,
-        progressOptions: {
-          res,
-          // parentMessageId: overrideParentMessageId || userMessageId,
-          plugins,
-        },
-        abortController,
-      });
-
-      if (overrideParentMessageId) {
-        response.parentMessageId = overrideParentMessageId;
-      }
-
-      logger.debug('[/ask/gptPlugins]', response);
-
-      const { conversation = {} } = await response.databasePromise;
-      delete response.databasePromise;
-      conversation.title =
-        conversation && !conversation.title ? null : conversation?.title || 'New Chat';
-
-      sendMessage(res, {
-        title: conversation.title,
-        final: true,
-        conversation,
-        requestMessage: userMessage,
-        responseMessage: response,
-      });
-      res.end();
-
-      if (parentMessageId === Constants.NO_PARENT && newConvo) {
-        addTitle(req, {
-          text,
-          response,
-          client,
-        });
-      }
-
-      response.plugins = plugins.map((p) => ({ ...p, loading: false }));
-      if (response.plugins?.length > 0) {
-        await updateMessage(
-          req,
-          { ...response, user },
-          { context: 'api/server/routes/ask/gptPlugins.js - save plugins used' },
-        );
-      }
-    } catch (error) {
-      const partialText = getPartialText();
-      handleAbortError(res, req, error, {
-        partialText,
-        conversationId,
-        sender,
-        messageId: responseMessageId,
-        parentMessageId: userMessageId ?? parentMessageId,
-      });
-    }
-  },
-);
-
-module.exports = router;

api/server/routes/ask/index.js

@@ -1,47 +0,0 @@
-const express = require('express');
-const { EModelEndpoint } = require('librechat-data-provider');
-const {
-  uaParser,
-  checkBan,
-  requireJwtAuth,
-  messageIpLimiter,
-  concurrentLimiter,
-  messageUserLimiter,
-  validateConvoAccess,
-} = require('~/server/middleware');
-const { isEnabled } = require('~/server/utils');
-const gptPlugins = require('./gptPlugins');
-const anthropic = require('./anthropic');
-const custom = require('./custom');
-const google = require('./google');
-const openAI = require('./openAI');
-
-const { LIMIT_CONCURRENT_MESSAGES, LIMIT_MESSAGE_IP, LIMIT_MESSAGE_USER } = process.env ?? {};
-
-const router = express.Router();
-
-router.use(requireJwtAuth);
-router.use(checkBan);
-router.use(uaParser);
-
-if (isEnabled(LIMIT_CONCURRENT_MESSAGES)) {
-  router.use(concurrentLimiter);
-}
-
-if (isEnabled(LIMIT_MESSAGE_IP)) {
-  router.use(messageIpLimiter);
-}
-
-if (isEnabled(LIMIT_MESSAGE_USER)) {
-  router.use(messageUserLimiter);
-}
-
-router.use(validateConvoAccess);
-
-router.use([`/${EModelEndpoint.azureOpenAI}`, `/${EModelEndpoint.openAI}`], openAI);
-router.use(`/${EModelEndpoint.gptPlugins}`, gptPlugins);
-router.use(`/${EModelEndpoint.anthropic}`, anthropic);
-router.use(`/${EModelEndpoint.google}`, google);
-router.use(`/${EModelEndpoint.custom}`, custom);
-
-module.exports = router;

api/server/routes/ask/openAI.js

@@ -1,27 +0,0 @@
-const express = require('express');
-const AskController = require('~/server/controllers/AskController');
-const { addTitle, initializeClient } = require('~/server/services/Endpoints/openAI');
-const {
-  handleAbort,
-  setHeaders,
-  validateModel,
-  validateEndpoint,
-  buildEndpointOption,
-  moderateText,
-} = require('~/server/middleware');
-
-const router = express.Router();
-router.use(moderateText);
-
-router.post(
-  '/',
-  validateEndpoint,
-  validateModel,
-  buildEndpointOption,
-  setHeaders,
-  async (req, res, next) => {
-    await AskController(req, res, next, initializeClient, addTitle);
-  },
-);
-
-module.exports = router;

api/server/routes/bedrock/chat.js

@@ -1,37 +0,0 @@
-const express = require('express');
-
-const router = express.Router();
-const {
-  setHeaders,
-  handleAbort,
-  moderateText,
-  // validateModel,
-  // validateEndpoint,
-  buildEndpointOption,
-} = require('~/server/middleware');
-const { initializeClient } = require('~/server/services/Endpoints/bedrock');
-const AgentController = require('~/server/controllers/agents/request');
-const addTitle = require('~/server/services/Endpoints/agents/title');
-
-router.use(moderateText);
-
-/**
- * @route POST /
- * @desc Chat with an assistant
- * @access Public
- * @param {express.Request} req - The request object, containing the request data.
- * @param {express.Response} res - The response object, used to send back a response.
- * @returns {void}
- */
-router.post(
-  '/',
-  // validateModel,
-  // validateEndpoint,
-  buildEndpointOption,
-  setHeaders,
-  async (req, res, next) => {
-    await AgentController(req, res, next, initializeClient, addTitle);
-  },
-);
-
-module.exports = router;

api/server/routes/bedrock/index.js

@@ -1,35 +0,0 @@
-const express = require('express');
-const {
-  uaParser,
-  checkBan,
-  requireJwtAuth,
-  messageIpLimiter,
-  concurrentLimiter,
-  messageUserLimiter,
-} = require('~/server/middleware');
-const { isEnabled } = require('~/server/utils');
-const chat = require('./chat');
-
-const { LIMIT_CONCURRENT_MESSAGES, LIMIT_MESSAGE_IP, LIMIT_MESSAGE_USER } = process.env ?? {};
-
-const router = express.Router();
-
-router.use(requireJwtAuth);
-router.use(checkBan);
-router.use(uaParser);
-
-if (isEnabled(LIMIT_CONCURRENT_MESSAGES)) {
-  router.use(concurrentLimiter);
-}
-
-if (isEnabled(LIMIT_MESSAGE_IP)) {
-  router.use(messageIpLimiter);
-}
-
-if (isEnabled(LIMIT_MESSAGE_USER)) {
-  router.use(messageUserLimiter);
-}
-
-router.use('/chat', chat);
-
-module.exports = router;

api/server/routes/edit/gptPlugins.js

@@ -1,207 +0,0 @@
-const express = require('express');
-const { getResponseSender } = require('librechat-data-provider');
-const {
-  setHeaders,
-  moderateText,
-  validateModel,
-  handleAbortError,
-  validateEndpoint,
-  buildEndpointOption,
-  createAbortController,
-} = require('~/server/middleware');
-const { sendMessage, createOnProgress, formatSteps, formatAction } = require('~/server/utils');
-const { initializeClient } = require('~/server/services/Endpoints/gptPlugins');
-const { saveMessage, updateMessage } = require('~/models');
-const { validateTools } = require('~/app');
-const { logger } = require('~/config');
-
-const router = express.Router();
-
-router.use(moderateText);
-
-router.post(
-  '/',
-  validateEndpoint,
-  validateModel,
-  buildEndpointOption,
-  setHeaders,
-  async (req, res) => {
-    let {
-      text,
-      generation,
-      endpointOption,
-      conversationId,
-      responseMessageId,
-      isContinued = false,
-      parentMessageId = null,
-      overrideParentMessageId = null,
-    } = req.body;
-
-    logger.debug('[/edit/gptPlugins]', {
-      text,
-      generation,
-      isContinued,
-      conversationId,
-      ...endpointOption,
-    });
-
-    let userMessage;
-    let userMessagePromise;
-    let promptTokens;
-    const sender = getResponseSender({
-      ...endpointOption,
-      model: endpointOption.modelOptions.model,
-    });
-    const userMessageId = parentMessageId;
-    const user = req.user.id;
-
-    const plugin = {
-      loading: true,
-      inputs: [],
-      latest: null,
-      outputs: null,
-    };
-
-    const getReqData = (data = {}) => {
-      for (let key in data) {
-        if (key === 'userMessage') {
-          userMessage = data[key];
-        } else if (key === 'userMessagePromise') {
-          userMessagePromise = data[key];
-        } else if (key === 'responseMessageId') {
-          responseMessageId = data[key];
-        } else if (key === 'promptTokens') {
-          promptTokens = data[key];
-        }
-      }
-    };
-
-    const {
-      onProgress: progressCallback,
-      sendIntermediateMessage,
-      getPartialText,
-    } = createOnProgress({
-      generation,
-      onProgress: () => {
-        if (plugin.loading === true) {
-          plugin.loading = false;
-        }
-      },
-    });
-
-    const onChainEnd = (data) => {
-      let { intermediateSteps: steps } = data;
-      plugin.outputs = steps && steps[0].action ? formatSteps(steps) : 'An error occurred.';
-      plugin.loading = false;
-      saveMessage(
-        req,
-        { ...userMessage, user },
-        { context: 'api/server/routes/ask/gptPlugins.js - onChainEnd' },
-      );
-      sendIntermediateMessage(res, {
-        plugin,
-        parentMessageId: userMessage.messageId,
-        messageId: responseMessageId,
-      });
-      // logger.debug('CHAIN END', plugin.outputs);
-    };
-
-    const getAbortData = () => ({
-      sender,
-      conversationId,
-      userMessagePromise,
-      messageId: responseMessageId,
-      parentMessageId: overrideParentMessageId ?? userMessageId,
-      text: getPartialText(),
-      plugin: { ...plugin, loading: false },
-      userMessage,
-      promptTokens,
-    });
-    const { abortController, onStart } = createAbortController(req, res, getAbortData, getReqData);
-
-    try {
-      endpointOption.tools = await validateTools(user, endpointOption.tools);
-      const { client } = await initializeClient({ req, res, endpointOption });
-
-      const onAgentAction = (action, start = false) => {
-        const formattedAction = formatAction(action);
-        plugin.inputs.push(formattedAction);
-        plugin.latest = formattedAction.plugin;
-        if (!start && !client.skipSaveUserMessage) {
-          saveMessage(
-            req,
-            { ...userMessage, user },
-            { context: 'api/server/routes/ask/gptPlugins.js - onAgentAction' },
-          );
-        }
-        sendIntermediateMessage(res, {
-          plugin,
-          parentMessageId: userMessage.messageId,
-          messageId: responseMessageId,
-        });
-        // logger.debug('PLUGIN ACTION', formattedAction);
-      };
-
-      let response = await client.sendMessage(text, {
-        user,
-        generation,
-        isContinued,
-        isEdited: true,
-        conversationId,
-        parentMessageId,
-        responseMessageId,
-        overrideParentMessageId,
-        getReqData,
-        onAgentAction,
-        onChainEnd,
-        onStart,
-        ...endpointOption,
-        progressCallback,
-        progressOptions: {
-          res,
-          plugin,
-          // parentMessageId: overrideParentMessageId || userMessageId,
-        },
-        abortController,
-      });
-
-      if (overrideParentMessageId) {
-        response.parentMessageId = overrideParentMessageId;
-      }
-
-      logger.debug('[/edit/gptPlugins] CLIENT RESPONSE', response);
-
-      const { conversation = {} } = await response.databasePromise;
-      delete response.databasePromise;
-      conversation.title =
-        conversation && !conversation.title ? null : conversation?.title || 'New Chat';
-
-      sendMessage(res, {
-        title: conversation.title,
-        final: true,
-        conversation,
-        requestMessage: userMessage,
-        responseMessage: response,
-      });
-      res.end();
-
-      response.plugin = { ...plugin, loading: false };
-      await updateMessage(
-        req,
-        { ...response, user },
-        { context: 'api/server/routes/edit/gptPlugins.js' },
-      );
-    } catch (error) {
-      const partialText = getPartialText();
-      handleAbortError(res, req, error, {
-        partialText,
-        conversationId,
-        sender,
-        messageId: responseMessageId,
-        parentMessageId: userMessageId ?? parentMessageId,
-      });
-    }
-  },
-);
-
-module.exports = router;

api/server/routes/edit/index.js

@@ -3,7 +3,6 @@ const openAI = require('./openAI');
 const custom = require('./custom');
 const google = require('./google');
 const anthropic = require('./anthropic');
-const gptPlugins = require('./gptPlugins');
 const { isEnabled } = require('~/server/utils');
 const { EModelEndpoint } = require('librechat-data-provider');
 const {
@@ -39,7 +38,6 @@ if (isEnabled(LIMIT_MESSAGE_USER)) {
 router.use(validateConvoAccess);
 
 router.use([`/${EModelEndpoint.azureOpenAI}`, `/${EModelEndpoint.openAI}`], openAI);
-router.use(`/${EModelEndpoint.gptPlugins}`, gptPlugins);
 router.use(`/${EModelEndpoint.anthropic}`, anthropic);
 router.use(`/${EModelEndpoint.google}`, google);
 router.use(`/${EModelEndpoint.custom}`, custom);

api/server/routes/files/files.js

@@ -283,7 +283,10 @@ router.post('/', async (req, res) => {
       message += ': ' + error.message;
     }
 
-    if (error.message?.includes('Invalid file format')) {
+    if (
+      error.message?.includes('Invalid file format') ||
+      error.message?.includes('No OCR result')
+    ) {
       message = error.message;
     }
 

api/server/routes/files/multer.spec.js

@@ -477,7 +477,9 @@ describe('Multer Configuration', () => {
         done(new Error('Expected mkdirSync to throw an error but no error was thrown'));
       } catch (error) {
         // This is the expected behavior - mkdirSync throws synchronously for invalid paths
-        expect(error.code).toBe('EACCES');
+        // On Linux, this typically returns EACCES (permission denied)
+        // On macOS/Darwin, this returns ENOENT (no such file or directory)
+        expect(['EACCES', 'ENOENT']).toContain(error.code);
         done();
       }
     });

api/server/routes/index.js

@@ -9,7 +9,6 @@ const presets = require('./presets');
 const prompts = require('./prompts');
 const balance = require('./balance');
 const plugins = require('./plugins');
-const bedrock = require('./bedrock');
 const actions = require('./actions');
 const banner = require('./banner');
 const search = require('./search');
@@ -26,11 +25,9 @@ const auth = require('./auth');
 const edit = require('./edit');
 const keys = require('./keys');
 const user = require('./user');
-const ask = require('./ask');
 const mcp = require('./mcp');
 
 module.exports = {
-  ask,
   edit,
   auth,
   keys,
@@ -46,7 +43,6 @@ module.exports = {
   search,
   config,
   models,
-  bedrock,
   prompts,
   plugins,
   actions,

api/server/routes/memories.js

@@ -1,37 +1,43 @@
 const express = require('express');
-const { Tokenizer } = require('@librechat/api');
+const { Tokenizer, generateCheckAccess } = require('@librechat/api');
 const { PermissionTypes, Permissions } = require('librechat-data-provider');
 const {
   getAllUserMemories,
   toggleUserMemories,
   createMemory,
-  setMemory,
   deleteMemory,
+  setMemory,
 } = require('~/models');
-const { requireJwtAuth, generateCheckAccess } = require('~/server/middleware');
+const { requireJwtAuth } = require('~/server/middleware');
+const { getRoleByName } = require('~/models/Role');
 
 const router = express.Router();
 
-const checkMemoryRead = generateCheckAccess(PermissionTypes.MEMORIES, [
-  Permissions.USE,
-  Permissions.READ,
-]);
-const checkMemoryCreate = generateCheckAccess(PermissionTypes.MEMORIES, [
-  Permissions.USE,
-  Permissions.CREATE,
-]);
-const checkMemoryUpdate = generateCheckAccess(PermissionTypes.MEMORIES, [
-  Permissions.USE,
-  Permissions.UPDATE,
-]);
-const checkMemoryDelete = generateCheckAccess(PermissionTypes.MEMORIES, [
-  Permissions.USE,
-  Permissions.UPDATE,
-]);
-const checkMemoryOptOut = generateCheckAccess(PermissionTypes.MEMORIES, [
-  Permissions.USE,
-  Permissions.OPT_OUT,
-]);
+const checkMemoryRead = generateCheckAccess({
+  permissionType: PermissionTypes.MEMORIES,
+  permissions: [Permissions.USE, Permissions.READ],
+  getRoleByName,
+});
+const checkMemoryCreate = generateCheckAccess({
+  permissionType: PermissionTypes.MEMORIES,
+  permissions: [Permissions.USE, Permissions.CREATE],
+  getRoleByName,
+});
+const checkMemoryUpdate = generateCheckAccess({
+  permissionType: PermissionTypes.MEMORIES,
+  permissions: [Permissions.USE, Permissions.UPDATE],
+  getRoleByName,
+});
+const checkMemoryDelete = generateCheckAccess({
+  permissionType: PermissionTypes.MEMORIES,
+  permissions: [Permissions.USE, Permissions.UPDATE],
+  getRoleByName,
+});
+const checkMemoryOptOut = generateCheckAccess({
+  permissionType: PermissionTypes.MEMORIES,
+  permissions: [Permissions.USE, Permissions.OPT_OUT],
+  getRoleByName,
+});
 
 router.use(requireJwtAuth);
 

api/server/routes/messages.js

@@ -235,12 +235,13 @@ router.put('/:conversationId/:messageId', validateMessageReq, async (req, res) =
       return res.status(400).json({ error: 'Content part not found' });
     }
 
-    if (updatedContent[index].type !== ContentTypes.TEXT) {
+    const currentPartType = updatedContent[index].type;
+    if (currentPartType !== ContentTypes.TEXT && currentPartType !== ContentTypes.THINK) {
       return res.status(400).json({ error: 'Cannot update non-text content' });
     }
 
-    const oldText = updatedContent[index].text;
-    updatedContent[index] = { type: ContentTypes.TEXT, text };
+    const oldText = updatedContent[index][currentPartType];
+    updatedContent[index] = { type: currentPartType, [currentPartType]: text };
 
     let tokenCount = message.tokenCount;
     if (tokenCount !== undefined) {

api/server/routes/prompts.js

@@ -1,5 +1,7 @@
 const express = require('express');
-const { PermissionTypes, Permissions, SystemRoles } = require('librechat-data-provider');
+const { logger } = require('@librechat/data-schemas');
+const { generateCheckAccess } = require('@librechat/api');
+const { Permissions, SystemRoles, PermissionTypes } = require('librechat-data-provider');
 const {
   getPrompt,
   getPrompts,
@@ -14,24 +16,30 @@ const {
   // updatePromptLabels,
   makePromptProduction,
 } = require('~/models/Prompt');
-const { requireJwtAuth, generateCheckAccess } = require('~/server/middleware');
-const { logger } = require('~/config');
+const { requireJwtAuth } = require('~/server/middleware');
+const { getRoleByName } = require('~/models/Role');
 
 const router = express.Router();
 
-const checkPromptAccess = generateCheckAccess(PermissionTypes.PROMPTS, [Permissions.USE]);
-const checkPromptCreate = generateCheckAccess(PermissionTypes.PROMPTS, [
-  Permissions.USE,
-  Permissions.CREATE,
-]);
+const checkPromptAccess = generateCheckAccess({
+  permissionType: PermissionTypes.PROMPTS,
+  permissions: [Permissions.USE],
+  getRoleByName,
+});
+const checkPromptCreate = generateCheckAccess({
+  permissionType: PermissionTypes.PROMPTS,
+  permissions: [Permissions.USE, Permissions.CREATE],
+  getRoleByName,
+});
 
-const checkGlobalPromptShare = generateCheckAccess(
-  PermissionTypes.PROMPTS,
-  [Permissions.USE, Permissions.CREATE],
-  {
+const checkGlobalPromptShare = generateCheckAccess({
+  permissionType: PermissionTypes.PROMPTS,
+  permissions: [Permissions.USE, Permissions.CREATE],
+  bodyProps: {
     [Permissions.SHARED_GLOBAL]: ['projectIds', 'removeProjectIds'],
   },
-);
+  getRoleByName,
+});
 
 router.use(requireJwtAuth);
 router.use(checkPromptAccess);

api/server/routes/tags.js

@@ -1,18 +1,24 @@
 const express = require('express');
+const { logger } = require('@librechat/data-schemas');
+const { generateCheckAccess } = require('@librechat/api');
 const { PermissionTypes, Permissions } = require('librechat-data-provider');
 const {
-  getConversationTags,
+  updateTagsForConversation,
   updateConversationTag,
   createConversationTag,
   deleteConversationTag,
-  updateTagsForConversation,
+  getConversationTags,
 } = require('~/models/ConversationTag');
-const { requireJwtAuth, generateCheckAccess } = require('~/server/middleware');
-const { logger } = require('~/config');
+const { requireJwtAuth } = require('~/server/middleware');
+const { getRoleByName } = require('~/models/Role');
 
 const router = express.Router();
 
-const checkBookmarkAccess = generateCheckAccess(PermissionTypes.BOOKMARKS, [Permissions.USE]);
+const checkBookmarkAccess = generateCheckAccess({
+  permissionType: PermissionTypes.BOOKMARKS,
+  permissions: [Permissions.USE],
+  getRoleByName,
+});
 
 router.use(requireJwtAuth);
 router.use(checkBookmarkAccess);

api/server/services/AssistantService.js

@@ -1,4 +1,7 @@
 const { klona } = require('klona');
+const { sleep } = require('@librechat/agents');
+const { sendEvent } = require('@librechat/api');
+const { logger } = require('@librechat/data-schemas');
 const {
   StepTypes,
   RunStatus,
@@ -11,11 +14,10 @@ const {
 } = require('librechat-data-provider');
 const { retrieveAndProcessFile } = require('~/server/services/Files/process');
 const { processRequiredActions } = require('~/server/services/ToolService');
-const { createOnProgress, sendMessage, sleep } = require('~/server/utils');
 const { RunManager, waitForRun } = require('~/server/services/Runs');
 const { processMessages } = require('~/server/services/Threads');
+const { createOnProgress } = require('~/server/utils');
 const { TextStream } = require('~/app/clients');
-const { logger } = require('~/config');
 
 /**
  * Sorts, processes, and flattens messages to a single string.
@@ -64,7 +66,7 @@ async function createOnTextProgress({
     };
 
     logger.debug('Content data:', contentData);
-    sendMessage(openai.res, contentData);
+    sendEvent(openai.res, contentData);
   };
 }
 

api/server/services/AuthService.js

@@ -1,4 +1,5 @@
 const bcrypt = require('bcryptjs');
+const jwt = require('jsonwebtoken');
 const { webcrypto } = require('node:crypto');
 const { isEnabled } = require('@librechat/api');
 const { logger } = require('@librechat/data-schemas');
@@ -499,6 +500,18 @@ const resendVerificationEmail = async (req) => {
     };
   }
 };
+/**
+ * Generate a short-lived JWT token
+ * @param {String} userId - The ID of the user
+ * @param {String} [expireIn='5m'] - The expiration time for the token (default is 5 minutes)
+ * @returns {String} - The generated JWT token
+ */
+const generateShortLivedToken = (userId, expireIn = '5m') => {
+  return jwt.sign({ id: userId }, process.env.JWT_SECRET, {
+    expiresIn: expireIn,
+    algorithm: 'HS256',
+  });
+};
 
 module.exports = {
   logoutUser,
@@ -506,7 +519,8 @@ module.exports = {
   registerUser,
   setAuthTokens,
   resetPassword,
+  setOpenIDAuthTokens,
   requestPasswordReset,
   resendVerificationEmail,
-  setOpenIDAuthTokens,
+  generateShortLivedToken,
 };

api/server/services/Config/EndpointService.js

@@ -1,5 +1,6 @@
+const { isUserProvided } = require('@librechat/api');
 const { EModelEndpoint } = require('librechat-data-provider');
-const { isUserProvided, generateConfig } = require('~/server/utils');
+const { generateConfig } = require('~/server/utils/handleText');
 
 const {
   OPENAI_API_KEY: openAIApiKey,

api/server/services/Config/getCustomConfig.js

@@ -40,6 +40,7 @@ async function getBalanceConfig() {
 /**
  *
  * @param {string | EModelEndpoint} endpoint
+ * @returns {Promise<TEndpoint | undefined>}
  */
 const getCustomEndpointConfig = async (endpoint) => {
   const customConfig = await getCustomConfig();

api/server/services/Config/loadAsyncEndpoints.js

@@ -1,3 +1,5 @@
+const path = require('path');
+const { loadServiceKey } = require('@librechat/api');
 const { EModelEndpoint } = require('librechat-data-provider');
 const { isUserProvided } = require('~/server/utils');
 const { config } = require('./EndpointService');
@@ -11,9 +13,13 @@ const { openAIApiKey, azureOpenAIApiKey, useAzurePlugins, userProvidedOpenAI, go
 async function loadAsyncEndpoints(req) {
   let i = 0;
   let serviceKey, googleUserProvides;
+  const serviceKeyPath =
+    process.env.GOOGLE_SERVICE_KEY_FILE_PATH ||
+    path.join(__dirname, '../../..', 'data', 'auth.json');
+
   try {
-    serviceKey = require('~/data/auth.json');
-  } catch (e) {
+    serviceKey = await loadServiceKey(serviceKeyPath);
+  } catch {
     if (i === 0) {
       i++;
     }
@@ -32,14 +38,14 @@ async function loadAsyncEndpoints(req) {
   const gptPlugins =
     useAzure || openAIApiKey || azureOpenAIApiKey
       ? {
-        availableAgents: ['classic', 'functions'],
-        userProvide: useAzure ? false : userProvidedOpenAI,
-        userProvideURL: useAzure
-          ? false
-          : config[EModelEndpoint.openAI]?.userProvideURL ||
+          availableAgents: ['classic', 'functions'],
+          userProvide: useAzure ? false : userProvidedOpenAI,
+          userProvideURL: useAzure
+            ? false
+            : config[EModelEndpoint.openAI]?.userProvideURL ||
               config[EModelEndpoint.azureOpenAI]?.userProvideURL,
-        azure: useAzurePlugins || useAzure,
-      }
+          azure: useAzurePlugins || useAzure,
+        }
       : false;
 
   return { google, gptPlugins };

api/server/services/Config/loadCustomConfig.js

@@ -1,18 +1,18 @@
 const path = require('path');
-const {
-  CacheKeys,
-  configSchema,
-  EImageOutputType,
-  validateSettingDefinitions,
-  agentParamSettings,
-  paramSettings,
-} = require('librechat-data-provider');
-const getLogStores = require('~/cache/getLogStores');
-const loadYaml = require('~/utils/loadYaml');
-const { logger } = require('~/config');
 const axios = require('axios');
 const yaml = require('js-yaml');
 const keyBy = require('lodash/keyBy');
+const { loadYaml } = require('@librechat/api');
+const { logger } = require('@librechat/data-schemas');
+const {
+  CacheKeys,
+  configSchema,
+  paramSettings,
+  EImageOutputType,
+  agentParamSettings,
+  validateSettingDefinitions,
+} = require('librechat-data-provider');
+const getLogStores = require('~/cache/getLogStores');
 
 const projectRoot = path.resolve(__dirname, '..', '..', '..', '..');
 const defaultConfigPath = path.resolve(projectRoot, 'librechat.yaml');

api/server/services/Config/loadCustomConfig.spec.js

@@ -1,6 +1,9 @@
 jest.mock('axios');
 jest.mock('~/cache/getLogStores');
-jest.mock('~/utils/loadYaml');
+jest.mock('@librechat/api', () => ({
+  ...jest.requireActual('@librechat/api'),
+  loadYaml: jest.fn(),
+}));
 jest.mock('librechat-data-provider', () => {
   const actual = jest.requireActual('librechat-data-provider');
   return {
@@ -30,11 +33,22 @@ jest.mock('librechat-data-provider', () => {
   };
 });
 
+jest.mock('@librechat/data-schemas', () => {
+  return {
+    logger: {
+      info: jest.fn(),
+      warn: jest.fn(),
+      debug: jest.fn(),
+      error: jest.fn(),
+    },
+  };
+});
+
 const axios = require('axios');
+const { loadYaml } = require('@librechat/api');
+const { logger } = require('@librechat/data-schemas');
 const loadCustomConfig = require('./loadCustomConfig');
 const getLogStores = require('~/cache/getLogStores');
-const loadYaml = require('~/utils/loadYaml');
-const { logger } = require('~/config');
 
 describe('loadCustomConfig', () => {
   const mockSet = jest.fn();

api/server/services/Endpoints/agents/agent.js

@@ -1,5 +1,9 @@
 const { Providers } = require('@librechat/agents');
-const { primeResources, optionalChainWithEmptyCheck } = require('@librechat/api');
+const {
+  primeResources,
+  extractLibreChatParams,
+  optionalChainWithEmptyCheck,
+} = require('@librechat/api');
 const {
   ErrorTypes,
   EModelEndpoint,
@@ -7,30 +11,12 @@ const {
   replaceSpecialVars,
   providerEndpointMap,
 } = require('librechat-data-provider');
-const initAnthropic = require('~/server/services/Endpoints/anthropic/initialize');
-const getBedrockOptions = require('~/server/services/Endpoints/bedrock/options');
-const initOpenAI = require('~/server/services/Endpoints/openAI/initialize');
-const initCustom = require('~/server/services/Endpoints/custom/initialize');
-const initGoogle = require('~/server/services/Endpoints/google/initialize');
+const { getProviderConfig } = require('~/server/services/Endpoints');
 const generateArtifactsPrompt = require('~/app/clients/prompts/artifacts');
-const { getCustomEndpointConfig } = require('~/server/services/Config');
 const { processFiles } = require('~/server/services/Files/process');
+const { getFiles, getToolFilesByIds } = require('~/models/File');
 const { getConvoFiles } = require('~/models/Conversation');
-const { getToolFilesByIds } = require('~/models/File');
 const { getModelMaxTokens } = require('~/utils');
-const { getFiles } = require('~/models/File');
-
-const providerConfigMap = {
-  [Providers.XAI]: initCustom,
-  [Providers.OLLAMA]: initCustom,
-  [Providers.DEEPSEEK]: initCustom,
-  [Providers.OPENROUTER]: initCustom,
-  [EModelEndpoint.openAI]: initOpenAI,
-  [EModelEndpoint.google]: initGoogle,
-  [EModelEndpoint.azureOpenAI]: initOpenAI,
-  [EModelEndpoint.anthropic]: initAnthropic,
-  [EModelEndpoint.bedrock]: getBedrockOptions,
-};
 
 /**
  * @param {object} params
@@ -71,7 +57,7 @@ const initializeAgent = async ({
     ),
   );
 
-  const { resendFiles = true, ...modelOptions } = _modelOptions;
+  const { resendFiles, maxContextTokens, modelOptions } = extractLibreChatParams(_modelOptions);
 
   if (isInitialAgent && conversationId != null && resendFiles) {
     const fileIds = (await getConvoFiles(conversationId)) ?? [];
@@ -99,7 +85,7 @@ const initializeAgent = async ({
   });
 
   const provider = agent.provider;
-  const { tools, toolContextMap } =
+  const { tools: structuredTools, toolContextMap } =
     (await loadTools?.({
       req,
       res,
@@ -111,17 +97,9 @@ const initializeAgent = async ({
     })) ?? {};
 
   agent.endpoint = provider;
-  let getOptions = providerConfigMap[provider];
-  if (!getOptions && providerConfigMap[provider.toLowerCase()] != null) {
-    agent.provider = provider.toLowerCase();
-    getOptions = providerConfigMap[agent.provider];
-  } else if (!getOptions) {
-    const customEndpointConfig = await getCustomEndpointConfig(provider);
-    if (!customEndpointConfig) {
-      throw new Error(`Provider ${provider} not supported`);
-    }
-    getOptions = initCustom;
-    agent.provider = Providers.OPENAI;
+  const { getOptions, overrideProvider } = await getProviderConfig(provider);
+  if (overrideProvider) {
+    agent.provider = overrideProvider;
   }
 
   const _endpointOption =
@@ -145,9 +123,8 @@ const initializeAgent = async ({
     modelOptions.maxTokens,
     0,
   );
-  const maxContextTokens = optionalChainWithEmptyCheck(
-    modelOptions.maxContextTokens,
-    modelOptions.max_context_tokens,
+  const agentMaxContextTokens = optionalChainWithEmptyCheck(
+    maxContextTokens,
     getModelMaxTokens(tokensModel, providerEndpointMap[provider]),
     4096,
   );
@@ -163,6 +140,22 @@ const initializeAgent = async ({
     agent.provider = options.provider;
   }
 
+  /** @type {import('@librechat/agents').GenericTool[]} */
+  let tools = options.tools?.length ? options.tools : structuredTools;
+  if (
+    (agent.provider === Providers.GOOGLE || agent.provider === Providers.VERTEXAI) &&
+    options.tools?.length &&
+    structuredTools?.length
+  ) {
+    throw new Error(`{ "type": "${ErrorTypes.GOOGLE_TOOL_CONFLICT}"}`);
+  } else if (
+    (agent.provider === Providers.OPENAI || agent.provider === Providers.AZURE) &&
+    options.tools?.length &&
+    structuredTools?.length
+  ) {
+    tools = structuredTools.concat(options.tools);
+  }
+
   /** @type {import('@librechat/agents').ClientOptions} */
   agent.model_parameters = { ...options.llmConfig };
   if (options.configOptions) {
@@ -185,11 +178,11 @@ const initializeAgent = async ({
 
   return {
     ...agent,
-    tools,
     attachments,
     resendFiles,
     toolContextMap,
-    maxContextTokens: (maxContextTokens - maxTokens) * 0.9,
+    tools,
+    maxContextTokens: (agentMaxContextTokens - maxTokens) * 0.9,
   };
 };
 

api/server/services/Endpoints/agents/build.js

@@ -1,10 +1,9 @@
-const { isAgentsEndpoint, Constants } = require('librechat-data-provider');
+const { isAgentsEndpoint, removeNullishValues, Constants } = require('librechat-data-provider');
 const { loadAgent } = require('~/models/Agent');
 const { logger } = require('~/config');
 
 const buildOptions = (req, endpoint, parsedBody, endpointType) => {
-  const { spec, iconURL, agent_id, instructions, maxContextTokens, ...model_parameters } =
-    parsedBody;
+  const { spec, iconURL, agent_id, instructions, ...model_parameters } = parsedBody;
   const agentPromise = loadAgent({
     req,
     agent_id: isAgentsEndpoint(endpoint) ? agent_id : Constants.EPHEMERAL_AGENT_ID,
@@ -15,19 +14,16 @@ const buildOptions = (req, endpoint, parsedBody, endpointType) => {
     return undefined;
   });
 
-  const endpointOption = {
+  return removeNullishValues({
     spec,
     iconURL,
     endpoint,
     agent_id,
     endpointType,
     instructions,
-    maxContextTokens,
     model_parameters,
     agent: agentPromise,
-  };
-
-  return endpointOption;
+  });
 };
 
 module.exports = { buildOptions };

api/server/services/Endpoints/agents/initialize.js

@@ -1,11 +1,17 @@
 const { logger } = require('@librechat/data-schemas');
 const { createContentAggregator } = require('@librechat/agents');
-const { Constants, EModelEndpoint, getResponseSender } = require('librechat-data-provider');
 const {
-  getDefaultHandlers,
+  Constants,
+  EModelEndpoint,
+  isAgentsEndpoint,
+  getResponseSender,
+} = require('librechat-data-provider');
+const {
   createToolEndCallback,
+  getDefaultHandlers,
 } = require('~/server/controllers/agents/callbacks');
 const { initializeAgent } = require('~/server/services/Endpoints/agents/agent');
+const { getCustomEndpointConfig } = require('~/server/services/Config');
 const { loadAgentTools } = require('~/server/services/ToolService');
 const AgentClient = require('~/server/controllers/agents/client');
 const { getAgent } = require('~/models/Agent');
@@ -61,6 +67,7 @@ const initializeClient = async ({ req, res, endpointOption }) => {
   }
 
   const primaryAgent = await endpointOption.agent;
+  delete endpointOption.agent;
   if (!primaryAgent) {
     throw new Error('Agent not found');
   }
@@ -108,11 +115,25 @@ const initializeClient = async ({ req, res, endpointOption }) => {
     }
   }
 
+  let endpointConfig = req.app.locals[primaryConfig.endpoint];
+  if (!isAgentsEndpoint(primaryConfig.endpoint) && !endpointConfig) {
+    try {
+      endpointConfig = await getCustomEndpointConfig(primaryConfig.endpoint);
+    } catch (err) {
+      logger.error(
+        '[api/server/controllers/agents/client.js #titleConvo] Error getting custom endpoint config',
+        err,
+      );
+    }
+  }
+
   const sender =
     primaryAgent.name ??
     getResponseSender({
       ...endpointOption,
       model: endpointOption.model_parameters.model,
+      modelDisplayLabel: endpointConfig?.modelDisplayLabel,
+      modelLabel: endpointOption.model_parameters.modelLabel,
     });
 
   const client = new AgentClient({

api/server/services/Endpoints/agents/title.js

@@ -23,7 +23,7 @@ const addTitle = async (req, { text, response, client }) => {
   let timeoutId;
   try {
     const timeoutPromise = new Promise((_, reject) => {
-      timeoutId = setTimeout(() => reject(new Error('Title generation timeout')), 25000);
+      timeoutId = setTimeout(() => reject(new Error('Title generation timeout')), 45000);
     }).catch((error) => {
       logger.error('Title error:', error);
     });

api/server/services/Endpoints/anthropic/initialize.js

@@ -41,7 +41,7 @@ const initializeClient = async ({ req, res, endpointOption, overrideModel, optio
       {
         reverseProxyUrl: ANTHROPIC_REVERSE_PROXY ?? null,
         proxy: PROXY ?? null,
-        modelOptions: endpointOption.model_parameters,
+        modelOptions: endpointOption?.model_parameters ?? {},
       },
       clientOptions,
     );

api/server/services/Endpoints/anthropic/llm.js

@@ -75,6 +75,7 @@ function getLLMConfig(apiKey, options = {}) {
 
   if (options.reverseProxyUrl) {
     requestOptions.clientOptions.baseURL = options.reverseProxyUrl;
+    requestOptions.anthropicApiUrl = options.reverseProxyUrl;
   }
 
   return {

api/server/services/Endpoints/anthropic/llm.spec.js

@@ -1,11 +1,45 @@
-const { anthropicSettings } = require('librechat-data-provider');
+const { anthropicSettings, removeNullishValues } = require('librechat-data-provider');
 const { getLLMConfig } = require('~/server/services/Endpoints/anthropic/llm');
+const { checkPromptCacheSupport, getClaudeHeaders, configureReasoning } = require('./helpers');
 
 jest.mock('https-proxy-agent', () => ({
   HttpsProxyAgent: jest.fn().mockImplementation((proxy) => ({ proxy })),
 }));
 
+jest.mock('./helpers', () => ({
+  checkPromptCacheSupport: jest.fn(),
+  getClaudeHeaders: jest.fn(),
+  configureReasoning: jest.fn((requestOptions) => requestOptions),
+}));
+
+jest.mock('librechat-data-provider', () => ({
+  anthropicSettings: {
+    model: { default: 'claude-3-opus-20240229' },
+    maxOutputTokens: { default: 4096, reset: jest.fn(() => 4096) },
+    thinking: { default: false },
+    promptCache: { default: false },
+    thinkingBudget: { default: null },
+  },
+  removeNullishValues: jest.fn((obj) => {
+    const result = {};
+    for (const key in obj) {
+      if (obj[key] !== null && obj[key] !== undefined) {
+        result[key] = obj[key];
+      }
+    }
+    return result;
+  }),
+}));
+
 describe('getLLMConfig', () => {
+  beforeEach(() => {
+    jest.clearAllMocks();
+    checkPromptCacheSupport.mockReturnValue(false);
+    getClaudeHeaders.mockReturnValue(undefined);
+    configureReasoning.mockImplementation((requestOptions) => requestOptions);
+    anthropicSettings.maxOutputTokens.reset.mockReturnValue(4096);
+  });
+
   it('should create a basic configuration with default values', () => {
     const result = getLLMConfig('test-api-key', { modelOptions: {} });
 
@@ -36,6 +70,7 @@ describe('getLLMConfig', () => {
     });
 
     expect(result.llmConfig.clientOptions).toHaveProperty('baseURL', 'http://reverse-proxy');
+    expect(result.llmConfig).toHaveProperty('anthropicApiUrl', 'http://reverse-proxy');
   });
 
   it('should include topK and topP for non-Claude-3.7 models', () => {
@@ -65,6 +100,11 @@ describe('getLLMConfig', () => {
   });
 
   it('should NOT include topK and topP for Claude-3-7 models (hyphen notation)', () => {
+    configureReasoning.mockImplementation((requestOptions) => {
+      requestOptions.thinking = { type: 'enabled' };
+      return requestOptions;
+    });
+
     const result = getLLMConfig('test-api-key', {
       modelOptions: {
         model: 'claude-3-7-sonnet',
@@ -78,6 +118,11 @@ describe('getLLMConfig', () => {
   });
 
   it('should NOT include topK and topP for Claude-3.7 models (decimal notation)', () => {
+    configureReasoning.mockImplementation((requestOptions) => {
+      requestOptions.thinking = { type: 'enabled' };
+      return requestOptions;
+    });
+
     const result = getLLMConfig('test-api-key', {
       modelOptions: {
         model: 'claude-3.7-sonnet',
@@ -154,4 +199,160 @@ describe('getLLMConfig', () => {
     expect(result3.llmConfig).toHaveProperty('topK', 10);
     expect(result3.llmConfig).toHaveProperty('topP', 0.9);
   });
+
+  describe('Edge cases', () => {
+    it('should handle missing apiKey', () => {
+      const result = getLLMConfig(undefined, { modelOptions: {} });
+      expect(result.llmConfig).not.toHaveProperty('apiKey');
+    });
+
+    it('should handle empty modelOptions', () => {
+      expect(() => {
+        getLLMConfig('test-api-key', {});
+      }).toThrow("Cannot read properties of undefined (reading 'thinking')");
+    });
+
+    it('should handle no options parameter', () => {
+      expect(() => {
+        getLLMConfig('test-api-key');
+      }).toThrow("Cannot read properties of undefined (reading 'thinking')");
+    });
+
+    it('should handle temperature, stop sequences, and stream settings', () => {
+      const result = getLLMConfig('test-api-key', {
+        modelOptions: {
+          temperature: 0.7,
+          stop: ['\n\n', 'END'],
+          stream: false,
+        },
+      });
+
+      expect(result.llmConfig).toHaveProperty('temperature', 0.7);
+      expect(result.llmConfig).toHaveProperty('stopSequences', ['\n\n', 'END']);
+      expect(result.llmConfig).toHaveProperty('stream', false);
+    });
+
+    it('should handle maxOutputTokens when explicitly set to falsy value', () => {
+      anthropicSettings.maxOutputTokens.reset.mockReturnValue(8192);
+      const result = getLLMConfig('test-api-key', {
+        modelOptions: {
+          model: 'claude-3-opus',
+          maxOutputTokens: null,
+        },
+      });
+
+      expect(anthropicSettings.maxOutputTokens.reset).toHaveBeenCalledWith('claude-3-opus');
+      expect(result.llmConfig).toHaveProperty('maxTokens', 8192);
+    });
+
+    it('should handle both proxy and reverseProxyUrl', () => {
+      const result = getLLMConfig('test-api-key', {
+        modelOptions: {},
+        proxy: 'http://proxy:8080',
+        reverseProxyUrl: 'https://reverse-proxy.com',
+      });
+
+      expect(result.llmConfig.clientOptions).toHaveProperty('fetchOptions');
+      expect(result.llmConfig.clientOptions.fetchOptions).toHaveProperty('dispatcher');
+      expect(result.llmConfig.clientOptions.fetchOptions.dispatcher).toBeDefined();
+      expect(result.llmConfig.clientOptions.fetchOptions.dispatcher.constructor.name).toBe(
+        'ProxyAgent',
+      );
+      expect(result.llmConfig.clientOptions).toHaveProperty('baseURL', 'https://reverse-proxy.com');
+      expect(result.llmConfig).toHaveProperty('anthropicApiUrl', 'https://reverse-proxy.com');
+    });
+
+    it('should handle prompt cache with supported model', () => {
+      checkPromptCacheSupport.mockReturnValue(true);
+      getClaudeHeaders.mockReturnValue({ 'anthropic-beta': 'prompt-caching-2024-07-31' });
+
+      const result = getLLMConfig('test-api-key', {
+        modelOptions: {
+          model: 'claude-3-5-sonnet',
+          promptCache: true,
+        },
+      });
+
+      expect(checkPromptCacheSupport).toHaveBeenCalledWith('claude-3-5-sonnet');
+      expect(getClaudeHeaders).toHaveBeenCalledWith('claude-3-5-sonnet', true);
+      expect(result.llmConfig.clientOptions.defaultHeaders).toEqual({
+        'anthropic-beta': 'prompt-caching-2024-07-31',
+      });
+    });
+
+    it('should handle thinking and thinkingBudget options', () => {
+      configureReasoning.mockImplementation((requestOptions, systemOptions) => {
+        if (systemOptions.thinking) {
+          requestOptions.thinking = { type: 'enabled' };
+        }
+        if (systemOptions.thinkingBudget) {
+          requestOptions.thinking = {
+            ...requestOptions.thinking,
+            budget_tokens: systemOptions.thinkingBudget,
+          };
+        }
+        return requestOptions;
+      });
+
+      getLLMConfig('test-api-key', {
+        modelOptions: {
+          model: 'claude-3-7-sonnet',
+          thinking: true,
+          thinkingBudget: 5000,
+        },
+      });
+
+      expect(configureReasoning).toHaveBeenCalledWith(
+        expect.any(Object),
+        expect.objectContaining({
+          thinking: true,
+          promptCache: false,
+          thinkingBudget: 5000,
+        }),
+      );
+    });
+
+    it('should remove system options from modelOptions', () => {
+      const modelOptions = {
+        model: 'claude-3-opus',
+        thinking: true,
+        promptCache: true,
+        thinkingBudget: 1000,
+        temperature: 0.5,
+      };
+
+      getLLMConfig('test-api-key', { modelOptions });
+
+      expect(modelOptions).not.toHaveProperty('thinking');
+      expect(modelOptions).not.toHaveProperty('promptCache');
+      expect(modelOptions).not.toHaveProperty('thinkingBudget');
+      expect(modelOptions).toHaveProperty('temperature', 0.5);
+    });
+
+    it('should handle all nullish values removal', () => {
+      removeNullishValues.mockImplementation((obj) => {
+        const cleaned = {};
+        Object.entries(obj).forEach(([key, value]) => {
+          if (value !== null && value !== undefined) {
+            cleaned[key] = value;
+          }
+        });
+        return cleaned;
+      });
+
+      const result = getLLMConfig('test-api-key', {
+        modelOptions: {
+          temperature: null,
+          topP: undefined,
+          topK: 0,
+          stop: [],
+        },
+      });
+
+      expect(result.llmConfig).not.toHaveProperty('temperature');
+      expect(result.llmConfig).not.toHaveProperty('topP');
+      expect(result.llmConfig).toHaveProperty('topK', 0);
+      expect(result.llmConfig).toHaveProperty('stopSequences', []);
+    });
+  });
 });

api/server/services/Endpoints/azureAssistants/initialize.js

@@ -1,12 +1,7 @@
 const OpenAI = require('openai');
 const { HttpsProxyAgent } = require('https-proxy-agent');
-const { constructAzureURL, isUserProvided } = require('@librechat/api');
-const {
-  ErrorTypes,
-  EModelEndpoint,
-  resolveHeaders,
-  mapModelToAzureConfig,
-} = require('librechat-data-provider');
+const { constructAzureURL, isUserProvided, resolveHeaders } = require('@librechat/api');
+const { ErrorTypes, EModelEndpoint, mapModelToAzureConfig } = require('librechat-data-provider');
 const {
   getUserKeyValues,
   getUserKeyExpiry,
@@ -114,11 +109,14 @@ const initializeClient = async ({ req, res, version, endpointOption, initAppClie
 
     apiKey = azureOptions.azureOpenAIApiKey;
     opts.defaultQuery = { 'api-version': azureOptions.azureOpenAIApiVersion };
-    opts.defaultHeaders = resolveHeaders({
-      ...headers,
-      'api-key': apiKey,
-      'OpenAI-Beta': `assistants=${version}`,
-    });
+    opts.defaultHeaders = resolveHeaders(
+      {
+        ...headers,
+        'api-key': apiKey,
+        'OpenAI-Beta': `assistants=${version}`,
+      },
+      req.user,
+    );
     opts.model = azureOptions.azureOpenAIApiDeploymentName;
 
     if (initAppClient) {

api/server/services/Endpoints/bedrock/options.js

@@ -64,7 +64,7 @@ const getOptions = async ({ req, overrideModel, endpointOption }) => {
 
   /** @type {BedrockClientOptions} */
   const requestOptions = {
-    model: overrideModel ?? endpointOption.model,
+    model: overrideModel ?? endpointOption?.model,
     region: BEDROCK_AWS_DEFAULT_REGION,
   };
 
@@ -76,7 +76,7 @@ const getOptions = async ({ req, overrideModel, endpointOption }) => {
 
   const llmConfig = bedrockOutputParser(
     bedrockInputParser.parse(
-      removeNullishValues(Object.assign(requestOptions, endpointOption.model_parameters)),
+      removeNullishValues(Object.assign(requestOptions, endpointOption?.model_parameters ?? {})),
     ),
   );
 

api/server/services/Endpoints/custom/initialize.js

@@ -6,7 +6,7 @@ const {
   extractEnvVariable,
 } = require('librechat-data-provider');
 const { Providers } = require('@librechat/agents');
-const { getOpenAIConfig, createHandleLLMNewToken } = require('@librechat/api');
+const { getOpenAIConfig, createHandleLLMNewToken, resolveHeaders } = require('@librechat/api');
 const { getUserKeyValues, checkUserKeyExpiry } = require('~/server/services/UserService');
 const { getCustomEndpointConfig } = require('~/server/services/Config');
 const { fetchModels } = require('~/server/services/ModelService');
@@ -28,12 +28,7 @@ const initializeClient = async ({ req, res, endpointOption, optionsOnly, overrid
   const CUSTOM_API_KEY = extractEnvVariable(endpointConfig.apiKey);
   const CUSTOM_BASE_URL = extractEnvVariable(endpointConfig.baseURL);
 
-  let resolvedHeaders = {};
-  if (endpointConfig.headers && typeof endpointConfig.headers === 'object') {
-    Object.keys(endpointConfig.headers).forEach((key) => {
-      resolvedHeaders[key] = extractEnvVariable(endpointConfig.headers[key]);
-    });
-  }
+  let resolvedHeaders = resolveHeaders(endpointConfig.headers, req.user);
 
   if (CUSTOM_API_KEY.match(envVarRegex)) {
     throw new Error(`Missing API Key for ${endpoint}.`);
@@ -134,7 +129,7 @@ const initializeClient = async ({ req, res, endpointOption, optionsOnly, overrid
   };
 
   if (optionsOnly) {
-    const modelOptions = endpointOption.model_parameters;
+    const modelOptions = endpointOption?.model_parameters ?? {};
     if (endpoint !== Providers.OLLAMA) {
       clientOptions = Object.assign(
         {

api/server/services/Endpoints/custom/initialize.spec.js

@@ -0,0 +1,93 @@
+const initializeClient = require('./initialize');
+
+jest.mock('@librechat/api', () => ({
+  resolveHeaders: jest.fn(),
+  getOpenAIConfig: jest.fn(),
+  createHandleLLMNewToken: jest.fn(),
+}));
+
+jest.mock('librechat-data-provider', () => ({
+  CacheKeys: { TOKEN_CONFIG: 'token_config' },
+  ErrorTypes: { NO_USER_KEY: 'NO_USER_KEY', NO_BASE_URL: 'NO_BASE_URL' },
+  envVarRegex: /\$\{([^}]+)\}/,
+  FetchTokenConfig: {},
+  extractEnvVariable: jest.fn((value) => value),
+}));
+
+jest.mock('@librechat/agents', () => ({
+  Providers: { OLLAMA: 'ollama' },
+}));
+
+jest.mock('~/server/services/UserService', () => ({
+  getUserKeyValues: jest.fn(),
+  checkUserKeyExpiry: jest.fn(),
+}));
+
+jest.mock('~/server/services/Config', () => ({
+  getCustomEndpointConfig: jest.fn().mockResolvedValue({
+    apiKey: 'test-key',
+    baseURL: 'https://test.com',
+    headers: { 'x-user': '{{LIBRECHAT_USER_ID}}', 'x-email': '{{LIBRECHAT_USER_EMAIL}}' },
+    models: { default: ['test-model'] },
+  }),
+}));
+
+jest.mock('~/server/services/ModelService', () => ({
+  fetchModels: jest.fn(),
+}));
+
+jest.mock('~/app/clients/OpenAIClient', () => {
+  return jest.fn().mockImplementation(() => ({
+    options: {},
+  }));
+});
+
+jest.mock('~/server/utils', () => ({
+  isUserProvided: jest.fn().mockReturnValue(false),
+}));
+
+jest.mock('~/cache/getLogStores', () =>
+  jest.fn().mockReturnValue({
+    get: jest.fn(),
+  }),
+);
+
+describe('custom/initializeClient', () => {
+  const mockRequest = {
+    body: { endpoint: 'test-endpoint' },
+    user: { id: 'user-123', email: 'test@example.com' },
+    app: { locals: {} },
+  };
+  const mockResponse = {};
+
+  beforeEach(() => {
+    jest.clearAllMocks();
+  });
+
+  it('calls resolveHeaders with headers and user', async () => {
+    const { resolveHeaders } = require('@librechat/api');
+    await initializeClient({ req: mockRequest, res: mockResponse, optionsOnly: true });
+    expect(resolveHeaders).toHaveBeenCalledWith(
+      { 'x-user': '{{LIBRECHAT_USER_ID}}', 'x-email': '{{LIBRECHAT_USER_EMAIL}}' },
+      { id: 'user-123', email: 'test@example.com' },
+    );
+  });
+
+  it('throws if endpoint config is missing', async () => {
+    const { getCustomEndpointConfig } = require('~/server/services/Config');
+    getCustomEndpointConfig.mockResolvedValueOnce(null);
+    await expect(
+      initializeClient({ req: mockRequest, res: mockResponse, optionsOnly: true }),
+    ).rejects.toThrow('Config not found for the test-endpoint custom endpoint.');
+  });
+
+  it('throws if user is missing', async () => {
+    await expect(
+      initializeClient({
+        req: { ...mockRequest, user: undefined },
+        res: mockResponse,
+        optionsOnly: true,
+      }),
+    ).rejects.toThrow("Cannot read properties of undefined (reading 'id')");
+  });
+});

api/server/services/Endpoints/google/initialize.js

@@ -1,7 +1,7 @@
+const path = require('path');
 const { EModelEndpoint, AuthKeys } = require('librechat-data-provider');
+const { getGoogleConfig, isEnabled, loadServiceKey } = require('@librechat/api');
 const { getUserKey, checkUserKeyExpiry } = require('~/server/services/UserService');
-const { getLLMConfig } = require('~/server/services/Endpoints/google/llm');
-const { isEnabled } = require('~/server/utils');
 const { GoogleClient } = require('~/app');
 
 const initializeClient = async ({ req, res, endpointOption, overrideModel, optionsOnly }) => {
@@ -16,9 +16,16 @@ const initializeClient = async ({ req, res, endpointOption, overrideModel, optio
   }
 
   let serviceKey = {};
+
   try {
-    serviceKey = require('~/data/auth.json');
-  } catch (e) {
+    const serviceKeyPath =
+      process.env.GOOGLE_SERVICE_KEY_FILE_PATH ||
+      path.join(__dirname, '../../../..', 'data', 'auth.json');
+    serviceKey = await loadServiceKey(serviceKeyPath);
+    if (!serviceKey) {
+      serviceKey = {};
+    }
+  } catch (_e) {
     // Do nothing
   }
 
@@ -58,14 +65,14 @@ const initializeClient = async ({ req, res, endpointOption, overrideModel, optio
   if (optionsOnly) {
     clientOptions = Object.assign(
       {
-        modelOptions: endpointOption.model_parameters,
+        modelOptions: endpointOption?.model_parameters ?? {},
       },
       clientOptions,
     );
     if (overrideModel) {
       clientOptions.modelOptions.model = overrideModel;
     }
-    return getLLMConfig(credentials, clientOptions);
+    return getGoogleConfig(credentials, clientOptions);
   }
 
   const client = new GoogleClient(credentials, clientOptions);

api/server/services/Endpoints/gptPlugins/build.js

@@ -1,41 +0,0 @@
-const { removeNullishValues } = require('librechat-data-provider');
-const generateArtifactsPrompt = require('~/app/clients/prompts/artifacts');
-
-const buildOptions = (endpoint, parsedBody) => {
-  const {
-    modelLabel,
-    chatGptLabel,
-    promptPrefix,
-    agentOptions,
-    tools = [],
-    iconURL,
-    greeting,
-    spec,
-    maxContextTokens,
-    artifacts,
-    ...modelOptions
-  } = parsedBody;
-  const endpointOption = removeNullishValues({
-    endpoint,
-    tools: tools
-      .map((tool) => tool?.pluginKey ?? tool)
-      .filter((toolName) => typeof toolName === 'string'),
-    modelLabel,
-    chatGptLabel,
-    promptPrefix,
-    agentOptions,
-    iconURL,
-    greeting,
-    spec,
-    maxContextTokens,
-    modelOptions,
-  });
-
-  if (typeof artifacts === 'string') {
-    endpointOption.artifactsPrompt = generateArtifactsPrompt({ endpoint, artifacts });
-  }
-
-  return endpointOption;
-};
-
-module.exports = buildOptions;

api/server/services/Endpoints/gptPlugins/index.js

@@ -1,7 +0,0 @@
-const buildOptions = require('./build');
-const initializeClient = require('./initialize');
-
-module.exports = {
-  buildOptions,
-  initializeClient,
-};

api/server/services/Endpoints/gptPlugins/initialize.js

@@ -1,134 +0,0 @@
-const {
-  EModelEndpoint,
-  resolveHeaders,
-  mapModelToAzureConfig,
-} = require('librechat-data-provider');
-const { isEnabled, isUserProvided, getAzureCredentials } = require('@librechat/api');
-const { getUserKeyValues, checkUserKeyExpiry } = require('~/server/services/UserService');
-const { PluginsClient } = require('~/app');
-
-const initializeClient = async ({ req, res, endpointOption }) => {
-  const {
-    PROXY,
-    OPENAI_API_KEY,
-    AZURE_API_KEY,
-    PLUGINS_USE_AZURE,
-    OPENAI_REVERSE_PROXY,
-    AZURE_OPENAI_BASEURL,
-    OPENAI_SUMMARIZE,
-    DEBUG_PLUGINS,
-  } = process.env;
-
-  const { key: expiresAt, model: modelName } = req.body;
-  const contextStrategy = isEnabled(OPENAI_SUMMARIZE) ? 'summarize' : null;
-
-  let useAzure = isEnabled(PLUGINS_USE_AZURE);
-  let endpoint = useAzure ? EModelEndpoint.azureOpenAI : EModelEndpoint.openAI;
-
-  /** @type {false | TAzureConfig} */
-  const azureConfig = req.app.locals[EModelEndpoint.azureOpenAI];
-  useAzure = useAzure || azureConfig?.plugins;
-
-  if (useAzure && endpoint !== EModelEndpoint.azureOpenAI) {
-    endpoint = EModelEndpoint.azureOpenAI;
-  }
-
-  const credentials = {
-    [EModelEndpoint.openAI]: OPENAI_API_KEY,
-    [EModelEndpoint.azureOpenAI]: AZURE_API_KEY,
-  };
-
-  const baseURLOptions = {
-    [EModelEndpoint.openAI]: OPENAI_REVERSE_PROXY,
-    [EModelEndpoint.azureOpenAI]: AZURE_OPENAI_BASEURL,
-  };
-
-  const userProvidesKey = isUserProvided(credentials[endpoint]);
-  const userProvidesURL = isUserProvided(baseURLOptions[endpoint]);
-
-  let userValues = null;
-  if (expiresAt && (userProvidesKey || userProvidesURL)) {
-    checkUserKeyExpiry(expiresAt, endpoint);
-    userValues = await getUserKeyValues({ userId: req.user.id, name: endpoint });
-  }
-
-  let apiKey = userProvidesKey ? userValues?.apiKey : credentials[endpoint];
-  let baseURL = userProvidesURL ? userValues?.baseURL : baseURLOptions[endpoint];
-
-  const clientOptions = {
-    contextStrategy,
-    debug: isEnabled(DEBUG_PLUGINS),
-    reverseProxyUrl: baseURL ? baseURL : null,
-    proxy: PROXY ?? null,
-    req,
-    res,
-    ...endpointOption,
-  };
-
-  if (useAzure && azureConfig) {
-    const { modelGroupMap, groupMap } = azureConfig;
-    const {
-      azureOptions,
-      baseURL,
-      headers = {},
-      serverless,
-    } = mapModelToAzureConfig({
-      modelName,
-      modelGroupMap,
-      groupMap,
-    });
-
-    clientOptions.reverseProxyUrl = baseURL ?? clientOptions.reverseProxyUrl;
-    clientOptions.headers = resolveHeaders({ ...headers, ...(clientOptions.headers ?? {}) });
-
-    clientOptions.titleConvo = azureConfig.titleConvo;
-    clientOptions.titleModel = azureConfig.titleModel;
-    clientOptions.titleMethod = azureConfig.titleMethod ?? 'completion';
-
-    const azureRate = modelName.includes('gpt-4') ? 30 : 17;
-    clientOptions.streamRate = azureConfig.streamRate ?? azureRate;
-
-    const groupName = modelGroupMap[modelName].group;
-    clientOptions.addParams = azureConfig.groupMap[groupName].addParams;
-    clientOptions.dropParams = azureConfig.groupMap[groupName].dropParams;
-    clientOptions.forcePrompt = azureConfig.groupMap[groupName].forcePrompt;
-
-    apiKey = azureOptions.azureOpenAIApiKey;
-    clientOptions.azure = !serverless && azureOptions;
-    if (serverless === true) {
-      clientOptions.defaultQuery = azureOptions.azureOpenAIApiVersion
-        ? { 'api-version': azureOptions.azureOpenAIApiVersion }
-        : undefined;
-      clientOptions.headers['api-key'] = apiKey;
-    }
-  } else if (useAzure || (apiKey && apiKey.includes('{"azure') && !clientOptions.azure)) {
-    clientOptions.azure = userProvidesKey ? JSON.parse(userValues.apiKey) : getAzureCredentials();
-    apiKey = clientOptions.azure.azureOpenAIApiKey;
-  }
-
-  /** @type {undefined | TBaseEndpoint} */
-  const pluginsConfig = req.app.locals[EModelEndpoint.gptPlugins];
-
-  if (!useAzure && pluginsConfig) {
-    clientOptions.streamRate = pluginsConfig.streamRate;
-  }
-
-  /** @type {undefined | TBaseEndpoint} */
-  const allConfig = req.app.locals.all;
-  if (allConfig) {
-    clientOptions.streamRate = allConfig.streamRate;
-  }
-
-  if (!apiKey) {
-    throw new Error(`${endpoint} API key not provided. Please provide it again.`);
-  }
-
-  const client = new PluginsClient(apiKey, clientOptions);
-  return {
-    client,
-    azure: clientOptions.azure,
-    openAIApiKey: apiKey,
-  };
-};
-
-module.exports = initializeClient;

api/server/services/Endpoints/gptPlugins/initialize.spec.js

@@ -1,410 +0,0 @@
-// gptPlugins/initializeClient.spec.js
-jest.mock('~/cache/getLogStores');
-const { EModelEndpoint, ErrorTypes, validateAzureGroups } = require('librechat-data-provider');
-const { getUserKey, getUserKeyValues } = require('~/server/services/UserService');
-const initializeClient = require('./initialize');
-const { PluginsClient } = require('~/app');
-
-// Mock getUserKey since it's the only function we want to mock
-jest.mock('~/server/services/UserService', () => ({
-  getUserKey: jest.fn(),
-  getUserKeyValues: jest.fn(),
-  checkUserKeyExpiry: jest.requireActual('~/server/services/UserService').checkUserKeyExpiry,
-}));
-
-describe('gptPlugins/initializeClient', () => {
-  // Set up environment variables
-  const originalEnvironment = process.env;
-  const app = {
-    locals: {},
-  };
-
-  const validAzureConfigs = [
-    {
-      group: 'librechat-westus',
-      apiKey: 'WESTUS_API_KEY',
-      instanceName: 'librechat-westus',
-      version: '2023-12-01-preview',
-      models: {
-        'gpt-4-vision-preview': {
-          deploymentName: 'gpt-4-vision-preview',
-          version: '2024-02-15-preview',
-        },
-        'gpt-3.5-turbo': {
-          deploymentName: 'gpt-35-turbo',
-        },
-        'gpt-3.5-turbo-1106': {
-          deploymentName: 'gpt-35-turbo-1106',
-        },
-        'gpt-4': {
-          deploymentName: 'gpt-4',
-        },
-        'gpt-4-1106-preview': {
-          deploymentName: 'gpt-4-1106-preview',
-        },
-      },
-    },
-    {
-      group: 'librechat-eastus',
-      apiKey: 'EASTUS_API_KEY',
-      instanceName: 'librechat-eastus',
-      deploymentName: 'gpt-4-turbo',
-      version: '2024-02-15-preview',
-      models: {
-        'gpt-4-turbo': true,
-      },
-      baseURL: 'https://eastus.example.com',
-      additionalHeaders: {
-        'x-api-key': 'x-api-key-value',
-      },
-    },
-    {
-      group: 'mistral-inference',
-      apiKey: 'AZURE_MISTRAL_API_KEY',
-      baseURL:
-        'https://Mistral-large-vnpet-serverless.region.inference.ai.azure.com/v1/chat/completions',
-      serverless: true,
-      models: {
-        'mistral-large': true,
-      },
-    },
-    {
-      group: 'llama-70b-chat',
-      apiKey: 'AZURE_LLAMA2_70B_API_KEY',
-      baseURL:
-        'https://Llama-2-70b-chat-qmvyb-serverless.region.inference.ai.azure.com/v1/chat/completions',
-      serverless: true,
-      models: {
-        'llama-70b-chat': true,
-      },
-    },
-  ];
-
-  const { modelNames, modelGroupMap, groupMap } = validateAzureGroups(validAzureConfigs);
-
-  beforeEach(() => {
-    jest.resetModules(); // Clears the cache
-    process.env = { ...originalEnvironment }; // Make a copy
-  });
-
-  afterAll(() => {
-    process.env = originalEnvironment; // Restore original env vars
-  });
-
-  test('should initialize PluginsClient with OpenAI API key and default options', async () => {
-    process.env.OPENAI_API_KEY = 'test-openai-api-key';
-    process.env.PLUGINS_USE_AZURE = 'false';
-    process.env.DEBUG_PLUGINS = 'false';
-    process.env.OPENAI_SUMMARIZE = 'false';
-
-    const req = {
-      body: { key: null },
-      user: { id: '123' },
-      app,
-    };
-    const res = {};
-    const endpointOption = { modelOptions: { model: 'default-model' } };
-
-    const { client, openAIApiKey } = await initializeClient({ req, res, endpointOption });
-
-    expect(openAIApiKey).toBe('test-openai-api-key');
-    expect(client).toBeInstanceOf(PluginsClient);
-  });
-
-  test('should initialize PluginsClient with Azure credentials when PLUGINS_USE_AZURE is true', async () => {
-    process.env.AZURE_API_KEY = 'test-azure-api-key';
-    (process.env.AZURE_OPENAI_API_INSTANCE_NAME = 'some-value'),
-      (process.env.AZURE_OPENAI_API_DEPLOYMENT_NAME = 'some-value'),
-      (process.env.AZURE_OPENAI_API_VERSION = 'some-value'),
-      (process.env.AZURE_OPENAI_API_COMPLETIONS_DEPLOYMENT_NAME = 'some-value'),
-      (process.env.AZURE_OPENAI_API_EMBEDDINGS_DEPLOYMENT_NAME = 'some-value'),
-      (process.env.PLUGINS_USE_AZURE = 'true');
-    process.env.DEBUG_PLUGINS = 'false';
-    process.env.OPENAI_SUMMARIZE = 'false';
-
-    const req = {
-      body: { key: null },
-      user: { id: '123' },
-      app,
-    };
-    const res = {};
-    const endpointOption = { modelOptions: { model: 'test-model' } };
-
-    const { client, azure } = await initializeClient({ req, res, endpointOption });
-
-    expect(azure.azureOpenAIApiKey).toBe('test-azure-api-key');
-    expect(client).toBeInstanceOf(PluginsClient);
-  });
-
-  test('should use the debug option when DEBUG_PLUGINS is enabled', async () => {
-    process.env.OPENAI_API_KEY = 'test-openai-api-key';
-    process.env.DEBUG_PLUGINS = 'true';
-
-    const req = {
-      body: { key: null },
-      user: { id: '123' },
-      app,
-    };
-    const res = {};
-    const endpointOption = { modelOptions: { model: 'default-model' } };
-
-    const { client } = await initializeClient({ req, res, endpointOption });
-
-    expect(client.options.debug).toBe(true);
-  });
-
-  test('should set contextStrategy to summarize when OPENAI_SUMMARIZE is enabled', async () => {
-    process.env.OPENAI_API_KEY = 'test-openai-api-key';
-    process.env.OPENAI_SUMMARIZE = 'true';
-
-    const req = {
-      body: { key: null },
-      user: { id: '123' },
-      app,
-    };
-    const res = {};
-    const endpointOption = { modelOptions: { model: 'default-model' } };
-
-    const { client } = await initializeClient({ req, res, endpointOption });
-
-    expect(client.options.contextStrategy).toBe('summarize');
-  });
-
-  // ... additional tests for reverseProxyUrl, proxy, user-provided keys, etc.
-
-  test('should throw an error if no API keys are provided in the environment', async () => {
-    // Clear the environment variables for API keys
-    delete process.env.OPENAI_API_KEY;
-    delete process.env.AZURE_API_KEY;
-
-    const req = {
-      body: { key: null },
-      user: { id: '123' },
-      app,
-    };
-    const res = {};
-    const endpointOption = { modelOptions: { model: 'default-model' } };
-
-    await expect(initializeClient({ req, res, endpointOption })).rejects.toThrow(
-      `${EModelEndpoint.openAI} API key not provided.`,
-    );
-  });
-
-  // Additional tests for gptPlugins/initializeClient.spec.js
-
-  // ... (previous test setup code)
-
-  test('should handle user-provided OpenAI keys and check expiry', async () => {
-    process.env.OPENAI_API_KEY = 'user_provided';
-    process.env.PLUGINS_USE_AZURE = 'false';
-
-    const futureDate = new Date(Date.now() + 10000).toISOString();
-    const req = {
-      body: { key: futureDate },
-      user: { id: '123' },
-      app,
-    };
-    const res = {};
-    const endpointOption = { modelOptions: { model: 'default-model' } };
-
-    getUserKeyValues.mockResolvedValue({ apiKey: 'test-user-provided-openai-api-key' });
-
-    const { openAIApiKey } = await initializeClient({ req, res, endpointOption });
-
-    expect(openAIApiKey).toBe('test-user-provided-openai-api-key');
-  });
-
-  test('should handle user-provided Azure keys and check expiry', async () => {
-    process.env.AZURE_API_KEY = 'user_provided';
-    process.env.PLUGINS_USE_AZURE = 'true';
-
-    const futureDate = new Date(Date.now() + 10000).toISOString();
-    const req = {
-      body: { key: futureDate },
-      user: { id: '123' },
-      app,
-    };
-    const res = {};
-    const endpointOption = { modelOptions: { model: 'test-model' } };
-
-    getUserKeyValues.mockResolvedValue({
-      apiKey: JSON.stringify({
-        azureOpenAIApiKey: 'test-user-provided-azure-api-key',
-        azureOpenAIApiDeploymentName: 'test-deployment',
-      }),
-    });
-
-    const { azure } = await initializeClient({ req, res, endpointOption });
-
-    expect(azure.azureOpenAIApiKey).toBe('test-user-provided-azure-api-key');
-  });
-
-  test('should throw an error if the user-provided key has expired', async () => {
-    process.env.OPENAI_API_KEY = 'user_provided';
-    process.env.PLUGINS_USE_AZURE = 'FALSE';
-    const expiresAt = new Date(Date.now() - 10000).toISOString(); // Expired
-    const req = {
-      body: { key: expiresAt },
-      user: { id: '123' },
-      app,
-    };
-    const res = {};
-    const endpointOption = { modelOptions: { model: 'default-model' } };
-
-    await expect(initializeClient({ req, res, endpointOption })).rejects.toThrow(
-      /expired_user_key/,
-    );
-  });
-
-  test('should throw an error if the user-provided Azure key is invalid JSON', async () => {
-    process.env.AZURE_API_KEY = 'user_provided';
-    process.env.PLUGINS_USE_AZURE = 'true';
-
-    const req = {
-      body: { key: new Date(Date.now() + 10000).toISOString() },
-      user: { id: '123' },
-      app,
-    };
-    const res = {};
-    const endpointOption = { modelOptions: { model: 'default-model' } };
-
-    // Simulate an invalid JSON string returned from getUserKey
-    getUserKey.mockResolvedValue('invalid-json');
-    getUserKeyValues.mockImplementation(() => {
-      let userValues = getUserKey();
-      try {
-        userValues = JSON.parse(userValues);
-      } catch (e) {
-        throw new Error(
-          JSON.stringify({
-            type: ErrorTypes.INVALID_USER_KEY,
-          }),
-        );
-      }
-      return userValues;
-    });
-
-    await expect(initializeClient({ req, res, endpointOption })).rejects.toThrow(
-      /invalid_user_key/,
-    );
-  });
-
-  test('should correctly handle the presence of a reverse proxy', async () => {
-    process.env.OPENAI_REVERSE_PROXY = 'http://reverse.proxy';
-    process.env.PROXY = 'http://proxy';
-    process.env.OPENAI_API_KEY = 'test-openai-api-key';
-
-    const req = {
-      body: { key: null },
-      user: { id: '123' },
-      app,
-    };
-    const res = {};
-    const endpointOption = { modelOptions: { model: 'default-model' } };
-
-    const { client } = await initializeClient({ req, res, endpointOption });
-
-    expect(client.options.reverseProxyUrl).toBe('http://reverse.proxy');
-    expect(client.options.proxy).toBe('http://proxy');
-  });
-
-  test('should throw an error when user-provided values are not valid JSON', async () => {
-    process.env.OPENAI_API_KEY = 'user_provided';
-    const req = {
-      body: { key: new Date(Date.now() + 10000).toISOString(), endpoint: 'openAI' },
-      user: { id: '123' },
-      app,
-    };
-    const res = {};
-    const endpointOption = {};
-
-    // Mock getUserKey to return a non-JSON string
-    getUserKey.mockResolvedValue('not-a-json');
-    getUserKeyValues.mockImplementation(() => {
-      let userValues = getUserKey();
-      try {
-        userValues = JSON.parse(userValues);
-      } catch (e) {
-        throw new Error(
-          JSON.stringify({
-            type: ErrorTypes.INVALID_USER_KEY,
-          }),
-        );
-      }
-      return userValues;
-    });
-
-    await expect(initializeClient({ req, res, endpointOption })).rejects.toThrow(
-      /invalid_user_key/,
-    );
-  });
-
-  test('should initialize client correctly for Azure OpenAI with valid configuration', async () => {
-    const req = {
-      body: {
-        key: null,
-        endpoint: EModelEndpoint.gptPlugins,
-        model: modelNames[0],
-      },
-      user: { id: '123' },
-      app: {
-        locals: {
-          [EModelEndpoint.azureOpenAI]: {
-            plugins: true,
-            modelNames,
-            modelGroupMap,
-            groupMap,
-          },
-        },
-      },
-    };
-    const res = {};
-    const endpointOption = {};
-
-    const client = await initializeClient({ req, res, endpointOption });
-    expect(client.client.options.azure).toBeDefined();
-  });
-
-  test('should initialize client with default options when certain env vars are not set', async () => {
-    delete process.env.OPENAI_SUMMARIZE;
-    process.env.OPENAI_API_KEY = 'some-api-key';
-
-    const req = {
-      body: { key: null, endpoint: EModelEndpoint.gptPlugins },
-      user: { id: '123' },
-      app,
-    };
-    const res = {};
-    const endpointOption = {};
-
-    const client = await initializeClient({ req, res, endpointOption });
-    expect(client.client.options.contextStrategy).toBe(null);
-  });
-
-  test('should correctly use user-provided apiKey and baseURL when provided', async () => {
-    process.env.OPENAI_API_KEY = 'user_provided';
-    process.env.OPENAI_REVERSE_PROXY = 'user_provided';
-    const req = {
-      body: {
-        key: new Date(Date.now() + 10000).toISOString(),
-        endpoint: 'openAI',
-      },
-      user: {
-        id: '123',
-      },
-      app,
-    };
-    const res = {};
-    const endpointOption = {};
-
-    getUserKeyValues.mockResolvedValue({
-      apiKey: 'test',
-      baseURL: 'https://user-provided-url.com',
-    });
-
-    const result = await initializeClient({ req, res, endpointOption });
-
-    expect(result.openAIApiKey).toBe('test');
-    expect(result.client.options.reverseProxyUrl).toBe('https://user-provided-url.com');
-  });
-});

api/server/services/Endpoints/index.js

@@ -0,0 +1,75 @@
+const { Providers } = require('@librechat/agents');
+const { EModelEndpoint } = require('librechat-data-provider');
+const initAnthropic = require('~/server/services/Endpoints/anthropic/initialize');
+const getBedrockOptions = require('~/server/services/Endpoints/bedrock/options');
+const initOpenAI = require('~/server/services/Endpoints/openAI/initialize');
+const initCustom = require('~/server/services/Endpoints/custom/initialize');
+const initGoogle = require('~/server/services/Endpoints/google/initialize');
+const { getCustomEndpointConfig } = require('~/server/services/Config');
+
+/** Check if the provider is a known custom provider
+ * @param {string | undefined} [provider] - The provider string
+ * @returns {boolean} - True if the provider is a known custom provider, false otherwise
+ */
+function isKnownCustomProvider(provider) {
+  return [Providers.XAI, Providers.OLLAMA, Providers.DEEPSEEK, Providers.OPENROUTER].includes(
+    provider || '',
+  );
+}
+
+const providerConfigMap = {
+  [Providers.XAI]: initCustom,
+  [Providers.OLLAMA]: initCustom,
+  [Providers.DEEPSEEK]: initCustom,
+  [Providers.OPENROUTER]: initCustom,
+  [EModelEndpoint.openAI]: initOpenAI,
+  [EModelEndpoint.google]: initGoogle,
+  [EModelEndpoint.azureOpenAI]: initOpenAI,
+  [EModelEndpoint.anthropic]: initAnthropic,
+  [EModelEndpoint.bedrock]: getBedrockOptions,
+};
+
+/**
+ * Get the provider configuration and override endpoint based on the provider string
+ * @param {string} provider - The provider string
+ * @returns {Promise<{
+ * getOptions: Function,
+ * overrideProvider?: string,
+ * customEndpointConfig?: TEndpoint
+ * }>}
+ */
+async function getProviderConfig(provider) {
+  let getOptions = providerConfigMap[provider];
+  let overrideProvider;
+  /** @type {TEndpoint | undefined} */
+  let customEndpointConfig;
+
+  if (!getOptions && providerConfigMap[provider.toLowerCase()] != null) {
+    overrideProvider = provider.toLowerCase();
+    getOptions = providerConfigMap[overrideProvider];
+  } else if (!getOptions) {
+    customEndpointConfig = await getCustomEndpointConfig(provider);
+    if (!customEndpointConfig) {
+      throw new Error(`Provider ${provider} not supported`);
+    }
+    getOptions = initCustom;
+    overrideProvider = Providers.OPENAI;
+  }
+
+  if (isKnownCustomProvider(overrideProvider)) {
+    customEndpointConfig = await getCustomEndpointConfig(provider);
+    if (!customEndpointConfig) {
+      throw new Error(`Provider ${provider} not supported`);
+    }
+  }
+
+  return {
+    getOptions,
+    overrideProvider,
+    customEndpointConfig,
+  };
+}
+
+module.exports = {
+  getProviderConfig,
+};

api/server/services/Endpoints/openAI/initialize.js

@@ -1,11 +1,7 @@
-const {
-  ErrorTypes,
-  EModelEndpoint,
-  resolveHeaders,
-  mapModelToAzureConfig,
-} = require('librechat-data-provider');
+const { ErrorTypes, EModelEndpoint, mapModelToAzureConfig } = require('librechat-data-provider');
 const {
   isEnabled,
+  resolveHeaders,
   isUserProvided,
   getOpenAIConfig,
   getAzureCredentials,
@@ -84,7 +80,10 @@ const initializeClient = async ({
     });
 
     clientOptions.reverseProxyUrl = baseURL ?? clientOptions.reverseProxyUrl;
-    clientOptions.headers = resolveHeaders({ ...headers, ...(clientOptions.headers ?? {}) });
+    clientOptions.headers = resolveHeaders(
+      { ...headers, ...(clientOptions.headers ?? {}) },
+      req.user,
+    );
 
     clientOptions.titleConvo = azureConfig.titleConvo;
     clientOptions.titleModel = azureConfig.titleModel;
@@ -139,7 +138,7 @@ const initializeClient = async ({
   }
 
   if (optionsOnly) {
-    const modelOptions = endpointOption.model_parameters;
+    const modelOptions = endpointOption?.model_parameters ?? {};
     modelOptions.model = modelName;
     clientOptions = Object.assign({ modelOptions }, clientOptions);
     clientOptions.modelOptions.user = req.user.id;

api/server/services/Files/Local/crud.js

@@ -1,10 +1,11 @@
 const fs = require('fs');
 const path = require('path');
 const axios = require('axios');
+const { logger } = require('@librechat/data-schemas');
 const { EModelEndpoint } = require('librechat-data-provider');
+const { generateShortLivedToken } = require('~/server/services/AuthService');
 const { getBufferMetadata } = require('~/server/utils');
 const paths = require('~/config/paths');
-const { logger } = require('~/config');
 
 /**
  * Saves a file to a specified output path with a new filename.
@@ -206,7 +207,7 @@ const deleteLocalFile = async (req, file) => {
   const cleanFilepath = file.filepath.split('?')[0];
 
   if (file.embedded && process.env.RAG_API_URL) {
-    const jwtToken = req.headers.authorization.split(' ')[1];
+    const jwtToken = generateShortLivedToken(req.user.id);
     axios.delete(`${process.env.RAG_API_URL}/documents`, {
       headers: {
         Authorization: `Bearer ${jwtToken}`,

api/server/services/Files/VectorDB/crud.js

@@ -4,6 +4,7 @@ const FormData = require('form-data');
 const { logAxiosError } = require('@librechat/api');
 const { logger } = require('@librechat/data-schemas');
 const { FileSources } = require('librechat-data-provider');
+const { generateShortLivedToken } = require('~/server/services/AuthService');
 
 /**
  * Deletes a file from the vector database. This function takes a file object, constructs the full path, and
@@ -23,7 +24,8 @@ const deleteVectors = async (req, file) => {
     return;
   }
   try {
-    const jwtToken = req.headers.authorization.split(' ')[1];
+    const jwtToken = generateShortLivedToken(req.user.id);
+
     return await axios.delete(`${process.env.RAG_API_URL}/documents`, {
       headers: {
         Authorization: `Bearer ${jwtToken}`,
@@ -70,7 +72,7 @@ async function uploadVectors({ req, file, file_id, entity_id }) {
   }
 
   try {
-    const jwtToken = req.headers.authorization.split(' ')[1];
+    const jwtToken = generateShortLivedToken(req.user.id);
     const formData = new FormData();
     formData.append('file_id', file_id);
     formData.append('file', fs.createReadStream(file.path));

api/server/services/Files/process.js

@@ -55,7 +55,9 @@ const processFiles = async (files, fileIds) => {
   }
 
   if (!fileIds) {
-    return await Promise.all(promises);
+    const results = await Promise.all(promises);
+    // Filter out null results from failed updateFileUsage calls
+    return results.filter((result) => result != null);
   }
 
   for (let file_id of fileIds) {
@@ -67,7 +69,9 @@ const processFiles = async (files, fileIds) => {
   }
 
   // TODO: calculate token cost when image is first uploaded
-  return await Promise.all(promises);
+  const results = await Promise.all(promises);
+  // Filter out null results from failed updateFileUsage calls
+  return results.filter((result) => result != null);
 };
 
 /**

api/server/services/Files/processFiles.test.js

@@ -0,0 +1,208 @@
+// Mock the updateFileUsage function before importing the actual processFiles
+jest.mock('~/models/File', () => ({
+  updateFileUsage: jest.fn(),
+}));
+
+// Mock winston and logger configuration to avoid dependency issues
+jest.mock('~/config', () => ({
+  logger: {
+    info: jest.fn(),
+    warn: jest.fn(),
+    debug: jest.fn(),
+    error: jest.fn(),
+  },
+}));
+
+// Mock all other dependencies that might cause issues
+jest.mock('librechat-data-provider', () => ({
+  isUUID: { parse: jest.fn() },
+  megabyte: 1024 * 1024,
+  FileContext: { message_attachment: 'message_attachment' },
+  FileSources: { local: 'local' },
+  EModelEndpoint: { assistants: 'assistants' },
+  EToolResources: { file_search: 'file_search' },
+  mergeFileConfig: jest.fn(),
+  removeNullishValues: jest.fn((obj) => obj),
+  isAssistantsEndpoint: jest.fn(),
+}));
+
+jest.mock('~/server/services/Files/images', () => ({
+  convertImage: jest.fn(),
+  resizeAndConvert: jest.fn(),
+  resizeImageBuffer: jest.fn(),
+}));
+
+jest.mock('~/server/controllers/assistants/v2', () => ({
+  addResourceFileId: jest.fn(),
+  deleteResourceFileId: jest.fn(),
+}));
+
+jest.mock('~/models/Agent', () => ({
+  addAgentResourceFile: jest.fn(),
+  removeAgentResourceFiles: jest.fn(),
+}));
+
+jest.mock('~/server/controllers/assistants/helpers', () => ({
+  getOpenAIClient: jest.fn(),
+}));
+
+jest.mock('~/server/services/Tools/credentials', () => ({
+  loadAuthValues: jest.fn(),
+}));
+
+jest.mock('~/server/services/Config', () => ({
+  checkCapability: jest.fn(),
+}));
+
+jest.mock('~/server/utils/queue', () => ({
+  LB_QueueAsyncCall: jest.fn(),
+}));
+
+jest.mock('./strategies', () => ({
+  getStrategyFunctions: jest.fn(),
+}));
+
+jest.mock('~/server/utils', () => ({
+  determineFileType: jest.fn(),
+}));
+
+// Import the actual processFiles function after all mocks are set up
+const { processFiles } = require('./process');
+const { updateFileUsage } = require('~/models/File');
+
+describe('processFiles', () => {
+  beforeEach(() => {
+    jest.clearAllMocks();
+  });
+
+  describe('null filtering functionality', () => {
+    it('should filter out null results from updateFileUsage when files do not exist', async () => {
+      const mockFiles = [
+        { file_id: 'existing-file-1' },
+        { file_id: 'non-existent-file' },
+        { file_id: 'existing-file-2' },
+      ];
+
+      // Mock updateFileUsage to return null for non-existent files
+      updateFileUsage.mockImplementation(({ file_id }) => {
+        if (file_id === 'non-existent-file') {
+          return Promise.resolve(null); // Simulate file not found in the database
+        }
+        return Promise.resolve({ file_id, usage: 1 });
+      });
+
+      const result = await processFiles(mockFiles);
+
+      expect(updateFileUsage).toHaveBeenCalledTimes(3);
+      expect(result).toEqual([
+        { file_id: 'existing-file-1', usage: 1 },
+        { file_id: 'existing-file-2', usage: 1 },
+      ]);
+
+      // Critical test - ensure no null values in result
+      expect(result).not.toContain(null);
+      expect(result).not.toContain(undefined);
+      expect(result.length).toBe(2); // Only valid files should be returned
+    });
+
+    it('should return empty array when all updateFileUsage calls return null', async () => {
+      const mockFiles = [{ file_id: 'non-existent-1' }, { file_id: 'non-existent-2' }];
+
+      // All updateFileUsage calls return null
+      updateFileUsage.mockResolvedValue(null);
+
+      const result = await processFiles(mockFiles);
+
+      expect(updateFileUsage).toHaveBeenCalledTimes(2);
+      expect(result).toEqual([]);
+      expect(result).not.toContain(null);
+      expect(result.length).toBe(0);
+    });
+
+    it('should work correctly when all files exist', async () => {
+      const mockFiles = [{ file_id: 'file-1' }, { file_id: 'file-2' }];
+
+      updateFileUsage.mockImplementation(({ file_id }) => {
+        return Promise.resolve({ file_id, usage: 1 });
+      });
+
+      const result = await processFiles(mockFiles);
+
+      expect(result).toEqual([
+        { file_id: 'file-1', usage: 1 },
+        { file_id: 'file-2', usage: 1 },
+      ]);
+      expect(result).not.toContain(null);
+      expect(result.length).toBe(2);
+    });
+
+    it('should handle fileIds parameter and filter nulls correctly', async () => {
+      const mockFiles = [{ file_id: 'file-1' }];
+      const mockFileIds = ['file-2', 'non-existent-file'];
+
+      updateFileUsage.mockImplementation(({ file_id }) => {
+        if (file_id === 'non-existent-file') {
+          return Promise.resolve(null);
+        }
+        return Promise.resolve({ file_id, usage: 1 });
+      });
+
+      const result = await processFiles(mockFiles, mockFileIds);
+
+      expect(result).toEqual([
+        { file_id: 'file-1', usage: 1 },
+        { file_id: 'file-2', usage: 1 },
+      ]);
+      expect(result).not.toContain(null);
+      expect(result).not.toContain(undefined);
+      expect(result.length).toBe(2);
+    });
+
+    it('should handle duplicate file_ids correctly', async () => {
+      const mockFiles = [
+        { file_id: 'duplicate-file' },
+        { file_id: 'duplicate-file' }, // Duplicate should be ignored
+        { file_id: 'unique-file' },
+      ];
+
+      updateFileUsage.mockImplementation(({ file_id }) => {
+        return Promise.resolve({ file_id, usage: 1 });
+      });
+
+      const result = await processFiles(mockFiles);
+
+      // Should only call updateFileUsage twice (duplicate ignored)
+      expect(updateFileUsage).toHaveBeenCalledTimes(2);
+      expect(result).toEqual([
+        { file_id: 'duplicate-file', usage: 1 },
+        { file_id: 'unique-file', usage: 1 },
+      ]);
+      expect(result.length).toBe(2);
+    });
+  });
+
+  describe('edge cases', () => {
+    it('should handle empty files array', async () => {
+      const result = await processFiles([]);
+      expect(result).toEqual([]);
+      expect(updateFileUsage).not.toHaveBeenCalled();
+    });
+
+    it('should handle mixed null and undefined returns from updateFileUsage', async () => {
+      const mockFiles = [{ file_id: 'file-1' }, { file_id: 'file-2' }, { file_id: 'file-3' }];
+
+      updateFileUsage.mockImplementation(({ file_id }) => {
+        if (file_id === 'file-1') return Promise.resolve(null);
+        if (file_id === 'file-2') return Promise.resolve(undefined);
+        return Promise.resolve({ file_id, usage: 1 });
+      });
+
+      const result = await processFiles(mockFiles);
+
+      expect(result).toEqual([{ file_id: 'file-3', usage: 1 }]);
+      expect(result).not.toContain(null);
+      expect(result).not.toContain(undefined);
+      expect(result.length).toBe(1);
+    });
+  });
+});

api/server/services/Files/strategies.js

@@ -1,5 +1,9 @@
 const { FileSources } = require('librechat-data-provider');
-const { uploadMistralOCR, uploadAzureMistralOCR } = require('@librechat/api');
+const {
+  uploadMistralOCR,
+  uploadAzureMistralOCR,
+  uploadGoogleVertexMistralOCR,
+} = require('@librechat/api');
 const {
   getFirebaseURL,
   prepareImageURL,
@@ -222,6 +226,26 @@ const azureMistralOCRStrategy = () => ({
   handleFileUpload: uploadAzureMistralOCR,
 });
 
+const vertexMistralOCRStrategy = () => ({
+  /** @type {typeof saveFileFromURL | null} */
+  saveURL: null,
+  /** @type {typeof getLocalFileURL | null} */
+  getFileURL: null,
+  /** @type {typeof saveLocalBuffer | null} */
+  saveBuffer: null,
+  /** @type {typeof processLocalAvatar | null} */
+  processAvatar: null,
+  /** @type {typeof uploadLocalImage | null} */
+  handleImageUpload: null,
+  /** @type {typeof prepareImagesLocal | null} */
+  prepareImagePayload: null,
+  /** @type {typeof deleteLocalFile | null} */
+  deleteFile: null,
+  /** @type {typeof getLocalFileStream | null} */
+  getDownloadStream: null,
+  handleFileUpload: uploadGoogleVertexMistralOCR,
+});
+
 // Strategy Selector
 const getStrategyFunctions = (fileSource) => {
   if (fileSource === FileSources.firebase) {
@@ -244,6 +268,8 @@ const getStrategyFunctions = (fileSource) => {
     return mistralOCRStrategy();
   } else if (fileSource === FileSources.azure_mistral_ocr) {
     return azureMistralOCRStrategy();
+  } else if (fileSource === FileSources.vertexai_mistral_ocr) {
+    return vertexMistralOCRStrategy();
   } else {
     throw new Error('Invalid file source');
   }

api/server/services/Runs/StreamRunManager.js

@@ -1,3 +1,6 @@
+const { sleep } = require('@librechat/agents');
+const { sendEvent } = require('@librechat/api');
+const { logger } = require('@librechat/data-schemas');
 const {
   Constants,
   StepTypes,
@@ -8,9 +11,8 @@ const {
 } = require('librechat-data-provider');
 const { retrieveAndProcessFile } = require('~/server/services/Files/process');
 const { processRequiredActions } = require('~/server/services/ToolService');
-const { createOnProgress, sendMessage, sleep } = require('~/server/utils');
 const { processMessages } = require('~/server/services/Threads');
-const { logger } = require('~/config');
+const { createOnProgress } = require('~/server/utils');
 
 /**
  * Implements the StreamRunManager functionality for managing the streaming
@@ -126,7 +128,7 @@ class StreamRunManager {
       conversationId: this.finalMessage.conversationId,
     };
 
-    sendMessage(this.res, contentData);
+    sendEvent(this.res, contentData);
   }
 
   /* <------------------ Misc. Helpers ------------------> */
@@ -302,7 +304,7 @@ class StreamRunManager {
 
           for (const d of delta[key]) {
             if (typeof d === 'object' && !Object.prototype.hasOwnProperty.call(d, 'index')) {
-              logger.warn('Expected an object with an \'index\' for array updates but got:', d);
+              logger.warn("Expected an object with an 'index' for array updates but got:", d);
               continue;
             }
 

api/server/services/initializeMCP.js

@@ -1,9 +1,9 @@
 const { logger } = require('@librechat/data-schemas');
-const { CacheKeys, processMCPEnv } = require('librechat-data-provider');
+const { CacheKeys } = require('librechat-data-provider');
+const { findToken, updateToken, createToken, deleteTokens } = require('~/models');
 const { getMCPManager, getFlowStateManager } = require('~/config');
 const { getCachedTools, setCachedTools } = require('./Config');
 const { getLogStores } = require('~/cache');
-const { findToken, updateToken, createToken, deleteTokens } = require('~/models');
 
 /**
  * Initialize MCP servers
@@ -30,7 +30,6 @@ async function initializeMCP(app) {
         createToken,
         deleteTokens,
       },
-      processMCPEnv,
     });
 
     delete app.locals.mcpConfig;

api/server/services/start/interface.js

@@ -41,6 +41,7 @@ async function loadDefaultInterface(config, configDefaults, roleName = SystemRol
     sidePanel: interfaceConfig?.sidePanel ?? defaults.sidePanel,
     privacyPolicy: interfaceConfig?.privacyPolicy ?? defaults.privacyPolicy,
     termsOfService: interfaceConfig?.termsOfService ?? defaults.termsOfService,
+    mcpServers: interfaceConfig?.mcpServers ?? defaults.mcpServers,
     bookmarks: interfaceConfig?.bookmarks ?? defaults.bookmarks,
     memories: shouldDisableMemories ? false : (interfaceConfig?.memories ?? defaults.memories),
     prompts: interfaceConfig?.prompts ?? defaults.prompts,