chore: remove test code

- Modified the user creation process to incorporate balance configuration retrieved from the new getBalanceConfig function.
- Adjusted imports for user methods to streamline the code structure.

.env.example

@@ -58,7 +58,7 @@ DEBUG_CONSOLE=false
 #                     Endpoints                     #
 #===================================================#
 
-# ENDPOINTS=openAI,assistants,azureOpenAI,google,anthropic
+# ENDPOINTS=openAI,assistants,azureOpenAI,google,gptPlugins,anthropic
 
 PROXY=
 
@@ -142,10 +142,10 @@ GOOGLE_KEY=user_provided
 # GOOGLE_AUTH_HEADER=true
 
 # Gemini API (AI Studio)
-# 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
+# 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
 
 # Vertex AI
-# 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_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_TITLE_MODEL=gemini-2.0-flash-lite-001
 
@@ -515,18 +515,6 @@ EMAIL_PASSWORD=
 EMAIL_FROM_NAME=
 EMAIL_FROM=noreply@librechat.ai
 
-#========================#
-#      Mailgun API       #
-#========================#
-
-# MAILGUN_API_KEY=your-mailgun-api-key
-# MAILGUN_DOMAIN=mg.yourdomain.com
-# EMAIL_FROM=noreply@yourdomain.com
-# EMAIL_FROM_NAME="LibreChat"
-
-# # Optional: For EU region
-# MAILGUN_HOST=https://api.eu.mailgun.net
-
 #========================#
 # Firebase CDN           #
 #========================#
@@ -657,4 +645,4 @@ 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

.github/CONTRIBUTING.md

@@ -30,8 +30,8 @@ Project maintainers have the right and responsibility to remove, edit, or reject
 2. Install typescript globally: `npm i -g typescript`.
 3. Run `npm ci` to install dependencies.
 4. Build the data provider: `npm run build:data-provider`.
-5. Build data schemas: `npm run build:data-schemas`.
-6. Build API methods: `npm run build:api`.
+5. Build MCP: `npm run build:mcp`.
+6. Build data schemas: `npm run build:data-schemas`.
 7. Setup and run unit tests:
     - Copy `.env.test`: `cp api/test/.env.test.example api/test/.env.test`.
     - Run backend unit tests: `npm run test:api`.

.github/workflows/backend-review.yml

@@ -7,7 +7,6 @@ on:
       - release/*
     paths:
       - 'api/**'
-      - 'packages/api/**'
 jobs:
   tests_Backend:
     name: Run Backend unit tests
@@ -37,12 +36,12 @@ jobs:
       - name: Install Data Provider Package
         run: npm run build:data-provider
 
+      - name: Install MCP Package
+        run: npm run build:mcp
+
       - name: Install Data Schemas Package
         run: npm run build:data-schemas
 
-      - name: Install API Package
-        run: npm run build:api
-
       - name: Create empty auth.json file
         run: |
           mkdir -p api/data
@@ -67,8 +66,5 @@ jobs:
       - name: Run librechat-data-provider unit tests
         run: cd packages/data-provider && npm run test:ci
 
-      - name: Run @librechat/data-schemas unit tests
-        run: cd packages/data-schemas && npm run test:ci
-
-      - name: Run @librechat/api unit tests
-        run: cd packages/api && npm run test:ci
+      - name: Run librechat-mcp unit tests
+        run: cd packages/mcp && npm run test:ci

.github/workflows/deploy-dev.yml

@@ -2,7 +2,7 @@ name: Update Test Server
 
 on:
   workflow_run:
-    workflows: ["Docker Dev Branch Images Build"]
+    workflows: ["Docker Dev Images Build"]
     types:
       - completed
   workflow_dispatch:
@@ -12,8 +12,7 @@ jobs:
     runs-on: ubuntu-latest
     if: |
       github.repository == 'danny-avila/LibreChat' &&
-      (github.event_name == 'workflow_dispatch' || 
-       (github.event.workflow_run.conclusion == 'success' && github.event.workflow_run.head_branch == 'dev'))
+      (github.event_name == 'workflow_dispatch' || github.event.workflow_run.conclusion == 'success')
     steps:
     - name: Checkout repository
       uses: actions/checkout@v4
@@ -30,17 +29,13 @@ jobs:
         DO_USER: ${{ secrets.DO_USER }}
       run: |
         ssh -o StrictHostKeyChecking=no ${DO_USER}@${DO_HOST} << EOF
-        sudo -i -u danny bash << 'EEOF'
+        sudo -i -u danny bash << EEOF
         cd ~/LibreChat && \
         git fetch origin main && \
-        sudo npm run stop:deployed && \
-        sudo docker images --format "{{.Repository}}:{{.ID}}" | grep -E "lc-dev|librechat" | cut -d: -f2 | xargs -r sudo docker rmi -f || true && \
-        sudo npm run update:deployed && \
-        git checkout dev && \
-        git pull origin dev && \
+        npm run update:deployed && \
         git checkout do-deploy && \
-        git rebase dev && \
-        sudo npm run start:deployed && \
+        git rebase main && \
+        npm run start:deployed && \
         echo "Update completed. Application should be running now."
         EEOF
         EOF

.github/workflows/dev-branch-images.yml

@@ -1,72 +0,0 @@
-name: Docker Dev Branch Images Build
-
-on:
-  workflow_dispatch:
-  push:
-    branches:
-      - dev
-    paths:
-      - 'api/**'
-      - 'client/**'
-      - 'packages/**'
-
-jobs:
-  build:
-    runs-on: ubuntu-latest
-    strategy:
-      matrix:
-        include:
-          - target: api-build
-            file: Dockerfile.multi
-            image_name: lc-dev-api
-          - target: node
-            file: Dockerfile
-            image_name: lc-dev
-
-    steps:
-      # Check out the repository
-      - name: Checkout
-        uses: actions/checkout@v4
-
-      # Set up QEMU
-      - name: Set up QEMU
-        uses: docker/setup-qemu-action@v3
-
-      # Set up Docker Buildx
-      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v3
-
-      # Log in to GitHub Container Registry
-      - name: Log in to GitHub Container Registry
-        uses: docker/login-action@v2
-        with:
-          registry: ghcr.io
-          username: ${{ github.actor }}
-          password: ${{ secrets.GITHUB_TOKEN }}
-
-      # Login to Docker Hub
-      - name: Login to Docker Hub
-        uses: docker/login-action@v3
-        with:
-          username: ${{ secrets.DOCKERHUB_USERNAME }}
-          password: ${{ secrets.DOCKERHUB_TOKEN }}
-
-      # Prepare the environment
-      - name: Prepare environment
-        run: |
-          cp .env.example .env
-
-      # Build and push Docker images for each target
-      - name: Build and push Docker images
-        uses: docker/build-push-action@v5
-        with:
-          context: .
-          file: ${{ matrix.file }}
-          push: true
-          tags: |
-            ghcr.io/${{ github.repository_owner }}/${{ matrix.image_name }}:${{ github.sha }}
-            ghcr.io/${{ github.repository_owner }}/${{ matrix.image_name }}:latest
-            ${{ secrets.DOCKERHUB_USERNAME }}/${{ matrix.image_name }}:${{ github.sha }}
-            ${{ secrets.DOCKERHUB_USERNAME }}/${{ matrix.image_name }}:latest
-          platforms: linux/amd64,linux/arm64
-          target: ${{ matrix.target }} 

.github/workflows/i18n-unused-keys.yml

@@ -5,13 +5,12 @@ on:
     paths:
       - "client/src/**"
       - "api/**"
-      - "packages/data-provider/src/**"
 
 jobs:
   detect-unused-i18n-keys:
     runs-on: ubuntu-latest
     permissions:
-      pull-requests: write
+      pull-requests: write  # Required for posting PR comments
     steps:
       - name: Checkout repository
         uses: actions/checkout@v3

.github/workflows/unused-packages.yml

@@ -98,8 +98,6 @@ jobs:
             cd client
             UNUSED=$(depcheck --json | jq -r '.dependencies | join("\n")' || echo "")
             UNUSED=$(comm -23 <(echo "$UNUSED" | sort) <(cat ../client_used_deps.txt ../client_used_code.txt | sort) || echo "")
-            # Filter out false positives
-            UNUSED=$(echo "$UNUSED" | grep -v "^micromark-extension-llm-math$" || echo "")
             echo "CLIENT_UNUSED<<EOF" >> $GITHUB_ENV
             echo "$UNUSED" >> $GITHUB_ENV
             echo "EOF" >> $GITHUB_ENV

.gitignore

@@ -55,7 +55,6 @@ bower_components/
 # AI
 .clineignore
 .cursor
-.aider*
 
 # Floobits
 .floo

Dockerfile.multi

@@ -14,7 +14,7 @@ RUN npm config set fetch-retry-maxtimeout 600000 && \
     npm config set fetch-retry-mintimeout 15000
 COPY package*.json ./
 COPY packages/data-provider/package*.json ./packages/data-provider/
-COPY packages/api/package*.json ./packages/api/
+COPY packages/mcp/package*.json ./packages/mcp/
 COPY packages/data-schemas/package*.json ./packages/data-schemas/
 COPY client/package*.json ./client/
 COPY api/package*.json ./api/
@@ -24,27 +24,26 @@ FROM base-min AS base
 WORKDIR /app
 RUN npm ci
 
-# Build `data-provider` package
+# Build data-provider
 FROM base AS data-provider-build
 WORKDIR /app/packages/data-provider
 COPY packages/data-provider ./
 RUN npm run build
 
-# Build `data-schemas` package
+# Build mcp package
+FROM base AS mcp-build
+WORKDIR /app/packages/mcp
+COPY packages/mcp ./
+COPY --from=data-provider-build /app/packages/data-provider/dist /app/packages/data-provider/dist
+RUN npm run build
+
+# Build data-schemas
 FROM base AS data-schemas-build
 WORKDIR /app/packages/data-schemas
 COPY packages/data-schemas ./
 COPY --from=data-provider-build /app/packages/data-provider/dist /app/packages/data-provider/dist
 RUN npm run build
 
-# Build `api` package
-FROM base AS api-package-build
-WORKDIR /app/packages/api
-COPY packages/api ./
-COPY --from=data-provider-build /app/packages/data-provider/dist /app/packages/data-provider/dist
-COPY --from=data-schemas-build /app/packages/data-schemas/dist /app/packages/data-schemas/dist
-RUN npm run build
-
 # Client build
 FROM base AS client-build
 WORKDIR /app/client
@@ -64,8 +63,8 @@ RUN npm ci --omit=dev
 COPY api ./api
 COPY config ./config
 COPY --from=data-provider-build /app/packages/data-provider/dist ./packages/data-provider/dist
+COPY --from=mcp-build /app/packages/mcp/dist ./packages/mcp/dist
 COPY --from=data-schemas-build /app/packages/data-schemas/dist ./packages/data-schemas/dist
-COPY --from=api-package-build /app/packages/api/dist ./packages/api/dist
 COPY --from=client-build /app/client/dist ./client/dist
 WORKDIR /app/api
 EXPOSE 3080

README.md

@@ -150,8 +150,8 @@ Click on the thumbnail to open the video☝️
 
 **Other:**
   - **Website:** [librechat.ai](https://librechat.ai)
-  - **Documentation:** [librechat.ai/docs](https://librechat.ai/docs)
-  - **Blog:** [librechat.ai/blog](https://librechat.ai/blog)
+  - **Documentation:** [docs.librechat.ai](https://docs.librechat.ai)
+  - **Blog:** [blog.librechat.ai](https://blog.librechat.ai)
 
 ---
 

api/app/clients/AnthropicClient.js

@@ -10,7 +10,6 @@ const {
   validateVisionModel,
 } = require('librechat-data-provider');
 const { SplitStreamHandler: _Handler } = require('@librechat/agents');
-const { Tokenizer, createFetch, createStreamEventHandlers } = require('@librechat/api');
 const {
   truncateText,
   formatMessage,
@@ -27,6 +26,8 @@ const {
 const { getModelMaxTokens, getModelMaxOutputTokens, matchModelName } = require('~/utils');
 const { spendTokens, spendStructuredTokens } = require('~/models/spendTokens');
 const { encodeAndFormat } = require('~/server/services/Files/images/encode');
+const { createFetch, createStreamEventHandlers } = require('./generators');
+const Tokenizer = require('~/server/services/Tokenizer');
 const { sleep } = require('~/server/utils');
 const BaseClient = require('./BaseClient');
 const { logger } = require('~/config');
@@ -190,11 +191,10 @@ class AnthropicClient extends BaseClient {
         reverseProxyUrl: this.options.reverseProxyUrl,
       }),
       apiKey: this.apiKey,
-      fetchOptions: {},
     };
 
     if (this.options.proxy) {
-      options.fetchOptions.agent = new HttpsProxyAgent(this.options.proxy);
+      options.httpAgent = new HttpsProxyAgent(this.options.proxy);
     }
 
     if (this.options.reverseProxyUrl) {

api/app/clients/BaseClient.js

@@ -792,8 +792,7 @@ class BaseClient {
 
     userMessage.tokenCount = userMessageTokenCount;
     /*
-      Note: `AgentController` saves the user message if not saved here
-      (noted by `savedMessageIds`), so we update the count of its `userMessage` reference
+      Note: `AskController` saves the user message, so we update the count of its `userMessage` reference
     */
     if (typeof opts?.getReqData === 'function') {
       opts.getReqData({
@@ -802,8 +801,7 @@ class BaseClient {
     }
     /*
       Note: we update the user message to be sure it gets the calculated token count;
-      though `AgentController` saves the user message if not saved here
-      (noted by `savedMessageIds`), EditController does not
+      though `AskController` saves the user message, EditController does not
     */
     await userMessagePromise;
     await this.updateMessageInDatabase({

api/app/clients/ChatGPTClient.js

@@ -0,0 +1,804 @@
+const { Keyv } = require('keyv');
+const crypto = require('crypto');
+const { CohereClient } = require('cohere-ai');
+const { fetchEventSource } = require('@waylaidwanderer/fetch-event-source');
+const { encoding_for_model: encodingForModel, get_encoding: getEncoding } = require('tiktoken');
+const {
+  ImageDetail,
+  EModelEndpoint,
+  resolveHeaders,
+  CohereConstants,
+  mapModelToAzureConfig,
+} = require('librechat-data-provider');
+const { extractBaseURL, constructAzureURL, genAzureChatCompletion } = require('~/utils');
+const { createContextHandlers } = require('./prompts');
+const { createCoherePayload } = require('./llm');
+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 +1,6 @@
 const { google } = require('googleapis');
 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');
@@ -12,14 +11,15 @@ 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 Tokenizer = require('~/server/services/Tokenizer');
 const { spendTokens } = require('~/models/spendTokens');
 const { getModelMaxTokens } = require('~/utils');
 const { sleep } = require('~/server/utils');
@@ -34,8 +34,7 @@ const BaseClient = require('./BaseClient');
 
 const loc = process.env.GOOGLE_LOC || 'us-central1';
 const publisher = 'google';
-const endpointPrefix =
-  loc === 'global' ? 'aiplatform.googleapis.com' : `${loc}-aiplatform.googleapis.com`;
+const endpointPrefix = `${loc}-aiplatform.googleapis.com`;
 
 const settings = endpointSettings[EModelEndpoint.google];
 const EXCLUDED_GENAI_MODELS = /gemini-(?:1\.0|1-0|pro)/;
@@ -166,16 +165,6 @@ 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({
@@ -247,11 +236,11 @@ class GoogleClient extends BaseClient {
       msg.content = (
         !Array.isArray(msg.content)
           ? [
-              {
-                type: ContentTypes.TEXT,
-                [ContentTypes.TEXT]: msg.content,
-              },
-            ]
+            {
+              type: ContentTypes.TEXT,
+              [ContentTypes.TEXT]: msg.content,
+            },
+          ]
           : msg.content
       ).concat(message.image_urls);
 

api/app/clients/OllamaClient.js

@@ -1,11 +1,10 @@
 const { z } = require('zod');
 const axios = require('axios');
 const { Ollama } = require('ollama');
-const { sleep } = require('@librechat/agents');
-const { logAxiosError } = require('@librechat/api');
-const { logger } = require('@librechat/data-schemas');
 const { Constants } = require('librechat-data-provider');
-const { deriveBaseURL } = require('~/utils');
+const { deriveBaseURL, logAxiosError } = require('~/utils');
+const { sleep } = require('~/server/utils');
+const { logger } = require('~/config');
 
 const ollamaPayloadSchema = z.object({
   mirostat: z.number().optional(),
@@ -68,7 +67,7 @@ class OllamaClient {
       return models;
     } catch (error) {
       const logMessage =
-        "Failed to fetch models from Ollama API. If you are not using Ollama directly, and instead, through some aggregator or reverse proxy that handles fetching via OpenAI spec, ensure the name of the endpoint doesn't start with `ollama` (case-insensitive).";
+        'Failed to fetch models from Ollama API. If you are not using Ollama directly, and instead, through some aggregator or reverse proxy that handles fetching via OpenAI spec, ensure the name of the endpoint doesn\'t start with `ollama` (case-insensitive).';
       logAxiosError({ message: logMessage, error });
       return [];
     }

api/app/clients/OpenAIClient.js

@@ -1,21 +1,13 @@
 const { OllamaClient } = require('./OllamaClient');
 const { HttpsProxyAgent } = require('https-proxy-agent');
 const { SplitStreamHandler, CustomOpenAIClient: OpenAI } = require('@librechat/agents');
-const {
-  isEnabled,
-  Tokenizer,
-  createFetch,
-  resolveHeaders,
-  constructAzureURL,
-  genAzureChatCompletion,
-  createStreamEventHandlers,
-} = require('@librechat/api');
 const {
   Constants,
   ImageDetail,
   ContentTypes,
   parseTextParts,
   EModelEndpoint,
+  resolveHeaders,
   KnownEndpoints,
   openAISettings,
   ImageDetailCost,
@@ -24,6 +16,13 @@ const {
   validateVisionModel,
   mapModelToAzureConfig,
 } = require('librechat-data-provider');
+const {
+  extractBaseURL,
+  constructAzureURL,
+  getModelMaxTokens,
+  genAzureChatCompletion,
+  getModelMaxOutputTokens,
+} = require('~/utils');
 const {
   truncateText,
   formatMessage,
@@ -31,12 +30,14 @@ const {
   titleInstruction,
   createContextHandlers,
 } = require('./prompts');
-const { extractBaseURL, getModelMaxTokens, getModelMaxOutputTokens } = require('~/utils');
 const { encodeAndFormat } = require('~/server/services/Files/images/encode');
-const { addSpaceIfNeeded, sleep } = require('~/server/utils');
+const { createFetch, createStreamEventHandlers } = require('./generators');
+const { addSpaceIfNeeded, isEnabled, sleep } = require('~/server/utils');
+const Tokenizer = require('~/server/services/Tokenizer');
 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');
@@ -46,6 +47,12 @@ 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';
@@ -372,12 +379,23 @@ class OpenAIClient extends BaseClient {
     return files;
   }
 
-  async buildMessages(messages, parentMessageId, { promptPrefix = null }, opts) {
+  async buildMessages(
+    messages,
+    parentMessageId,
+    { isChatCompletion = false, 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;
@@ -1141,7 +1159,6 @@ ${convo}
       logger.debug('[OpenAIClient] chatCompletion', { baseURL, modelOptions });
       const opts = {
         baseURL,
-        fetchOptions: {},
       };
 
       if (this.useOpenRouter) {
@@ -1160,7 +1177,7 @@ ${convo}
       }
 
       if (this.options.proxy) {
-        opts.fetchOptions.agent = new HttpsProxyAgent(this.options.proxy);
+        opts.httpAgent = new HttpsProxyAgent(this.options.proxy);
       }
 
       /** @type {TAzureConfig | undefined} */
@@ -1378,7 +1395,7 @@ ${convo}
           ...modelOptions,
           stream: true,
         };
-        const stream = await openai.chat.completions
+        const stream = await openai.beta.chat.completions
           .stream(params)
           .on('abort', () => {
             /* Do nothing here */

api/app/clients/PluginsClient.js

@@ -0,0 +1,542 @@
+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/generators.js

@@ -0,0 +1,71 @@
+const fetch = require('node-fetch');
+const { GraphEvents } = require('@librechat/agents');
+const { logger, sendEvent } = require('~/config');
+const { sleep } = require('~/server/utils');
+
+/**
+ * Makes a function to make HTTP request and logs the process.
+ * @param {Object} params
+ * @param {boolean} [params.directEndpoint] - Whether to use a direct endpoint.
+ * @param {string} [params.reverseProxyUrl] - The reverse proxy URL to use for the request.
+ * @returns {Promise<Response>} - A promise that resolves to the response of the fetch request.
+ */
+function createFetch({ directEndpoint = false, reverseProxyUrl = '' }) {
+  /**
+   * Makes an HTTP request and logs the process.
+   * @param {RequestInfo} url - The URL to make the request to. Can be a string or a Request object.
+   * @param {RequestInit} [init] - Optional init options for the request.
+   * @returns {Promise<Response>} - A promise that resolves to the response of the fetch request.
+   */
+  return async (_url, init) => {
+    let url = _url;
+    if (directEndpoint) {
+      url = reverseProxyUrl;
+    }
+    logger.debug(`Making request to ${url}`);
+    if (typeof Bun !== 'undefined') {
+      return await fetch(url, init);
+    }
+    return await fetch(url, init);
+  };
+}
+
+// Add this at the module level outside the class
+/**
+ * Creates event handlers for stream events that don't capture client references
+ * @param {Object} res - The response object to send events to
+ * @returns {Object} Object containing handler functions
+ */
+function createStreamEventHandlers(res) {
+  return {
+    [GraphEvents.ON_RUN_STEP]: (event) => {
+      if (res) {
+        sendEvent(res, event);
+      }
+    },
+    [GraphEvents.ON_MESSAGE_DELTA]: (event) => {
+      if (res) {
+        sendEvent(res, event);
+      }
+    },
+    [GraphEvents.ON_REASONING_DELTA]: (event) => {
+      if (res) {
+        sendEvent(res, event);
+      }
+    },
+  };
+}
+
+function createHandleLLMNewToken(streamRate) {
+  return async () => {
+    if (streamRate) {
+      await sleep(streamRate);
+    }
+  };
+}
+
+module.exports = {
+  createFetch,
+  createHandleLLMNewToken,
+  createStreamEventHandlers,
+};

api/app/clients/index.js

@@ -1,11 +1,15 @@
+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/llm/createLLM.js

@@ -1,5 +1,6 @@
 const { ChatOpenAI } = require('@langchain/openai');
-const { isEnabled, sanitizeModelName, constructAzureURL } = require('@librechat/api');
+const { sanitizeModelName, constructAzureURL } = require('~/utils');
+const { isEnabled } = require('~/server/utils');
 
 /**
  * Creates a new instance of a language model (LLM) for chat interactions.

api/app/clients/prompts/createContextHandlers.js

@@ -96,35 +96,35 @@ 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/AnthropicClient.test.js

@@ -309,7 +309,7 @@ describe('AnthropicClient', () => {
       };
       client.setOptions({ modelOptions, promptCache: true });
       const anthropicClient = client.getClient(modelOptions);
-      expect(anthropicClient._options.defaultHeaders).toBeUndefined();
+      expect(anthropicClient.defaultHeaders).not.toHaveProperty('anthropic-beta');
     });
 
     it('should not add beta header for other models', () => {
@@ -320,7 +320,7 @@ describe('AnthropicClient', () => {
         },
       });
       const anthropicClient = client.getClient();
-      expect(anthropicClient._options.defaultHeaders).toBeUndefined();
+      expect(anthropicClient.defaultHeaders).not.toHaveProperty('anthropic-beta');
     });
   });
 

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

@@ -33,9 +33,7 @@ jest.mock('~/models', () => ({
 const { getConvo, saveConvo } = require('~/models');
 
 jest.mock('@librechat/agents', () => {
-  const { Providers } = jest.requireActual('@librechat/agents');
   return {
-    Providers,
     ChatOpenAI: jest.fn().mockImplementation(() => {
       return {};
     }),

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

@@ -531,6 +531,44 @@ 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

@@ -0,0 +1,314 @@
+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/dynamic/OpenAPIPlugin.js

@@ -0,0 +1,184 @@
+require('dotenv').config();
+const fs = require('fs');
+const { z } = require('zod');
+const path = require('path');
+const yaml = require('js-yaml');
+const { createOpenAPIChain } = require('langchain/chains');
+const { DynamicStructuredTool } = require('@langchain/core/tools');
+const { ChatPromptTemplate, HumanMessagePromptTemplate } = require('@langchain/core/prompts');
+const { logger } = require('~/config');
+
+function addLinePrefix(text, prefix = '// ') {
+  return text
+    .split('\n')
+    .map((line) => prefix + line)
+    .join('\n');
+}
+
+function createPrompt(name, functions) {
+  const prefix = `// The ${name} tool has the following functions. Determine the desired or most optimal function for the user's query:`;
+  const functionDescriptions = functions
+    .map((func) => `// - ${func.name}: ${func.description}`)
+    .join('\n');
+  return `${prefix}\n${functionDescriptions}
+// You are an expert manager and scrum master. You must provide a detailed intent to better execute the function.
+// Always format as such: {{"func": "function_name", "intent": "intent and expected result"}}`;
+}
+
+const AuthBearer = z
+  .object({
+    type: z.string().includes('service_http'),
+    authorization_type: z.string().includes('bearer'),
+    verification_tokens: z.object({
+      openai: z.string(),
+    }),
+  })
+  .catch(() => false);
+
+const AuthDefinition = z
+  .object({
+    type: z.string(),
+    authorization_type: z.string(),
+    verification_tokens: z.object({
+      openai: z.string(),
+    }),
+  })
+  .catch(() => false);
+
+async function readSpecFile(filePath) {
+  try {
+    const fileContents = await fs.promises.readFile(filePath, 'utf8');
+    if (path.extname(filePath) === '.json') {
+      return JSON.parse(fileContents);
+    }
+    return yaml.load(fileContents);
+  } catch (e) {
+    logger.error('[readSpecFile] error', e);
+    return false;
+  }
+}
+
+async function getSpec(url) {
+  const RegularUrl = z
+    .string()
+    .url()
+    .catch(() => false);
+
+  if (RegularUrl.parse(url) && path.extname(url) === '.json') {
+    const response = await fetch(url);
+    return await response.json();
+  }
+
+  const ValidSpecPath = z
+    .string()
+    .url()
+    .catch(async () => {
+      const spec = path.join(__dirname, '..', '.well-known', 'openapi', url);
+      if (!fs.existsSync(spec)) {
+        return false;
+      }
+
+      return await readSpecFile(spec);
+    });
+
+  return ValidSpecPath.parse(url);
+}
+
+async function createOpenAPIPlugin({ data, llm, user, message, memory, signal }) {
+  let spec;
+  try {
+    spec = await getSpec(data.api.url);
+  } catch (error) {
+    logger.error('[createOpenAPIPlugin] getSpec error', error);
+    return null;
+  }
+
+  if (!spec) {
+    logger.warn('[createOpenAPIPlugin] No spec found');
+    return null;
+  }
+
+  const headers = {};
+  const { auth, name_for_model, description_for_model, description_for_human } = data;
+  if (auth && AuthDefinition.parse(auth)) {
+    logger.debug('[createOpenAPIPlugin] auth detected', auth);
+    const { openai } = auth.verification_tokens;
+    if (AuthBearer.parse(auth)) {
+      headers.authorization = `Bearer ${openai}`;
+      logger.debug('[createOpenAPIPlugin] added auth bearer', headers);
+    }
+  }
+
+  const chainOptions = { llm };
+
+  if (data.headers && data.headers['librechat_user_id']) {
+    logger.debug('[createOpenAPIPlugin] id detected', headers);
+    headers[data.headers['librechat_user_id']] = user;
+  }
+
+  if (Object.keys(headers).length > 0) {
+    logger.debug('[createOpenAPIPlugin] headers detected', headers);
+    chainOptions.headers = headers;
+  }
+
+  if (data.params) {
+    logger.debug('[createOpenAPIPlugin] params detected', data.params);
+    chainOptions.params = data.params;
+  }
+
+  let history = '';
+  if (memory) {
+    logger.debug('[createOpenAPIPlugin] openAPI chain: memory detected', memory);
+    const { history: chat_history } = await memory.loadMemoryVariables({});
+    history = chat_history?.length > 0 ? `\n\n## Chat History:\n${chat_history}\n` : '';
+  }
+
+  chainOptions.prompt = ChatPromptTemplate.fromMessages([
+    HumanMessagePromptTemplate.fromTemplate(
+      `# Use the provided API's to respond to this query:\n\n{query}\n\n## Instructions:\n${addLinePrefix(
+        description_for_model,
+      )}${history}`,
+    ),
+  ]);
+
+  const chain = await createOpenAPIChain(spec, chainOptions);
+
+  const { functions } = chain.chains[0].lc_kwargs.llmKwargs;
+
+  return new DynamicStructuredTool({
+    name: name_for_model,
+    description_for_model: `${addLinePrefix(description_for_human)}${createPrompt(
+      name_for_model,
+      functions,
+    )}`,
+    description: `${description_for_human}`,
+    schema: z.object({
+      func: z
+        .string()
+        .describe(
+          `The function to invoke. The functions available are: ${functions
+            .map((func) => func.name)
+            .join(', ')}`,
+        ),
+      intent: z
+        .string()
+        .describe('Describe your intent with the function and your expected result'),
+    }),
+    func: async ({ func = '', intent = '' }) => {
+      const filteredFunctions = functions.filter((f) => f.name === func);
+      chain.chains[0].lc_kwargs.llmKwargs.functions = filteredFunctions;
+      const query = `${message}${func?.length > 0 ? `\n// Intent: ${intent}` : ''}`;
+      const result = await chain.call({
+        query,
+        signal,
+      });
+      return result.response;
+    },
+  });
+}
+
+module.exports = {
+  getSpec,
+  readSpecFile,
+  createOpenAPIPlugin,
+};

api/app/clients/tools/dynamic/OpenAPIPlugin.spec.js

@@ -0,0 +1,72 @@
+const fs = require('fs');
+const { createOpenAPIPlugin, getSpec, readSpecFile } = require('./OpenAPIPlugin');
+
+global.fetch = jest.fn().mockImplementationOnce(() => {
+  return new Promise((resolve) => {
+    resolve({
+      ok: true,
+      json: () => Promise.resolve({ key: 'value' }),
+    });
+  });
+});
+jest.mock('fs', () => ({
+  promises: {
+    readFile: jest.fn(),
+  },
+  existsSync: jest.fn(),
+}));
+
+describe('readSpecFile', () => {
+  it('reads JSON file correctly', async () => {
+    fs.promises.readFile.mockResolvedValue(JSON.stringify({ test: 'value' }));
+    const result = await readSpecFile('test.json');
+    expect(result).toEqual({ test: 'value' });
+  });
+
+  it('reads YAML file correctly', async () => {
+    fs.promises.readFile.mockResolvedValue('test: value');
+    const result = await readSpecFile('test.yaml');
+    expect(result).toEqual({ test: 'value' });
+  });
+
+  it('handles error correctly', async () => {
+    fs.promises.readFile.mockRejectedValue(new Error('test error'));
+    const result = await readSpecFile('test.json');
+    expect(result).toBe(false);
+  });
+});
+
+describe('getSpec', () => {
+  it('fetches spec from url correctly', async () => {
+    const parsedJson = await getSpec('https://www.instacart.com/.well-known/ai-plugin.json');
+    const isObject = typeof parsedJson === 'object';
+    expect(isObject).toEqual(true);
+  });
+
+  it('reads spec from file correctly', async () => {
+    fs.existsSync.mockReturnValue(true);
+    fs.promises.readFile.mockResolvedValue(JSON.stringify({ test: 'value' }));
+    const result = await getSpec('test.json');
+    expect(result).toEqual({ test: 'value' });
+  });
+
+  it('returns false when file does not exist', async () => {
+    fs.existsSync.mockReturnValue(false);
+    const result = await getSpec('test.json');
+    expect(result).toBe(false);
+  });
+});
+
+describe('createOpenAPIPlugin', () => {
+  it('returns null when getSpec throws an error', async () => {
+    const result = await createOpenAPIPlugin({ data: { api: { url: 'invalid' } } });
+    expect(result).toBe(null);
+  });
+
+  it('returns null when no spec is found', async () => {
+    const result = await createOpenAPIPlugin({});
+    expect(result).toBe(null);
+  });
+
+  // Add more tests here for different scenarios
+});

api/app/clients/tools/structured/DALLE3.js

@@ -8,10 +8,10 @@ const { HttpsProxyAgent } = require('https-proxy-agent');
 const { FileContext, ContentTypes } = require('librechat-data-provider');
 const { getImageBasename } = require('~/server/services/Files/images');
 const extractBaseURL = require('~/utils/extractBaseURL');
-const logger = require('~/config/winston');
+const { logger } = require('~/config');
 
 const displayMessage =
-  "DALL-E displayed an image. All generated images are already plainly visible, so don't repeat the descriptions in detail. Do not list download links as they are available in the UI already. The user may download the images by clicking on them, but do not mention anything about downloading to the user.";
+  'DALL-E displayed an image. All generated images are already plainly visible, so don\'t repeat the descriptions in detail. Do not list download links as they are available in the UI already. The user may download the images by clicking on them, but do not mention anything about downloading to the user.';
 class DALLE3 extends Tool {
   constructor(fields = {}) {
     super();

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

@@ -4,13 +4,12 @@ const { v4 } = require('uuid');
 const OpenAI = require('openai');
 const FormData = require('form-data');
 const { tool } = require('@langchain/core/tools');
-const { logAxiosError } = require('@librechat/api');
-const { logger } = require('@librechat/data-schemas');
 const { HttpsProxyAgent } = require('https-proxy-agent');
 const { ContentTypes, EImageOutputType } = require('librechat-data-provider');
 const { getStrategyFunctions } = require('~/server/services/Files/strategies');
-const { extractBaseURL } = require('~/utils');
+const { logAxiosError, extractBaseURL } = require('~/utils');
 const { getFiles } = require('~/models/File');
+const { logger } = require('~/config');
 
 /** Default descriptions for image generation tool  */
 const DEFAULT_IMAGE_GEN_DESCRIPTION = `
@@ -65,7 +64,7 @@ const DEFAULT_IMAGE_EDIT_PROMPT_DESCRIPTION = `Describe the changes, enhancement
       Always base this prompt on the most recently uploaded reference images.`;
 
 const displayMessage =
-  "The tool displayed an image. All generated images are already plainly visible, so don't repeat the descriptions in detail. Do not list download links as they are available in the UI already. The user may download the images by clicking on them, but do not mention anything about downloading to the user.";
+  'The tool displayed an image. All generated images are already plainly visible, so don\'t repeat the descriptions in detail. Do not list download links as they are available in the UI already. The user may download the images by clicking on them, but do not mention anything about downloading to the user.';
 
 /**
  * Replaces unwanted characters from the input string
@@ -107,12 +106,6 @@ 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
@@ -207,18 +200,10 @@ function createOpenAIImageTools(fields = {}) {
       }
 
       let resp;
-      /** @type {AbortSignal} */
-      let derivedSignal = null;
-      /** @type {() => void} */
-      let abortHandler = null;
-
       try {
-        if (runnableConfig?.signal) {
-          derivedSignal = AbortSignal.any([runnableConfig.signal]);
-          abortHandler = createAbortHandler();
-          derivedSignal.addEventListener('abort', abortHandler, { once: true });
-        }
-
+        const derivedSignal = runnableConfig?.signal
+          ? AbortSignal.any([runnableConfig.signal])
+          : undefined;
         resp = await openai.images.generate(
           {
             model: 'gpt-image-1',
@@ -242,10 +227,6 @@ 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) {
@@ -427,17 +408,10 @@ Error Message: ${error.message}`);
         headers['Authorization'] = `Bearer ${apiKey}`;
       }
 
-      /** @type {AbortSignal} */
-      let derivedSignal = null;
-      /** @type {() => void} */
-      let abortHandler = null;
-
       try {
-        if (runnableConfig?.signal) {
-          derivedSignal = AbortSignal.any([runnableConfig.signal]);
-          abortHandler = createAbortHandler();
-          derivedSignal.addEventListener('abort', abortHandler, { once: true });
-        }
+        const derivedSignal = runnableConfig?.signal
+          ? AbortSignal.any([runnableConfig.signal])
+          : undefined;
 
         /** @type {import('axios').AxiosRequestConfig} */
         const axiosConfig = {
@@ -492,10 +466,6 @@ 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/structured/specs/DALLE3.spec.js

@@ -1,29 +1,10 @@
 const OpenAI = require('openai');
 const DALLE3 = require('../DALLE3');
-const logger = require('~/config/winston');
+
+const { logger } = require('~/config');
 
 jest.mock('openai');
 
-jest.mock('@librechat/data-schemas', () => {
-  return {
-    logger: {
-      info: jest.fn(),
-      warn: jest.fn(),
-      debug: jest.fn(),
-      error: jest.fn(),
-    },
-  };
-});
-
-jest.mock('tiktoken', () => {
-  return {
-    encoding_for_model: jest.fn().mockReturnValue({
-      encode: jest.fn(),
-      decode: jest.fn(),
-    }),
-  };
-});
-
 const processFileURL = jest.fn();
 
 jest.mock('~/server/services/Files/images', () => ({
@@ -56,11 +37,6 @@ jest.mock('fs', () => {
   return {
     existsSync: jest.fn(),
     mkdirSync: jest.fn(),
-    promises: {
-      writeFile: jest.fn(),
-      readFile: jest.fn(),
-      unlink: jest.fn(),
-    },
   };
 });
 

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

@@ -135,7 +135,7 @@ const createFileSearchTool = async ({ req, files, entity_id }) => {
         query: z
           .string()
           .describe(
-            "A natural language query to search for relevant information in the files. Be specific and use keywords related to the information you're looking for. The query will be used for semantic similarity matching against the file contents.",
+            'A natural language query to search for relevant information in the files. Be specific and use keywords related to the information you\'re looking for. The query will be used for semantic similarity matching against the file contents.',
           ),
       }),
     },

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

@@ -1,14 +1,14 @@
-const { mcpToolPattern } = require('@librechat/api');
-const { logger } = require('@librechat/data-schemas');
 const { SerpAPI } = require('@langchain/community/tools/serpapi');
 const { Calculator } = require('@langchain/community/tools/calculator');
 const { EnvVar, createCodeExecutionTool, createSearchTool } = require('@librechat/agents');
 const {
   Tools,
+  Constants,
   EToolResources,
   loadWebSearchAuth,
   replaceSpecialVars,
 } = require('librechat-data-provider');
+const { getUserPluginAuthValue } = require('~/server/services/PluginService');
 const {
   availableTools,
   manifestToolMap,
@@ -28,10 +28,11 @@ const {
 } = require('../');
 const { primeFiles: primeCodeFiles } = require('~/server/services/Files/Code/process');
 const { createFileSearchTool, primeFiles: primeSearchFiles } = require('./fileSearch');
-const { getUserPluginAuthValue } = require('~/server/services/PluginService');
 const { loadAuthValues } = require('~/server/services/Tools/credentials');
-const { getCachedTools } = require('~/server/services/Config');
 const { createMCPTool } = require('~/server/services/MCP');
+const { logger } = require('~/config');
+
+const mcpToolPattern = new RegExp(`^.+${Constants.mcp_delimiter}.+$`);
 
 /**
  * Validates the availability and authentication of tools for a user based on environment variables or user-specific plugin authentication values.
@@ -92,7 +93,7 @@ const validateTools = async (user, tools = []) => {
     return Array.from(validToolsSet.values());
   } catch (err) {
     logger.error('[validateTools] There was a problem validating tools', err);
-    throw new Error(err);
+    throw new Error('There was a problem validating tools');
   }
 };
 
@@ -235,7 +236,7 @@ const loadTools = async ({
 
   /** @type {Record<string, string>} */
   const toolContextMap = {};
-  const appTools = (await getCachedTools({ includeGlobal: true })) ?? {};
+  const appTools = options.req?.app?.locals?.availableTools ?? {};
 
   for (const tool of tools) {
     if (tool === Tools.execute_code) {
@@ -298,7 +299,6 @@ Current Date & Time: ${replaceSpecialVars({ text: '{{iso_datetime}}' })}
       requestedTools[tool] = async () =>
         createMCPTool({
           req: options.req,
-          res: options.res,
           toolKey: tool,
           model: agent?.model ?? model,
           provider: agent?.provider ?? endpoint,

api/app/clients/tools/util/handleTools.test.js

@@ -1,5 +1,8 @@
-const mongoose = require('mongoose');
-const { MongoMemoryServer } = require('mongodb-memory-server');
+const mockUser = {
+  _id: 'fakeId',
+  save: jest.fn(),
+  findByIdAndDelete: jest.fn(),
+};
 
 const mockPluginService = {
   updateUserPluginAuth: jest.fn(),
@@ -7,18 +10,29 @@ const mockPluginService = {
   getUserPluginAuthValue: jest.fn(),
 };
 
+const mockModels = {
+  User: mockUser,
+};
+jest.mock('~/db/connect', () => {
+  return {
+    connectDb: jest.fn(),
+    User: mockModels.mockUser,
+  };
+});
+jest.mock('~/models/File', () => ({
+  File: jest.fn(),
+}));
+
 jest.mock('~/server/services/PluginService', () => mockPluginService);
 
 const { BaseLLM } = require('@langchain/openai');
 const { Calculator } = require('@langchain/community/tools/calculator');
 
-const { User } = require('~/db/models');
 const PluginService = require('~/server/services/PluginService');
 const { validateTools, loadTools, loadToolWithAuth } = require('./handleTools');
 const { StructuredSD, availableTools, DALLE3 } = require('../');
 
 describe('Tool Handlers', () => {
-  let mongoServer;
   let fakeUser;
   const pluginKey = 'dalle';
   const pluginKey2 = 'wolfram';
@@ -29,9 +43,7 @@ describe('Tool Handlers', () => {
   const authConfigs = mainPlugin.authConfig;
 
   beforeAll(async () => {
-    mongoServer = await MongoMemoryServer.create();
-    const mongoUri = mongoServer.getUri();
-    await mongoose.connect(mongoUri);
+    mockUser.save.mockResolvedValue(undefined);
 
     const userAuthValues = {};
     mockPluginService.getUserPluginAuthValue.mockImplementation((userId, authField) => {
@@ -46,7 +58,7 @@ describe('Tool Handlers', () => {
       },
     );
 
-    fakeUser = new User({
+    fakeUser = await mockModels.User.createUser({
       name: 'Fake User',
       username: 'fakeuser',
       email: 'fakeuser@example.com',
@@ -72,36 +84,9 @@ describe('Tool Handlers', () => {
   });
 
   afterAll(async () => {
-    await mongoose.disconnect();
-    await mongoServer.stop();
-  });
-
-  beforeEach(async () => {
-    // Clear mocks but not the database since we need the user to persist
-    jest.clearAllMocks();
-
-    // Reset the mock implementations
-    const userAuthValues = {};
-    mockPluginService.getUserPluginAuthValue.mockImplementation((userId, authField) => {
-      return userAuthValues[`${userId}-${authField}`];
-    });
-    mockPluginService.updateUserPluginAuth.mockImplementation(
-      (userId, authField, _pluginKey, credential) => {
-        const fields = authField.split('||');
-        fields.forEach((field) => {
-          userAuthValues[`${userId}-${field}`] = credential;
-        });
-      },
-    );
-
-    // Re-add the auth configs for the user
+    await mockUser.findByIdAndDelete(fakeUser._id);
     for (const authConfig of authConfigs) {
-      await PluginService.updateUserPluginAuth(
-        fakeUser._id,
-        authConfig.authField,
-        pluginKey,
-        mockCredential,
-      );
+      await PluginService.deleteUserPluginAuth(fakeUser._id, authConfig.authField);
     }
   });
 

api/cache/banViolation.js

@@ -1,8 +1,7 @@
 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/banViolation.spec.js

@@ -1,28 +1,69 @@
-const mongoose = require('mongoose');
-const { MongoMemoryServer } = require('mongodb-memory-server');
 const banViolation = require('./banViolation');
 
-// Mock deleteAllUserSessions since we're testing ban logic, not session deletion
-jest.mock('~/models', () => ({
-  ...jest.requireActual('~/models'),
-  deleteAllUserSessions: jest.fn().mockResolvedValue(true),
+const mockModels = {
+  Session: {
+    deleteAllUserSessions: jest.fn(),
+  },
+};
+
+jest.mock('~/db/connect', () => {
+  return {
+    connectDb: jest.fn(),
+    get models() {
+      return mockModels;
+    },
+  };
+});
+
+jest.mock('~/server/utils', () => ({
+  isEnabled: jest.fn(() => true), // default to false, override per test if needed
+  math: jest.fn(() => 20), // default to false, override per test if needed
+  removePorts: jest.fn(),
 }));
 
+jest.mock('keyv');
+// jest.mock('../models/Session');
+// Mocking the getLogStores function
+jest.mock('./getLogStores', () => {
+  return jest.fn().mockImplementation(() => {
+    const EventEmitter = require('events');
+    const { CacheKeys } = require('librechat-data-provider');
+    const math = require('../server/utils/math');
+    const mockGet = jest.fn();
+    const mockSet = jest.fn();
+    class KeyvMongo extends EventEmitter {
+      constructor(url = 'mongodb://127.0.0.1:27017', options) {
+        super();
+        this.ttlSupport = false;
+        url = url ?? {};
+        if (typeof url === 'string') {
+          url = { url };
+        }
+        if (url.uri) {
+          url = { url: url.uri, ...url };
+        }
+        this.opts = {
+          url,
+          collection: 'keyv',
+          ...url,
+          ...options,
+        };
+      }
+
+      get = mockGet;
+      set = mockSet;
+    }
+
+    return new KeyvMongo('', {
+      namespace: CacheKeys.BANS,
+      ttl: math(process.env.BAN_DURATION, 7200000),
+    });
+  });
+});
+
 describe('banViolation', () => {
-  let mongoServer;
   let req, res, errorMessage;
 
-  beforeAll(async () => {
-    mongoServer = await MongoMemoryServer.create();
-    const mongoUri = mongoServer.getUri();
-    await mongoose.connect(mongoUri);
-  });
-
-  afterAll(async () => {
-    await mongoose.disconnect();
-    await mongoServer.stop();
-  });
-
   beforeEach(() => {
     req = {
       ip: '127.0.0.1',
@@ -35,7 +76,7 @@ describe('banViolation', () => {
     };
     errorMessage = {
       type: 'someViolation',
-      user_id: new mongoose.Types.ObjectId().toString(), // Use valid ObjectId
+      user_id: '12345',
       prev_count: 0,
       violation_count: 0,
     };

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 { math, isEnabled } = require('~/server/utils');
 const keyvRedis = require('./keyvRedis');
 const keyvMongo = require('./keyvMongo');
 
@@ -29,10 +29,6 @@ const roles = isRedisEnabled
   ? new Keyv({ store: keyvRedis })
   : new Keyv({ namespace: CacheKeys.ROLES });
 
-const mcpTools = isRedisEnabled
-  ? new Keyv({ store: keyvRedis })
-  : new Keyv({ namespace: CacheKeys.MCP_TOOLS });
-
 const audioRuns = isRedisEnabled
   ? new Keyv({ store: keyvRedis, ttl: Time.TEN_MINUTES })
   : new Keyv({ namespace: CacheKeys.AUDIO_RUNS, ttl: Time.TEN_MINUTES });
@@ -71,7 +67,6 @@ const openIdExchangedTokensCache = isRedisEnabled
 
 const namespaces = {
   [CacheKeys.ROLES]: roles,
-  [CacheKeys.MCP_TOOLS]: mcpTools,
   [CacheKeys.CONFIG_STORE]: config,
   [CacheKeys.PENDING_REQ]: pending_req,
   [ViolationTypes.BAN]: new Keyv({ store: keyvMongo, namespace: CacheKeys.BANS, ttl: duration }),

api/config/index.js

@@ -1,6 +1,7 @@
+const axios = require('axios');
 const { EventSource } = require('eventsource');
-const { Time } = require('librechat-data-provider');
-const { MCPManager, FlowStateManager } = require('@librechat/api');
+const { Time, CacheKeys } = require('librechat-data-provider');
+const { MCPManager, FlowStateManager } = require('librechat-mcp');
 const logger = require('./winston');
 
 global.EventSource = EventSource;
@@ -15,7 +16,7 @@ let flowManager = null;
  */
 function getMCPManager(userId) {
   if (!mcpManager) {
-    mcpManager = MCPManager.getInstance();
+    mcpManager = MCPManager.getInstance(logger);
   } else {
     mcpManager.checkIdleConnections(userId);
   }
@@ -30,13 +31,66 @@ function getFlowStateManager(flowsCache) {
   if (!flowManager) {
     flowManager = new FlowStateManager(flowsCache, {
       ttl: Time.ONE_MINUTE * 3,
+      logger,
     });
   }
   return flowManager;
 }
 
+/**
+ * Sends message data in Server Sent Events format.
+ * @param {ServerResponse} res - The server response.
+ * @param {{ data: string | Record<string, unknown>, event?: string }} event - The message event.
+ * @param {string} event.event - The type of event.
+ * @param {string} event.data - The message to be sent.
+ */
+const sendEvent = (res, event) => {
+  if (typeof event.data === 'string' && event.data.length === 0) {
+    return;
+  }
+  res.write(`event: message\ndata: ${JSON.stringify(event)}\n\n`);
+};
+
+/**
+ * Creates and configures an Axios instance with optional proxy settings.
+ *
+ * @typedef {import('axios').AxiosInstance} AxiosInstance
+ * @typedef {import('axios').AxiosProxyConfig} AxiosProxyConfig
+ *
+ * @returns {AxiosInstance} A configured Axios instance
+ * @throws {Error} If there's an issue creating the Axios instance or parsing the proxy URL
+ */
+function createAxiosInstance() {
+  const instance = axios.create();
+
+  if (process.env.proxy) {
+    try {
+      const url = new URL(process.env.proxy);
+
+      /** @type {AxiosProxyConfig} */
+      const proxyConfig = {
+        host: url.hostname.replace(/^\[|\]$/g, ''),
+        protocol: url.protocol.replace(':', ''),
+      };
+
+      if (url.port) {
+        proxyConfig.port = parseInt(url.port, 10);
+      }
+
+      instance.defaults.proxy = proxyConfig;
+    } catch (error) {
+      console.error('Error parsing proxy URL:', error);
+      throw new Error(`Invalid proxy URL: ${process.env.proxy}`);
+    }
+  }
+
+  return instance;
+}
+
 module.exports = {
   logger,
+  sendEvent,
   getMCPManager,
+  createAxiosInstance,
   getFlowStateManager,
 };

api/config/index.spec.js

@@ -1,6 +1,7 @@
-import axios from 'axios';
-import { createAxiosInstance } from './axios';
+const axios = require('axios');
+const { createAxiosInstance } = require('./index');
 
+// Mock axios
 jest.mock('axios', () => ({
   interceptors: {
     request: { use: jest.fn(), eject: jest.fn() },
@@ -19,13 +20,7 @@ jest.mock('axios', () => ({
   post: jest.fn().mockResolvedValue({ data: {} }),
   put: jest.fn().mockResolvedValue({ data: {} }),
   delete: jest.fn().mockResolvedValue({ data: {} }),
-  reset: jest.fn().mockImplementation(function (this: {
-    get: jest.Mock;
-    post: jest.Mock;
-    put: jest.Mock;
-    delete: jest.Mock;
-    create: jest.Mock;
-  }) {
+  reset: jest.fn().mockImplementation(function () {
     this.get.mockClear();
     this.post.mockClear();
     this.put.mockClear();

api/db/indexSync.js

@@ -1,11 +1,8 @@
 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;
@@ -31,123 +28,43 @@ 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;
   }
-
-  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();
+    const client = MeiliSearchClient.getInstance();
+
+    const { status } = await client.health();
+    if (status !== 'available') {
+      throw new Error('Meilisearch not available');
     }
 
-    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');
+    if (indexingDisabled === true) {
+      logger.info('[indexSync] Indexing is disabled, skipping...');
       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/Action.js

@@ -1,4 +1,5 @@
-const { Action } = require('~/db/models');
+const mongoose = require('mongoose');
+const Action = require('~/db/models').Action;
 
 /**
  * Update an action with new data without overwriting existing properties,

api/models/Agent.js

@@ -11,10 +11,10 @@ const {
   removeAgentIdsFromProject,
   removeAgentFromAllProjects,
 } = require('./Project');
-const { getCachedTools } = require('~/server/services/Config');
 const getLogStores = require('~/cache/getLogStores');
 const { getActions } = require('./Action');
-const { Agent } = require('~/db/models');
+
+const Agent = require('~/db/models').Agent;
 
 /**
  * Create an agent with the provided data.
@@ -56,12 +56,12 @@ const getAgent = async (searchParameter) => await Agent.findOne(searchParameter)
  * @param {string} params.agent_id
  * @param {string} params.endpoint
  * @param {import('@librechat/agents').ClientOptions} [params.model_parameters]
- * @returns {Promise<Agent|null>} The agent document as a plain object, or null if not found.
+ * @returns {Agent|null} The agent document as a plain object, or null if not found.
  */
-const loadEphemeralAgent = async ({ req, agent_id, endpoint, model_parameters: _m }) => {
+const loadEphemeralAgent = ({ req, agent_id, endpoint, model_parameters: _m }) => {
   const { model, ...model_parameters } = _m;
   /** @type {Record<string, FunctionTool>} */
-  const availableTools = await getCachedTools({ includeGlobal: true });
+  const availableTools = req.app.locals.availableTools;
   /** @type {TEphemeralAgent | null} */
   const ephemeralAgent = req.body.ephemeralAgent;
   const mcpServers = new Set(ephemeralAgent?.mcp);
@@ -70,9 +70,6 @@ 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);
   }
@@ -115,7 +112,7 @@ const loadAgent = async ({ req, agent_id, endpoint, model_parameters }) => {
     return null;
   }
   if (agent_id === EPHEMERAL_AGENT_ID) {
-    return await loadEphemeralAgent({ req, agent_id, endpoint, model_parameters });
+    return loadEphemeralAgent({ req, agent_id, endpoint, model_parameters });
   }
   const agent = await getAgent({
     id: agent_id,
@@ -174,6 +171,7 @@ const isDuplicateVersion = (updateData, currentData, versions, actionsHash = nul
     'created_at',
     'updated_at',
     '__v',
+    'agent_ids',
     'versions',
     'actionsHash', // Exclude actionsHash from direct comparison
   ];
@@ -263,12 +261,11 @@ const isDuplicateVersion = (updateData, currentData, versions, actionsHash = nul
  * @param {Object} [options] - Optional configuration object.
  * @param {string} [options.updatingUserId] - The ID of the user performing the update (used for tracking non-author updates).
  * @param {boolean} [options.forceVersion] - Force creation of a new version even if no fields changed.
- * @param {boolean} [options.skipVersioning] - Skip version creation entirely (useful for isolated operations like sharing).
  * @returns {Promise<Agent>} The updated or newly created agent document as a plain object.
  * @throws {Error} If the update would create a duplicate version
  */
 const updateAgent = async (searchParameter, updateData, options = {}) => {
-  const { updatingUserId = null, forceVersion = false, skipVersioning = false } = options;
+  const { updatingUserId = null, forceVersion = false } = options;
   const mongoOptions = { new: true, upsert: false };
 
   const currentAgent = await Agent.findOne(searchParameter);
@@ -305,8 +302,10 @@ const updateAgent = async (searchParameter, updateData, options = {}) => {
     }
 
     const shouldCreateVersion =
-      !skipVersioning &&
-      (forceVersion || Object.keys(directUpdates).length > 0 || $push || $pull || $addToSet);
+      forceVersion ||
+      (versions &&
+        versions.length > 0 &&
+        (Object.keys(directUpdates).length > 0 || $push || $pull || $addToSet));
 
     if (shouldCreateVersion) {
       const duplicateVersion = isDuplicateVersion(updateData, versionData, versions, actionsHash);
@@ -341,7 +340,7 @@ const updateAgent = async (searchParameter, updateData, options = {}) => {
       versionEntry.updatedBy = new mongoose.Types.ObjectId(updatingUserId);
     }
 
-    if (shouldCreateVersion) {
+    if (shouldCreateVersion || forceVersion) {
       updateData.$push = {
         ...($push || {}),
         versions: versionEntry,
@@ -552,10 +551,7 @@ const updateAgentProjects = async ({ user, agentId, projectIds, removeProjectIds
     delete updateQuery.author;
   }
 
-  const updatedAgent = await updateAgent(updateQuery, updateOps, {
-    updatingUserId: user.id,
-    skipVersioning: true,
-  });
+  const updatedAgent = await updateAgent(updateQuery, updateOps, { updatingUserId: user.id });
   if (updatedAgent) {
     return updatedAgent;
   }

api/models/Assistant.js

@@ -1,4 +1,5 @@
-const { Assistant } = require('~/db/models');
+const mongoose = require('mongoose');
+const Assistant = require('~/db/models').Assistant;
 
 /**
  * Update an assistant with new data without overwriting existing properties,

api/models/Banner.js

@@ -1,5 +1,7 @@
+const mongoose = require('mongoose');
 const { logger } = require('@librechat/data-schemas');
-const { Banner } = require('~/db/models');
+
+const Banner = require('~/db/models').Banner;
 
 /**
  * Retrieves the current active banner.

api/models/Conversation.js

@@ -1,8 +1,8 @@
+const mongoose = require('mongoose');
 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');
+
+const Conversation = require('~/db/models').Conversation;
 
 /**
  * Searches for a conversation by conversationId and returns a lean document with only conversationId and user.
@@ -100,15 +100,10 @@ module.exports = {
         update.conversationId = newConversationId;
       }
 
-      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;
-        }
+      if (req.body.isTemporary) {
+        const expiredAt = new Date();
+        expiredAt.setDate(expiredAt.getDate() + 30);
+        update.expiredAt = expiredAt;
       } else {
         update.expiredAt = null;
       }

api/models/ConversationTag.js

@@ -1,5 +1,8 @@
+const mongoose = require('mongoose');
 const { logger } = require('@librechat/data-schemas');
-const { ConversationTag, Conversation } = require('~/db/models');
+
+const ConversationTag = require('~/db/models').ConversationTag;
+const Conversation = require('~/db/models').Conversation;
 
 /**
  * Retrieves all conversation tags for a user.

api/models/File.js

@@ -1,6 +1,8 @@
+const mongoose = require('mongoose');
 const { logger } = require('@librechat/data-schemas');
-const { EToolResources, FileContext } = require('librechat-data-provider');
-const { File } = require('~/db/models');
+const { EToolResources } = require('librechat-data-provider');
+
+const File = require('~/db/models').File;
 
 /**
  * Finds a file by its file_id with additional query options.
@@ -32,19 +34,19 @@ 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 || !toolResourceSet?.size) {
+  if (!fileIds || !fileIds.length) {
     return [];
   }
 
   try {
     const filter = {
       file_id: { $in: fileIds },
-      $or: [],
     };
 
-    if (toolResourceSet.has(EToolResources.ocr)) {
-      filter.$or.push({ text: { $exists: true, $ne: null }, context: FileContext.agents });
+    if (toolResourceSet.size) {
+      filter.$or = [];
     }
+
     if (toolResourceSet.has(EToolResources.file_search)) {
       filter.$or.push({ embedded: true });
     }

api/models/Message.js

@@ -1,9 +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 Message = require('~/db/models').Message;
 const idSchema = z.string().uuid();
 
 /**
@@ -56,14 +54,9 @@ async function saveMessage(req, params, metadata) {
     };
 
     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(`---\`saveMessage\` context: ${metadata?.context}`);
-        update.expiredAt = null;
-      }
+      const expiredAt = new Date();
+      expiredAt.setDate(expiredAt.getDate() + 30);
+      update.expiredAt = expiredAt;
     } else {
       update.expiredAt = null;
     }
@@ -260,7 +253,6 @@ async function updateMessage(req, message, metadata) {
       text: updatedMessage.text,
       isCreatedByUser: updatedMessage.isCreatedByUser,
       tokenCount: updatedMessage.tokenCount,
-      feedback: updatedMessage.feedback,
     };
   } catch (err) {
     logger.error('Error updating message:', err);

api/models/Message.spec.js

@@ -1,7 +1,37 @@
-const mongoose = require('mongoose');
-const { MongoMemoryServer } = require('mongodb-memory-server');
 const { v4: uuidv4 } = require('uuid');
-const { messageSchema } = require('@librechat/data-schemas');
+
+jest.mock('mongoose');
+
+const mockFindQuery = {
+  select: jest.fn().mockReturnThis(),
+  sort: jest.fn().mockReturnThis(),
+  lean: jest.fn().mockReturnThis(),
+  deleteMany: jest.fn().mockResolvedValue({ deletedCount: 1 }),
+};
+
+const mockSchema = {
+  findOneAndUpdate: jest.fn(),
+  updateOne: jest.fn(),
+  findOne: jest.fn(() => ({
+    lean: jest.fn(),
+  })),
+  find: jest.fn(() => mockFindQuery),
+  deleteMany: jest.fn(),
+};
+
+jest.mock('~/config/winston', () => ({
+  error: jest.fn(),
+}));
+
+const mockModels = {
+  Message: {
+    findOneAndUpdate: mockSchema.findOneAndUpdate,
+    updateOne: mockSchema.updateOne,
+    findOne: mockSchema.findOne,
+    find: mockSchema.find,
+    deleteMany: mockSchema.deleteMany,
+  },
+};
 
 const {
   saveMessage,
@@ -10,102 +40,77 @@ const {
   deleteMessages,
   updateMessageText,
   deleteMessagesSince,
-} = require('./Message');
-
-/**
- * @type {import('mongoose').Model<import('@librechat/data-schemas').IMessage>}
- */
-let Message;
+} = require('~/models/Message');
 
 describe('Message Operations', () => {
-  let mongoServer;
   let mockReq;
-  let mockMessageData;
+  let mockMessage;
 
-  beforeAll(async () => {
-    mongoServer = await MongoMemoryServer.create();
-    const mongoUri = mongoServer.getUri();
-    Message = mongoose.models.Message || mongoose.model('Message', messageSchema);
-    await mongoose.connect(mongoUri);
-  });
-
-  afterAll(async () => {
-    await mongoose.disconnect();
-    await mongoServer.stop();
-  });
-
-  beforeEach(async () => {
-    // Clear database
-    await Message.deleteMany({});
+  beforeEach(() => {
+    jest.clearAllMocks();
 
     mockReq = {
       user: { id: 'user123' },
     };
 
-    mockMessageData = {
+    mockMessage = {
       messageId: 'msg123',
       conversationId: uuidv4(),
       text: 'Hello, world!',
       user: 'user123',
     };
+
+    mockSchema.findOneAndUpdate.mockResolvedValue({
+      toObject: () => mockMessage,
+    });
   });
 
   describe('saveMessage', () => {
     it('should save a message for an authenticated user', async () => {
-      const result = await saveMessage(mockReq, mockMessageData);
-
-      expect(result.messageId).toBe('msg123');
-      expect(result.user).toBe('user123');
-      expect(result.text).toBe('Hello, world!');
-
-      // Verify the message was actually saved to the database
-      const savedMessage = await Message.findOne({ messageId: 'msg123', user: 'user123' });
-      expect(savedMessage).toBeTruthy();
-      expect(savedMessage.text).toBe('Hello, world!');
+      const result = await saveMessage(mockReq, mockMessage);
+      expect(result).toEqual(mockMessage);
+      expect(mockSchema.findOneAndUpdate).toHaveBeenCalledWith(
+        { messageId: 'msg123', user: 'user123' },
+        expect.objectContaining({ user: 'user123' }),
+        expect.any(Object),
+      );
     });
 
     it('should throw an error for unauthenticated user', async () => {
       mockReq.user = null;
-      await expect(saveMessage(mockReq, mockMessageData)).rejects.toThrow('User not authenticated');
+      await expect(saveMessage(mockReq, mockMessage)).rejects.toThrow('User not authenticated');
     });
 
-    it('should handle invalid conversation ID gracefully', async () => {
-      mockMessageData.conversationId = 'invalid-id';
-      const result = await saveMessage(mockReq, mockMessageData);
-      expect(result).toBeUndefined();
+    it('should throw an error for invalid conversation ID', async () => {
+      mockMessage.conversationId = 'invalid-id';
+      await expect(saveMessage(mockReq, mockMessage)).resolves.toBeUndefined();
     });
   });
 
   describe('updateMessageText', () => {
     it('should update message text for the authenticated user', async () => {
-      // First save a message
-      await saveMessage(mockReq, mockMessageData);
-
-      // Then update it
       await updateMessageText(mockReq, { messageId: 'msg123', text: 'Updated text' });
-
-      // Verify the update
-      const updatedMessage = await Message.findOne({ messageId: 'msg123', user: 'user123' });
-      expect(updatedMessage.text).toBe('Updated text');
+      expect(mockSchema.updateOne).toHaveBeenCalledWith(
+        { messageId: 'msg123', user: 'user123' },
+        { text: 'Updated text' },
+      );
     });
   });
 
   describe('updateMessage', () => {
     it('should update a message for the authenticated user', async () => {
-      // First save a message
-      await saveMessage(mockReq, mockMessageData);
-
+      mockSchema.findOneAndUpdate.mockResolvedValue(mockMessage);
       const result = await updateMessage(mockReq, { messageId: 'msg123', text: 'Updated text' });
-
-      expect(result.messageId).toBe('msg123');
-      expect(result.text).toBe('Updated text');
-
-      // Verify in database
-      const updatedMessage = await Message.findOne({ messageId: 'msg123', user: 'user123' });
-      expect(updatedMessage.text).toBe('Updated text');
+      expect(result).toEqual(
+        expect.objectContaining({
+          messageId: 'msg123',
+          text: 'Hello, world!',
+        }),
+      );
     });
 
     it('should throw an error if message is not found', async () => {
+      mockSchema.findOneAndUpdate.mockResolvedValue(null);
       await expect(
         updateMessage(mockReq, { messageId: 'nonexistent', text: 'Test' }),
       ).rejects.toThrow('Message not found or user not authorized.');
@@ -114,45 +119,19 @@ describe('Message Operations', () => {
 
   describe('deleteMessagesSince', () => {
     it('should delete messages only for the authenticated user', async () => {
-      const conversationId = uuidv4();
-
-      // Create multiple messages in the same conversation
-      const message1 = await saveMessage(mockReq, {
-        messageId: 'msg1',
-        conversationId,
-        text: 'First message',
-        user: 'user123',
+      mockSchema.findOne().lean.mockResolvedValueOnce({ createdAt: new Date() });
+      mockFindQuery.deleteMany.mockResolvedValueOnce({ deletedCount: 1 });
+      const result = await deleteMessagesSince(mockReq, {
+        messageId: 'msg123',
+        conversationId: 'convo123',
       });
-
-      const message2 = await saveMessage(mockReq, {
-        messageId: 'msg2',
-        conversationId,
-        text: 'Second message',
-        user: 'user123',
-      });
-
-      const message3 = await saveMessage(mockReq, {
-        messageId: 'msg3',
-        conversationId,
-        text: 'Third message',
-        user: 'user123',
-      });
-
-      // Delete messages since message2 (this should only delete messages created AFTER msg2)
-      await deleteMessagesSince(mockReq, {
-        messageId: 'msg2',
-        conversationId,
-      });
-
-      // Verify msg1 and msg2 remain, msg3 is deleted
-      const remainingMessages = await Message.find({ conversationId, user: 'user123' });
-      expect(remainingMessages).toHaveLength(2);
-      expect(remainingMessages.map((m) => m.messageId)).toContain('msg1');
-      expect(remainingMessages.map((m) => m.messageId)).toContain('msg2');
-      expect(remainingMessages.map((m) => m.messageId)).not.toContain('msg3');
+      expect(mockSchema.findOne).toHaveBeenCalledWith({ messageId: 'msg123', user: 'user123' });
+      expect(mockSchema.find).not.toHaveBeenCalled();
+      expect(result).toBeUndefined();
     });
 
     it('should return undefined if no message is found', async () => {
+      mockSchema.findOne().lean.mockResolvedValueOnce(null);
       const result = await deleteMessagesSince(mockReq, {
         messageId: 'nonexistent',
         conversationId: 'convo123',
@@ -163,71 +142,29 @@ describe('Message Operations', () => {
 
   describe('getMessages', () => {
     it('should retrieve messages with the correct filter', async () => {
-      const conversationId = uuidv4();
-
-      // Save some messages
-      await saveMessage(mockReq, {
-        messageId: 'msg1',
-        conversationId,
-        text: 'First message',
-        user: 'user123',
-      });
-
-      await saveMessage(mockReq, {
-        messageId: 'msg2',
-        conversationId,
-        text: 'Second message',
-        user: 'user123',
-      });
-
-      const messages = await getMessages({ conversationId });
-      expect(messages).toHaveLength(2);
-      expect(messages[0].text).toBe('First message');
-      expect(messages[1].text).toBe('Second message');
+      const filter = { conversationId: 'convo123' };
+      await getMessages(filter);
+      expect(mockSchema.find).toHaveBeenCalledWith(filter);
+      expect(mockFindQuery.sort).toHaveBeenCalledWith({ createdAt: 1 });
+      expect(mockFindQuery.lean).toHaveBeenCalled();
     });
   });
 
   describe('deleteMessages', () => {
     it('should delete messages with the correct filter', async () => {
-      // Save some messages for different users
-      await saveMessage(mockReq, mockMessageData);
-      await saveMessage(
-        { user: { id: 'user456' } },
-        {
-          messageId: 'msg456',
-          conversationId: uuidv4(),
-          text: 'Other user message',
-          user: 'user456',
-        },
-      );
-
       await deleteMessages({ user: 'user123' });
-
-      // Verify only user123's messages were deleted
-      const user123Messages = await Message.find({ user: 'user123' });
-      const user456Messages = await Message.find({ user: 'user456' });
-
-      expect(user123Messages).toHaveLength(0);
-      expect(user456Messages).toHaveLength(1);
+      expect(mockSchema.deleteMany).toHaveBeenCalledWith({ user: 'user123' });
     });
   });
 
   describe('Conversation Hijacking Prevention', () => {
     it("should not allow editing a message in another user's conversation", async () => {
       const attackerReq = { user: { id: 'attacker123' } };
-      const victimConversationId = uuidv4();
+      const victimConversationId = 'victim-convo-123';
       const victimMessageId = 'victim-msg-123';
 
-      // First, save a message as the victim (but we'll try to edit as attacker)
-      const victimReq = { user: { id: 'victim123' } };
-      await saveMessage(victimReq, {
-        messageId: victimMessageId,
-        conversationId: victimConversationId,
-        text: 'Victim message',
-        user: 'victim123',
-      });
+      mockSchema.findOneAndUpdate.mockResolvedValue(null);
 
-      // Attacker tries to edit the victim's message
       await expect(
         updateMessage(attackerReq, {
           messageId: victimMessageId,
@@ -236,82 +173,71 @@ describe('Message Operations', () => {
         }),
       ).rejects.toThrow('Message not found or user not authorized.');
 
-      // Verify the original message is unchanged
-      const originalMessage = await Message.findOne({
-        messageId: victimMessageId,
-        user: 'victim123',
-      });
-      expect(originalMessage.text).toBe('Victim message');
+      expect(mockSchema.findOneAndUpdate).toHaveBeenCalledWith(
+        { messageId: victimMessageId, user: 'attacker123' },
+        expect.anything(),
+        expect.anything(),
+      );
     });
 
     it("should not allow deleting messages from another user's conversation", async () => {
       const attackerReq = { user: { id: 'attacker123' } };
-      const victimConversationId = uuidv4();
+      const victimConversationId = 'victim-convo-123';
       const victimMessageId = 'victim-msg-123';
 
-      // Save a message as the victim
-      const victimReq = { user: { id: 'victim123' } };
-      await saveMessage(victimReq, {
-        messageId: victimMessageId,
-        conversationId: victimConversationId,
-        text: 'Victim message',
-        user: 'victim123',
-      });
-
-      // Attacker tries to delete from victim's conversation
+      mockSchema.findOne().lean.mockResolvedValueOnce(null); // Simulating message not found for this user
       const result = await deleteMessagesSince(attackerReq, {
         messageId: victimMessageId,
         conversationId: victimConversationId,
       });
 
       expect(result).toBeUndefined();
-
-      // Verify the victim's message still exists
-      const victimMessage = await Message.findOne({
+      expect(mockSchema.findOne).toHaveBeenCalledWith({
         messageId: victimMessageId,
-        user: 'victim123',
+        user: 'attacker123',
       });
-      expect(victimMessage).toBeTruthy();
-      expect(victimMessage.text).toBe('Victim message');
     });
 
     it("should not allow inserting a new message into another user's conversation", async () => {
       const attackerReq = { user: { id: 'attacker123' } };
-      const victimConversationId = uuidv4();
+      const victimConversationId = uuidv4(); // Use a valid UUID
 
-      // Attacker tries to save a message - this should succeed but with attacker's user ID
-      const result = await saveMessage(attackerReq, {
-        conversationId: victimConversationId,
-        text: 'Inserted malicious message',
-        messageId: 'new-msg-123',
-        user: 'attacker123',
-      });
+      await expect(
+        saveMessage(attackerReq, {
+          conversationId: victimConversationId,
+          text: 'Inserted malicious message',
+          messageId: 'new-msg-123',
+        }),
+      ).resolves.not.toThrow(); // It should not throw an error
 
-      expect(result).toBeTruthy();
-      expect(result.user).toBe('attacker123');
-
-      // Verify the message was saved with the attacker's user ID, not as an anonymous message
-      const savedMessage = await Message.findOne({ messageId: 'new-msg-123' });
-      expect(savedMessage.user).toBe('attacker123');
-      expect(savedMessage.conversationId).toBe(victimConversationId);
+      // Check that the message was saved with the attacker's user ID
+      expect(mockSchema.findOneAndUpdate).toHaveBeenCalledWith(
+        { messageId: 'new-msg-123', user: 'attacker123' },
+        expect.objectContaining({
+          user: 'attacker123',
+          conversationId: victimConversationId,
+        }),
+        expect.anything(),
+      );
     });
 
     it('should allow retrieving messages from any conversation', async () => {
-      const victimConversationId = uuidv4();
+      const victimConversationId = 'victim-convo-123';
 
-      // Save a message in the victim's conversation
-      const victimReq = { user: { id: 'victim123' } };
-      await saveMessage(victimReq, {
-        messageId: 'victim-msg',
+      await getMessages({ conversationId: victimConversationId });
+
+      expect(mockSchema.find).toHaveBeenCalledWith({
         conversationId: victimConversationId,
-        text: 'Victim message',
-        user: 'victim123',
       });
 
-      // Anyone should be able to retrieve messages by conversation ID
-      const messages = await getMessages({ conversationId: victimConversationId });
-      expect(messages).toHaveLength(1);
-      expect(messages[0].text).toBe('Victim message');
+      mockSchema.find.mockReturnValueOnce({
+        select: jest.fn().mockReturnThis(),
+        sort: jest.fn().mockReturnThis(),
+        lean: jest.fn().mockResolvedValue([{ text: 'Test message' }]),
+      });
+
+      const result = await getMessages({ conversationId: victimConversationId });
+      expect(result).toEqual([{ text: 'Test message' }]);
     });
   });
 });

api/models/Preset.js

@@ -1,5 +1,7 @@
+const mongoose = require('mongoose');
 const { logger } = require('@librechat/data-schemas');
-const { Preset } = require('~/db/models');
+
+const Preset = require('~/db/models').Preset;
 
 const getPreset = async (user, presetId) => {
   try {

api/models/Project.js

@@ -1,5 +1,7 @@
+const mongoose = require('mongoose');
 const { GLOBAL_PROJECT_NAME } = require('librechat-data-provider').Constants;
-const { Project } = require('~/db/models');
+
+const Project = require('~/db/models').Project;
 
 /**
  * Retrieve a project by ID and convert the found project document to a plain object.

api/models/Prompt.js

@@ -1,3 +1,4 @@
+const mongoose = require('mongoose');
 const { ObjectId } = require('mongodb');
 const { logger } = require('@librechat/data-schemas');
 const { SystemRoles, SystemCategories, Constants } = require('librechat-data-provider');
@@ -7,9 +8,11 @@ const {
   removeGroupIdsFromProject,
   removeGroupFromAllProjects,
 } = require('./Project');
-const { PromptGroup, Prompt } = require('~/db/models');
 const { escapeRegExp } = require('~/server/utils');
 
+const PromptGroup = require('~/db/models').PromptGroup;
+const Prompt = require('~/db/models').Prompt;
+
 /**
  * Create a pipeline for the aggregation to get prompt groups
  * @param {Object} query

api/models/Role.js

@@ -8,7 +8,8 @@ const {
 } = require('librechat-data-provider');
 const { logger } = require('@librechat/data-schemas');
 const getLogStores = require('~/cache/getLogStores');
-const { Role } = require('~/db/models');
+
+const Role = require('~/db/models').Role;
 
 /**
  * Retrieve a role by name and convert the found role document to a plain object.
@@ -170,6 +171,35 @@ async function updateAccessPermissions(roleName, permissionsUpdate) {
   }
 }
 
+/**
+ * Initialize default roles in the system.
+ * Creates the default roles (ADMIN, USER) if they don't exist in the database.
+ * Updates existing roles with new permission types if they're missing.
+ *
+ * @returns {Promise<void>}
+ */
+const initializeRoles = async function () {
+  for (const roleName of [SystemRoles.ADMIN, SystemRoles.USER]) {
+    let role = await Role.findOne({ name: roleName });
+    const defaultPerms = roleDefaults[roleName].permissions;
+
+    if (!role) {
+      // Create new role if it doesn't exist.
+      role = new Role(roleDefaults[roleName]);
+    } else {
+      // Ensure role.permissions is defined.
+      role.permissions = role.permissions || {};
+      // For each permission type in defaults, add it if missing.
+      for (const permType of Object.keys(defaultPerms)) {
+        if (role.permissions[permType] == null) {
+          role.permissions[permType] = defaultPerms[permType];
+        }
+      }
+    }
+    await role.save();
+  }
+};
+
 /**
  * Migrates roles from old schema to new schema structure.
  * This can be called directly to fix existing roles.
@@ -251,7 +281,8 @@ const migrateRoleSchema = async function (roleName) {
 
 module.exports = {
   getRoleByName,
+  initializeRoles,
   updateRoleByName,
-  migrateRoleSchema,
   updateAccessPermissions,
+  migrateRoleSchema,
 };

api/models/Role.spec.js

@@ -6,10 +6,10 @@ const {
   roleDefaults,
   PermissionTypes,
 } = require('librechat-data-provider');
-const { getRoleByName, updateAccessPermissions } = require('~/models/Role');
+const { getRoleByName, updateAccessPermissions, initializeRoles } = require('~/models/Role');
 const getLogStores = require('~/cache/getLogStores');
-const { initializeRoles } = require('~/models');
-const { Role } = require('~/db/models');
+
+const Role = require('~/db/models').Role;
 
 // Mock the cache
 jest.mock('~/cache/getLogStores', () =>

api/models/Share.js

@@ -0,0 +1,349 @@
+const { nanoid } = require('nanoid');
+const mongoose = require('mongoose');
+const { Constants } = require('librechat-data-provider');
+const { logger } = require('@librechat/data-schemas');
+const { getMessages } = require('./Message');
+
+const Conversation = require('~/db/models').Conversation;
+const SharedLink = require('~/db/models').SharedLink;
+
+class ShareServiceError extends Error {
+  constructor(message, code) {
+    super(message);
+    this.name = 'ShareServiceError';
+    this.code = code;
+  }
+}
+
+const memoizedAnonymizeId = (prefix) => {
+  const memo = new Map();
+  return (id) => {
+    if (!memo.has(id)) {
+      memo.set(id, `${prefix}_${nanoid()}`);
+    }
+    return memo.get(id);
+  };
+};
+
+const anonymizeConvoId = memoizedAnonymizeId('convo');
+const anonymizeAssistantId = memoizedAnonymizeId('a');
+const anonymizeMessageId = (id) =>
+  id === Constants.NO_PARENT ? id : memoizedAnonymizeId('msg')(id);
+
+function anonymizeConvo(conversation) {
+  if (!conversation) {
+    return null;
+  }
+
+  const newConvo = { ...conversation };
+  if (newConvo.assistant_id) {
+    newConvo.assistant_id = anonymizeAssistantId(newConvo.assistant_id);
+  }
+  return newConvo;
+}
+
+function anonymizeMessages(messages, newConvoId) {
+  if (!Array.isArray(messages)) {
+    return [];
+  }
+
+  const idMap = new Map();
+  return messages.map((message) => {
+    const newMessageId = anonymizeMessageId(message.messageId);
+    idMap.set(message.messageId, newMessageId);
+
+    const anonymizedAttachments = message.attachments?.map((attachment) => {
+      return {
+        ...attachment,
+        messageId: newMessageId,
+        conversationId: newConvoId,
+      };
+    });
+
+    return {
+      ...message,
+      messageId: newMessageId,
+      parentMessageId:
+        idMap.get(message.parentMessageId) || anonymizeMessageId(message.parentMessageId),
+      conversationId: newConvoId,
+      model: message.model?.startsWith('asst_')
+        ? anonymizeAssistantId(message.model)
+        : message.model,
+      attachments: anonymizedAttachments,
+    };
+  });
+}
+
+async function getSharedMessages(shareId) {
+  try {
+    const share = await SharedLink.findOne({ shareId, isPublic: true })
+      .populate({
+        path: 'messages',
+        select: '-_id -__v -user',
+      })
+      .select('-_id -__v -user')
+      .lean();
+
+    if (!share?.conversationId || !share.isPublic) {
+      return null;
+    }
+
+    const newConvoId = anonymizeConvoId(share.conversationId);
+    const result = {
+      ...share,
+      conversationId: newConvoId,
+      messages: anonymizeMessages(share.messages, newConvoId),
+    };
+
+    return result;
+  } catch (error) {
+    logger.error('[getShare] Error getting share link', {
+      error: error.message,
+      shareId,
+    });
+    throw new ShareServiceError('Error getting share link', 'SHARE_FETCH_ERROR');
+  }
+}
+
+async function getSharedLinks(user, pageParam, pageSize, isPublic, sortBy, sortDirection, search) {
+  try {
+    const query = { user, isPublic };
+
+    if (pageParam) {
+      if (sortDirection === 'desc') {
+        query[sortBy] = { $lt: pageParam };
+      } else {
+        query[sortBy] = { $gt: pageParam };
+      }
+    }
+
+    if (search && search.trim()) {
+      try {
+        const searchResults = await Conversation.meiliSearch(search);
+
+        if (!searchResults?.hits?.length) {
+          return {
+            links: [],
+            nextCursor: undefined,
+            hasNextPage: false,
+          };
+        }
+
+        const conversationIds = searchResults.hits.map((hit) => hit.conversationId);
+        query['conversationId'] = { $in: conversationIds };
+      } catch (searchError) {
+        logger.error('[getSharedLinks] Meilisearch error', {
+          error: searchError.message,
+          user,
+        });
+        return {
+          links: [],
+          nextCursor: undefined,
+          hasNextPage: false,
+        };
+      }
+    }
+
+    const sort = {};
+    sort[sortBy] = sortDirection === 'desc' ? -1 : 1;
+
+    if (Array.isArray(query.conversationId)) {
+      query.conversationId = { $in: query.conversationId };
+    }
+
+    const sharedLinks = await SharedLink.find(query)
+      .sort(sort)
+      .limit(pageSize + 1)
+      .select('-__v -user')
+      .lean();
+
+    const hasNextPage = sharedLinks.length > pageSize;
+    const links = sharedLinks.slice(0, pageSize);
+
+    const nextCursor = hasNextPage ? links[links.length - 1][sortBy] : undefined;
+
+    return {
+      links: links.map((link) => ({
+        shareId: link.shareId,
+        title: link?.title || 'Untitled',
+        isPublic: link.isPublic,
+        createdAt: link.createdAt,
+        conversationId: link.conversationId,
+      })),
+      nextCursor,
+      hasNextPage,
+    };
+  } catch (error) {
+    logger.error('[getSharedLinks] Error getting shares', {
+      error: error.message,
+      user,
+    });
+    throw new ShareServiceError('Error getting shares', 'SHARES_FETCH_ERROR');
+  }
+}
+
+async function deleteAllSharedLinks(user) {
+  try {
+    const result = await SharedLink.deleteMany({ user });
+    return {
+      message: 'All shared links deleted successfully',
+      deletedCount: result.deletedCount,
+    };
+  } catch (error) {
+    logger.error('[deleteAllSharedLinks] Error deleting shared links', {
+      error: error.message,
+      user,
+    });
+    throw new ShareServiceError('Error deleting shared links', 'BULK_DELETE_ERROR');
+  }
+}
+
+async function createSharedLink(user, conversationId) {
+  if (!user || !conversationId) {
+    throw new ShareServiceError('Missing required parameters', 'INVALID_PARAMS');
+  }
+  try {
+    const [existingShare, conversationMessages] = await Promise.all([
+      SharedLink.findOne({ conversationId, isPublic: true }).select('-_id -__v -user').lean(),
+      getMessages({ conversationId }),
+    ]);
+
+    if (existingShare && existingShare.isPublic) {
+      throw new ShareServiceError('Share already exists', 'SHARE_EXISTS');
+    } else if (existingShare) {
+      await SharedLink.deleteOne({ conversationId });
+    }
+
+    const conversation = await Conversation.findOne({ conversationId }).lean();
+    const title = conversation?.title || 'Untitled';
+
+    const shareId = nanoid();
+    await SharedLink.create({
+      shareId,
+      conversationId,
+      messages: conversationMessages,
+      title,
+      user,
+    });
+
+    return { shareId, conversationId };
+  } catch (error) {
+    logger.error('[createSharedLink] Error creating shared link', {
+      error: error.message,
+      user,
+      conversationId,
+    });
+    throw new ShareServiceError('Error creating shared link', 'SHARE_CREATE_ERROR');
+  }
+}
+
+async function getSharedLink(user, conversationId) {
+  if (!user || !conversationId) {
+    throw new ShareServiceError('Missing required parameters', 'INVALID_PARAMS');
+  }
+
+  try {
+    const share = await SharedLink.findOne({ conversationId, user, isPublic: true })
+      .select('shareId -_id')
+      .lean();
+
+    if (!share) {
+      return { shareId: null, success: false };
+    }
+
+    return { shareId: share.shareId, success: true };
+  } catch (error) {
+    logger.error('[getSharedLink] Error getting shared link', {
+      error: error.message,
+      user,
+      conversationId,
+    });
+    throw new ShareServiceError('Error getting shared link', 'SHARE_FETCH_ERROR');
+  }
+}
+
+async function updateSharedLink(user, shareId) {
+  if (!user || !shareId) {
+    throw new ShareServiceError('Missing required parameters', 'INVALID_PARAMS');
+  }
+
+  try {
+    const share = await SharedLink.findOne({ shareId }).select('-_id -__v -user').lean();
+
+    if (!share) {
+      throw new ShareServiceError('Share not found', 'SHARE_NOT_FOUND');
+    }
+
+    const [updatedMessages] = await Promise.all([
+      getMessages({ conversationId: share.conversationId }),
+    ]);
+
+    const newShareId = nanoid();
+    const update = {
+      messages: updatedMessages,
+      user,
+      shareId: newShareId,
+    };
+
+    const updatedShare = await SharedLink.findOneAndUpdate({ shareId, user }, update, {
+      new: true,
+      upsert: false,
+      runValidators: true,
+    }).lean();
+
+    if (!updatedShare) {
+      throw new ShareServiceError('Share update failed', 'SHARE_UPDATE_ERROR');
+    }
+
+    anonymizeConvo(updatedShare);
+
+    return { shareId: newShareId, conversationId: updatedShare.conversationId };
+  } catch (error) {
+    logger.error('[updateSharedLink] Error updating shared link', {
+      error: error.message,
+      user,
+      shareId,
+    });
+    throw new ShareServiceError(
+      error.code === 'SHARE_UPDATE_ERROR' ? error.message : 'Error updating shared link',
+      error.code || 'SHARE_UPDATE_ERROR',
+    );
+  }
+}
+
+async function deleteSharedLink(user, shareId) {
+  if (!user || !shareId) {
+    throw new ShareServiceError('Missing required parameters', 'INVALID_PARAMS');
+  }
+
+  try {
+    const result = await SharedLink.findOneAndDelete({ shareId, user }).lean();
+
+    if (!result) {
+      return null;
+    }
+
+    return {
+      success: true,
+      shareId,
+      message: 'Share deleted successfully',
+    };
+  } catch (error) {
+    logger.error('[deleteSharedLink] Error deleting shared link', {
+      error: error.message,
+      user,
+      shareId,
+    });
+    throw new ShareServiceError('Error deleting shared link', 'SHARE_DELETE_ERROR');
+  }
+}
+
+module.exports = {
+  getSharedLink,
+  getSharedLinks,
+  createSharedLink,
+  updateSharedLink,
+  deleteSharedLink,
+  getSharedMessages,
+  deleteAllSharedLinks,
+};

api/models/Token.js

@@ -0,0 +1,42 @@
+const { findToken, updateToken, createToken } = require('~/models');
+const { encryptV2 } = require('~/server/utils/crypto');
+
+/**
+ * Handles the OAuth token by creating or updating the token.
+ * @param {object} fields
+ * @param {string} fields.userId - The user's ID.
+ * @param {string} fields.token - The full token to store.
+ * @param {string} fields.identifier - Unique, alternative identifier for the token.
+ * @param {number} fields.expiresIn - The number of seconds until the token expires.
+ * @param {object} fields.metadata - Additional metadata to store with the token.
+ * @param {string} [fields.type="oauth"] - The type of token. Default is 'oauth'.
+ */
+async function handleOAuthToken({
+  token,
+  userId,
+  identifier,
+  expiresIn,
+  metadata,
+  type = 'oauth',
+}) {
+  const encrypedToken = await encryptV2(token);
+  const tokenData = {
+    type,
+    userId,
+    metadata,
+    identifier,
+    token: encrypedToken,
+    expiresIn: parseInt(expiresIn, 10) || 3600,
+  };
+
+  const existingToken = await findToken({ userId, identifier });
+  if (existingToken) {
+    return await updateToken({ identifier }, tokenData);
+  } else {
+    return await createToken(tokenData);
+  }
+}
+
+module.exports = {
+  handleOAuthToken,
+};

api/models/ToolCall.js

@@ -1,4 +1,6 @@
-const { ToolCall } = require('~/db/models');
+const mongoose = require('mongoose');
+
+const ToolCall = require('~/db/models').ToolCall;
 
 /**
  * Create a new tool call

api/models/Transaction.js

@@ -1,7 +1,10 @@
+const mongoose = require('mongoose');
 const { logger } = require('@librechat/data-schemas');
 const { getBalanceConfig } = require('~/server/services/Config');
 const { getMultiplier, getCacheMultiplier } = require('./tx');
-const { Transaction, Balance } = require('~/db/models');
+
+const Transaction = require('~/db/models').Transaction;
+const Balance = require('~/db/models').Balance;
 
 const cancelRate = 1.15;
 

api/models/Transaction.spec.js

@@ -4,7 +4,7 @@ const { spendTokens, spendStructuredTokens } = require('./spendTokens');
 const { getBalanceConfig } = require('~/server/services/Config');
 const { getMultiplier, getCacheMultiplier } = require('./tx');
 const { createTransaction } = require('./Transaction');
-const { Balance } = require('~/db/models');
+const Balance = require('~/db/models').Balance;
 
 // Mock the custom config module so we can control the balance flag.
 jest.mock('~/server/services/Config');

api/models/balanceMethods.js

@@ -1,9 +1,11 @@
+const mongoose = require('mongoose');
 const { logger } = require('@librechat/data-schemas');
 const { ViolationTypes } = require('librechat-data-provider');
 const { createAutoRefillTransaction } = require('./Transaction');
 const { logViolation } = require('~/cache');
 const { getMultiplier } = require('./tx');
-const { Balance } = require('~/db/models');
+
+const Balance = require('~/db/models').Balance;
 
 function isInvalidDate(date) {
   return isNaN(date);

api/models/convoStructure.spec.js

@@ -1,7 +1,8 @@
 const mongoose = require('mongoose');
 const { MongoMemoryServer } = require('mongodb-memory-server');
 const { getMessages, bulkSaveMessages } = require('./Message');
-const { Message } = require('~/db/models');
+
+const Message = require('~/db/models').Message;
 
 // Original version of buildTree function
 function buildTree({ messages, fileMap }) {

api/models/inviteUser.js

@@ -1,6 +1,6 @@
 const mongoose = require('mongoose');
-const { getRandomValues } = require('@librechat/api');
 const { logger, hashToken } = require('@librechat/data-schemas');
+const { getRandomValues } = require('~/server/utils/crypto');
 const { createToken, findToken } = require('~/models');
 
 /**

api/models/spendTokens.spec.js

@@ -2,8 +2,8 @@ const mongoose = require('mongoose');
 const { MongoMemoryServer } = require('mongodb-memory-server');
 const { spendTokens, spendStructuredTokens } = require('./spendTokens');
 const { createTransaction, createAutoRefillTransaction } = require('./Transaction');
-
-require('~/db/models');
+const Transaction = require('~/db/models').Transaction;
+const Balance = require('~/db/models').Balance;
 
 // Mock the logger to prevent console output during tests
 jest.mock('~/config', () => ({
@@ -20,15 +20,10 @@ jest.mock('~/server/services/Config');
 describe('spendTokens', () => {
   let mongoServer;
   let userId;
-  let Transaction;
-  let Balance;
 
   beforeAll(async () => {
     mongoServer = await MongoMemoryServer.create();
     await mongoose.connect(mongoServer.getUri());
-
-    Transaction = mongoose.model('Transaction');
-    Balance = mongoose.model('Balance');
   });
 
   afterAll(async () => {

api/models/tx.js

@@ -78,7 +78,7 @@ const tokenValues = Object.assign(
     'gpt-3.5-turbo-1106': { prompt: 1, completion: 2 },
     'o4-mini': { prompt: 1.1, completion: 4.4 },
     'o3-mini': { prompt: 1.1, completion: 4.4 },
-    o3: { prompt: 2, completion: 8 },
+    o3: { prompt: 10, completion: 40 },
     'o1-mini': { prompt: 1.1, completion: 4.4 },
     'o1-preview': { prompt: 15, completion: 60 },
     o1: { prompt: 15, completion: 60 },

api/models/userMethods.js

@@ -12,10 +12,6 @@ const comparePassword = async (user, candidatePassword) => {
     throw new Error('No user provided');
   }
 
-  if (!user.password) {
-    throw new Error('No password, likely an email first registered via Social/OIDC login');
-  }
-
   return new Promise((resolve, reject) => {
     bcrypt.compare(candidatePassword, user.password, (err, isMatch) => {
       if (err) {

api/package.json

@@ -34,27 +34,27 @@
   },
   "homepage": "https://librechat.ai",
   "dependencies": {
-    "@anthropic-ai/sdk": "^0.52.0",
+    "@anthropic-ai/sdk": "^0.37.0",
     "@aws-sdk/client-s3": "^3.758.0",
     "@aws-sdk/s3-request-presigner": "^3.758.0",
     "@azure/identity": "^4.7.0",
     "@azure/search-documents": "^12.0.0",
     "@azure/storage-blob": "^12.27.0",
-    "@google/generative-ai": "^0.24.0",
+    "@google/generative-ai": "^0.23.0",
     "@googleapis/youtube": "^20.0.0",
     "@keyv/redis": "^4.3.3",
-    "@langchain/community": "^0.3.47",
-    "@langchain/core": "^0.3.60",
-    "@langchain/google-genai": "^0.2.13",
-    "@langchain/google-vertexai": "^0.2.13",
+    "@langchain/community": "^0.3.44",
+    "@langchain/core": "^0.3.57",
+    "@langchain/google-genai": "^0.2.9",
+    "@langchain/google-vertexai": "^0.2.9",
     "@langchain/textsplitters": "^0.1.0",
-    "@librechat/agents": "^2.4.46",
-    "@librechat/api": "*",
+    "@librechat/agents": "^2.4.37",
     "@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",
@@ -81,15 +81,15 @@
     "keyv-file": "^5.1.2",
     "klona": "^2.0.6",
     "librechat-data-provider": "*",
+    "librechat-mcp": "*",
     "lodash": "^4.17.21",
     "meilisearch": "^0.38.0",
     "memorystore": "^1.6.7",
     "mime": "^3.0.0",
     "module-alias": "^2.2.3",
     "mongoose": "^8.12.1",
-    "multer": "^2.0.1",
+    "multer": "^2.0.0",
     "nanoid": "^3.3.7",
-    "node-fetch": "^2.7.0",
     "nodemailer": "^6.9.15",
     "ollama": "^0.5.0",
     "openai": "^4.96.2",
@@ -109,9 +109,8 @@
     "tiktoken": "^1.0.15",
     "traverse": "^0.6.7",
     "ua-parser-js": "^1.0.36",
-    "undici": "^7.10.0",
     "winston": "^3.11.0",
-    "winston-daily-rotate-file": "^5.0.0",
+    "winston-daily-rotate-file": "^4.7.1",
     "youtube-transcript": "^1.2.1",
     "zod": "^3.22.4"
   },

api/server/cleanup.js

@@ -169,6 +169,9 @@ function disposeClient(client) {
       client.isGenerativeModel = null;
     }
     // Properties specific to OpenAIClient
+    if (client.ChatGPTClient) {
+      client.ChatGPTClient = null;
+    }
     if (client.completionsUrl) {
       client.completionsUrl = null;
     }
@@ -217,9 +220,6 @@ function disposeClient(client) {
     if (client.maxResponseTokens) {
       client.maxResponseTokens = null;
     }
-    if (client.processMemory) {
-      client.processMemory = null;
-    }
     if (client.run) {
       // Break circular references in run
       if (client.run.Graph) {

api/server/controllers/AskController.js

@@ -0,0 +1,282 @@
+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 {
-  requestPasswordReset,
-  setOpenIDAuthTokens,
+  registerUser,
   resetPassword,
   setAuthTokens,
-  registerUser,
+  requestPasswordReset,
+  setOpenIDAuthTokens,
 } = 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/Balance.js

@@ -1,4 +1,6 @@
-const { Balance } = require('~/db/models');
+const mongoose = require('mongoose');
+
+const Balance = require('~/db/models').Balance;
 
 async function balanceController(req, res) {
   const balanceData = await Balance.findOne(

api/server/controllers/EditController.js

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

api/server/controllers/PluginController.js

@@ -1,11 +1,9 @@
-const { logger } = require('@librechat/data-schemas');
 const { CacheKeys, AuthType } = require('librechat-data-provider');
-const { getCustomConfig, getCachedTools } = require('~/server/services/Config');
 const { getToolkitKey } = require('~/server/services/ToolService');
-const { getMCPManager, getFlowStateManager } = require('~/config');
+const { getCustomConfig } = require('~/server/services/Config');
 const { availableTools } = require('~/app/clients/tools');
+const { getMCPManager } = require('~/config');
 const { getLogStores } = require('~/cache');
-const { Constants } = require('librechat-data-provider');
 
 /**
  * Filters out duplicate plugins from the list of plugins.
@@ -86,45 +84,6 @@ const getAvailablePluginsController = async (req, res) => {
   }
 };
 
-function createServerToolsCallback() {
-  /**
-   * @param {string} serverName
-   * @param {TPlugin[] | null} serverTools
-   */
-  return async function (serverName, serverTools) {
-    try {
-      const mcpToolsCache = getLogStores(CacheKeys.MCP_TOOLS);
-      if (!serverName || !mcpToolsCache) {
-        return;
-      }
-      await mcpToolsCache.set(serverName, serverTools);
-      logger.debug(`MCP tools for ${serverName} added to cache.`);
-    } catch (error) {
-      logger.error('Error retrieving MCP tools from cache:', error);
-    }
-  };
-}
-
-function createGetServerTools() {
-  /**
-   * Retrieves cached server tools
-   * @param {string} serverName
-   * @returns {Promise<TPlugin[] | null>}
-   */
-  return async function (serverName) {
-    try {
-      const mcpToolsCache = getLogStores(CacheKeys.MCP_TOOLS);
-      if (!mcpToolsCache) {
-        return null;
-      }
-      return await mcpToolsCache.get(serverName);
-    } catch (error) {
-      logger.error('Error retrieving MCP tools from cache:', error);
-      return null;
-    }
-  };
-}
-
 /**
  * Retrieves and returns a list of available tools, either from a cache or by reading a plugin manifest file.
  *
@@ -150,16 +109,7 @@ const getAvailableTools = async (req, res) => {
     const customConfig = await getCustomConfig();
     if (customConfig?.mcpServers != null) {
       const mcpManager = getMCPManager();
-      const flowsCache = getLogStores(CacheKeys.FLOWS);
-      const flowManager = flowsCache ? getFlowStateManager(flowsCache) : null;
-      const serverToolsCallback = createServerToolsCallback();
-      const getServerTools = createGetServerTools();
-      const mcpTools = await mcpManager.loadManifestTools({
-        flowManager,
-        serverToolsCallback,
-        getServerTools,
-      });
-      pluginManifest = [...mcpTools, ...pluginManifest];
+      pluginManifest = await mcpManager.loadManifestTools(pluginManifest);
     }
 
     /** @type {TPlugin[]} */
@@ -173,57 +123,17 @@ const getAvailableTools = async (req, res) => {
       }
     });
 
-    const toolDefinitions = await getCachedTools({ includeGlobal: true });
+    const toolDefinitions = req.app.locals.availableTools;
+    const tools = authenticatedPlugins.filter(
+      (plugin) =>
+        toolDefinitions[plugin.pluginKey] !== undefined ||
+        (plugin.toolkit === true &&
+          Object.keys(toolDefinitions).some((key) => getToolkitKey(key) === plugin.pluginKey)),
+    );
 
-    const toolsOutput = [];
-    for (const plugin of authenticatedPlugins) {
-      const isToolDefined = toolDefinitions[plugin.pluginKey] !== undefined;
-      const isToolkit =
-        plugin.toolkit === true &&
-        Object.keys(toolDefinitions).some((key) => getToolkitKey(key) === plugin.pluginKey);
-
-      if (!isToolDefined && !isToolkit) {
-        continue;
-      }
-
-      const toolToAdd = { ...plugin };
-
-      if (!plugin.pluginKey.includes(Constants.mcp_delimiter)) {
-        toolsOutput.push(toolToAdd);
-        continue;
-      }
-
-      const parts = plugin.pluginKey.split(Constants.mcp_delimiter);
-      const serverName = parts[parts.length - 1];
-      const serverConfig = customConfig?.mcpServers?.[serverName];
-
-      if (!serverConfig?.customUserVars) {
-        toolsOutput.push(toolToAdd);
-        continue;
-      }
-
-      const customVarKeys = Object.keys(serverConfig.customUserVars);
-
-      if (customVarKeys.length === 0) {
-        toolToAdd.authConfig = [];
-        toolToAdd.authenticated = true;
-      } else {
-        toolToAdd.authConfig = Object.entries(serverConfig.customUserVars).map(([key, value]) => ({
-          authField: key,
-          label: value.title || key,
-          description: value.description || '',
-        }));
-        toolToAdd.authenticated = false;
-      }
-
-      toolsOutput.push(toolToAdd);
-    }
-
-    const finalTools = filterUniquePlugins(toolsOutput);
-    await cache.set(CacheKeys.TOOLS, finalTools);
-    res.status(200).json(finalTools);
+    await cache.set(CacheKeys.TOOLS, tools);
+    res.status(200).json(tools);
   } catch (error) {
-    logger.error('[getAvailableTools]', error);
     res.status(500).json({ message: error.message });
   }
 };

api/server/controllers/TwoFactorController.js

@@ -1,4 +1,3 @@
-const { encryptV3 } = require('@librechat/api');
 const { logger } = require('@librechat/data-schemas');
 const {
   verifyTOTP,
@@ -8,6 +7,7 @@ const {
   generateBackupCodes,
 } = require('~/server/services/twoFactorService');
 const { getUserById, updateUser } = require('~/models');
+const { encryptV3 } = require('~/server/utils/crypto');
 
 const safeAppTitle = (process.env.APP_TITLE || 'LibreChat').replace(/\s+/g, '');
 

api/server/controllers/UserController.js

@@ -1,6 +1,5 @@
 const {
   Tools,
-  Constants,
   FileSources,
   webSearchKeys,
   extractWebSearchEnvVars,
@@ -21,10 +20,12 @@ const { updateUserPluginsService, deleteUserKey } = require('~/server/services/U
 const { verifyEmail, resendVerificationEmail } = require('~/server/services/AuthService');
 const { needsRefresh, getNewS3URL } = require('~/server/services/Files/S3/crud');
 const { processDeleteRequest } = require('~/server/services/Files/process');
-const { Transaction, Balance, User } = require('~/db/models');
+const { deleteAllSharedLinks } = require('~/models/Share');
 const { deleteToolCalls } = require('~/models/ToolCall');
-const { deleteAllSharedLinks } = require('~/models');
-const { getMCPManager } = require('~/config');
+
+const Transaction = require('~/db/models').Transaction;
+const Balance = require('~/db/models').Balance;
+const User = require('~/db/models').User;
 
 const getUserController = async (req, res) => {
   /** @type {MongoUser} */
@@ -104,22 +105,10 @@ const updateUserPluginsController = async (req, res) => {
     }
 
     let keys = Object.keys(auth);
-    const values = Object.values(auth); // Used in 'install' block
-
-    const isMCPTool = pluginKey.startsWith('mcp_') || pluginKey.includes(Constants.mcp_delimiter);
-
-    // Early exit condition:
-    // If keys are empty (meaning auth: {} was likely sent for uninstall, or auth was empty for install)
-    // AND it's not web_search (which has special key handling to populate `keys` for uninstall)
-    // AND it's NOT (an uninstall action FOR an MCP tool - we need to proceed for this case to clear all its auth)
-    // THEN return.
-    if (
-      keys.length === 0 &&
-      pluginKey !== Tools.web_search &&
-      !(action === 'uninstall' && isMCPTool)
-    ) {
+    if (keys.length === 0 && pluginKey !== Tools.web_search) {
       return res.status(200).send();
     }
+    const values = Object.values(auth);
 
     /** @type {number} */
     let status = 200;
@@ -146,53 +135,16 @@ const updateUserPluginsController = async (req, res) => {
         }
       }
     } else if (action === 'uninstall') {
-      // const isMCPTool was defined earlier
-      if (isMCPTool && keys.length === 0) {
-        // This handles the case where auth: {} is sent for an MCP tool uninstall.
-        // It means "delete all credentials associated with this MCP pluginKey".
-        authService = await deleteUserPluginAuth(user.id, null, true, pluginKey);
+      for (let i = 0; i < keys.length; i++) {
+        authService = await deleteUserPluginAuth(user.id, keys[i]);
         if (authService instanceof Error) {
-          logger.error(
-            `[authService] Error deleting all auth for MCP tool ${pluginKey}:`,
-            authService,
-          );
+          logger.error('[authService]', authService);
           ({ status, message } = authService);
         }
-      } else {
-        // This handles:
-        // 1. Web_search uninstall (keys will be populated with all webSearchKeys if auth was {}).
-        // 2. Other tools uninstall (if keys were provided).
-        // 3. MCP tool uninstall if specific keys were provided in `auth` (not current frontend behavior).
-        // If keys is empty for non-MCP tools (and not web_search), this loop won't run, and nothing is deleted.
-        for (let i = 0; i < keys.length; i++) {
-          authService = await deleteUserPluginAuth(user.id, keys[i]); // Deletes by authField name
-          if (authService instanceof Error) {
-            logger.error('[authService] Error deleting specific auth key:', authService);
-            ({ status, message } = authService);
-          }
-        }
       }
     }
 
     if (status === 200) {
-      // If auth was updated successfully, disconnect MCP sessions as they might use these credentials
-      if (pluginKey.startsWith(Constants.mcp_prefix)) {
-        try {
-          const mcpManager = getMCPManager(user.id);
-          if (mcpManager) {
-            logger.info(
-              `[updateUserPluginsController] Disconnecting MCP connections for user ${user.id} after plugin auth update for ${pluginKey}.`,
-            );
-            await mcpManager.disconnectUserConnections(user.id);
-          }
-        } catch (disconnectError) {
-          logger.error(
-            `[updateUserPluginsController] Error disconnecting MCP connections for user ${user.id} after plugin auth update:`,
-            disconnectError,
-          );
-          // Do not fail the request for this, but log it.
-        }
-      }
       return res.status(status).send();
     }
 
@@ -214,11 +166,7 @@ const deleteUserController = async (req, res) => {
     await Balance.deleteMany({ user: user._id }); // delete user balances
     await deletePresets(user.id); // delete user presets
     /* TODO: Delete Assistant Threads */
-    try {
-      await deleteConvos(user.id); // delete user convos
-    } catch (error) {
-      logger.error('[deleteUserController] Error deleting user convos, likely no convos', error);
-    }
+    await deleteConvos(user.id); // delete user convos
     await deleteUserPluginAuth(user.id, null, true); // delete user plugin auth
     await deleteUserById(user.id); // delete user
     await deleteAllSharedLinks(user.id); // delete user shared links

api/server/controllers/agents/callbacks.js

@@ -1,6 +1,4 @@
 const { nanoid } = require('nanoid');
-const { sendEvent } = require('@librechat/api');
-const { logger } = require('@librechat/data-schemas');
 const { Tools, StepTypes, FileContext } = require('librechat-data-provider');
 const {
   EnvVar,
@@ -14,6 +12,7 @@ const {
 const { processCodeOutput } = require('~/server/services/Files/Code/process');
 const { loadAuthValues } = require('~/server/services/Tools/credentials');
 const { saveBase64Image } = require('~/server/services/Files/process');
+const { logger, sendEvent } = require('~/config');
 
 class ModelEndHandler {
   /**
@@ -241,7 +240,9 @@ function createToolEndCallback({ req, res, artifactPromises }) {
     if (output.artifact[Tools.web_search]) {
       artifactPromises.push(
         (async () => {
+          const name = `${output.name}_${output.tool_call_id}_${nanoid()}`;
           const attachment = {
+            name,
             type: Tools.web_search,
             messageId: metadata.run_id,
             toolCallId: output.tool_call_id,

api/server/controllers/agents/client.js

@@ -1,15 +1,15 @@
+// const { HttpsProxyAgent } = require('https-proxy-agent');
+// const {
+// Constants,
+// ImageDetail,
+// EModelEndpoint,
+// resolveHeaders,
+// validateVisionModel,
+// mapModelToAzureConfig,
+// } = require('librechat-data-provider');
 require('events').EventEmitter.defaultMaxListeners = 100;
-const { logger } = require('@librechat/data-schemas');
-const {
-  sendEvent,
-  createRun,
-  Tokenizer,
-  memoryInstructions,
-  createMemoryProcessor,
-} = require('@librechat/api');
 const {
   Callback,
-  Providers,
   GraphEvents,
   formatMessage,
   formatAgentMessages,
@@ -19,41 +19,25 @@ const {
 } = require('@librechat/agents');
 const {
   Constants,
-  Permissions,
   VisionModes,
   ContentTypes,
   EModelEndpoint,
   KnownEndpoints,
-  PermissionTypes,
   isAgentsEndpoint,
   AgentCapabilities,
   bedrockInputSchema,
   removeNullishValues,
 } = require('librechat-data-provider');
-const { DynamicStructuredTool } = require('@langchain/core/tools');
-const { getBufferString, HumanMessage } = require('@langchain/core/messages');
-const { createGetMCPAuthMap, checkCapability } = require('~/server/services/Config');
+const { getCustomEndpointConfig, 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 { getBufferString, HumanMessage } = require('@langchain/core/messages');
 const { encodeAndFormat } = require('~/server/services/Files/images/encode');
-const { getProviderConfig } = require('~/server/services/Endpoints');
-const { checkAccess } = require('~/server/middleware/roles/access');
+const initOpenAI = require('~/server/services/Endpoints/openAI/initialize');
+const Tokenizer = require('~/server/services/Tokenizer');
 const BaseClient = require('~/app/clients/BaseClient');
-const { loadAgent } = require('~/models/Agent');
-const { getMCPManager } = require('~/config');
-
-const omitTitleOptions = new Set([
-  'stream',
-  'thinking',
-  'streaming',
-  'clientOptions',
-  'thinkingConfig',
-  'thinkingBudget',
-  'includeThoughts',
-  'maxOutputTokens',
-]);
+const { logger, sendEvent } = require('~/config');
+const { createRun } = require('./run');
 
 /**
  * @param {ServerRequest} req
@@ -73,8 +57,12 @@ const legacyContentEndpoints = new Set([KnownEndpoints.groq, KnownEndpoints.deep
 
 const noSystemModelRegex = [/\b(o1-preview|o1-mini|amazon\.titan-text)\b/gi];
 
+// const { processMemory, memoryInstructions } = require('~/server/services/Endpoints/agents/memory');
+// const { getFormattedMemories } = require('~/models/Memory');
+// const { getCurrentDateTime } = require('~/utils');
+
 function createTokenCounter(encoding) {
-  return function (message) {
+  return (message) => {
     const countTokens = (text) => Tokenizer.getTokenCount(text, encoding);
     return getTokenCountForMessage(message, countTokens);
   };
@@ -135,8 +123,6 @@ class AgentClient extends BaseClient {
     this.usage;
     /** @type {Record<string, number>} */
     this.indexTokenCountMap = {};
-    /** @type {(messages: BaseMessage[]) => Promise<void>} */
-    this.processMemory;
   }
 
   /**
@@ -151,10 +137,55 @@ class AgentClient extends BaseClient {
   }
 
   /**
-   * `AgentClient` is not opinionated about vision requests, so we don't do anything here
+   *
+   * Checks if the model is a vision model based on request attachments and sets the appropriate options:
+   * - Sets `this.modelOptions.model` to `gpt-4-vision-preview` if the request is a vision request.
+   * - Sets `this.isVisionModel` to `true` if vision request.
+   * - Deletes `this.modelOptions.stop` if vision request.
    * @param {MongoFile[]} attachments
    */
-  checkVisionRequest() {}
+  checkVisionRequest(attachments) {
+    // if (!attachments) {
+    //   return;
+    // }
+    // const availableModels = this.options.modelsConfig?.[this.options.endpoint];
+    // if (!availableModels) {
+    //   return;
+    // }
+    // let visionRequestDetected = false;
+    // for (const file of attachments) {
+    //   if (file?.type?.includes('image')) {
+    //     visionRequestDetected = true;
+    //     break;
+    //   }
+    // }
+    // if (!visionRequestDetected) {
+    //   return;
+    // }
+    // this.isVisionModel = validateVisionModel({ model: this.modelOptions.model, availableModels });
+    // if (this.isVisionModel) {
+    //   delete this.modelOptions.stop;
+    //   return;
+    // }
+    // for (const model of availableModels) {
+    //   if (!validateVisionModel({ model, availableModels })) {
+    //     continue;
+    //   }
+    //   this.modelOptions.model = model;
+    //   this.isVisionModel = true;
+    //   delete this.modelOptions.stop;
+    //   return;
+    // }
+    // if (!availableModels.includes(this.defaultVisionModel)) {
+    //   return;
+    // }
+    // if (!validateVisionModel({ model: this.defaultVisionModel, availableModels })) {
+    //   return;
+    // }
+    // this.modelOptions.model = this.defaultVisionModel;
+    // this.isVisionModel = true;
+    // delete this.modelOptions.stop;
+  }
 
   getSaveOptions() {
     // TODO:
@@ -238,6 +269,24 @@ class AgentClient extends BaseClient {
       .filter(Boolean)
       .join('\n')
       .trim();
+    // this.systemMessage = getCurrentDateTime();
+    // const { withKeys, withoutKeys } = await getFormattedMemories({
+    //   userId: this.options.req.user.id,
+    // });
+    // processMemory({
+    //   userId: this.options.req.user.id,
+    //   message: this.options.req.body.text,
+    //   parentMessageId,
+    //   memory: withKeys,
+    //   thread_id: this.conversationId,
+    // }).catch((error) => {
+    //   logger.error('Memory Agent failed to process memory', error);
+    // });
+
+    // this.systemMessage += '\n\n' + memoryInstructions;
+    // if (withoutKeys) {
+    //   this.systemMessage += `\n\n# Existing memory about the user:\n${withoutKeys}`;
+    // }
 
     if (this.options.attachments) {
       const attachments = await this.options.attachments;
@@ -321,37 +370,6 @@ class AgentClient extends BaseClient {
       systemContent = this.augmentedPrompt + systemContent;
     }
 
-    // Inject MCP server instructions if available
-    const ephemeralAgent = this.options.req.body.ephemeralAgent;
-    let mcpServers = [];
-
-    // Check for ephemeral agent MCP servers
-    if (ephemeralAgent && ephemeralAgent.mcp && ephemeralAgent.mcp.length > 0) {
-      mcpServers = ephemeralAgent.mcp;
-    }
-    // Check for regular agent MCP tools
-    else if (this.options.agent && this.options.agent.tools) {
-      mcpServers = this.options.agent.tools
-        .filter(
-          (tool) =>
-            tool instanceof DynamicStructuredTool && tool.name.includes(Constants.mcp_delimiter),
-        )
-        .map((tool) => tool.name.split(Constants.mcp_delimiter).pop())
-        .filter(Boolean);
-    }
-
-    if (mcpServers.length > 0) {
-      try {
-        const mcpInstructions = getMCPManager().formatInstructionsForContext(mcpServers);
-        if (mcpInstructions) {
-          systemContent = [systemContent, mcpInstructions].filter(Boolean).join('\n\n');
-          logger.debug('[AgentClient] Injected MCP instructions for servers:', mcpServers);
-        }
-      } catch (error) {
-        logger.error('[AgentClient] Failed to inject MCP instructions:', error);
-      }
-    }
-
     if (systemContent) {
       this.options.agent.instructions = systemContent;
     }
@@ -381,150 +399,9 @@ class AgentClient extends BaseClient {
       opts.getReqData({ promptTokens });
     }
 
-    const withoutKeys = await this.useMemory();
-    if (withoutKeys) {
-      systemContent += `${memoryInstructions}\n\n# Existing memory about the user:\n${withoutKeys}`;
-    }
-
-    if (systemContent) {
-      this.options.agent.instructions = systemContent;
-    }
-
     return result;
   }
 
-  /**
-   * @returns {Promise<string | undefined>}
-   */
-  async useMemory() {
-    const user = this.options.req.user;
-    if (user.personalization?.memories === false) {
-      return;
-    }
-    const hasAccess = await checkAccess(user, PermissionTypes.MEMORIES, [Permissions.USE]);
-
-    if (!hasAccess) {
-      logger.debug(
-        `[api/server/controllers/agents/client.js #useMemory] User ${user.id} does not have USE permission for memories`,
-      );
-      return;
-    }
-    /** @type {TCustomConfig['memory']} */
-    const memoryConfig = this.options.req?.app?.locals?.memory;
-    if (!memoryConfig || memoryConfig.disabled === true) {
-      return;
-    }
-
-    /** @type {Agent} */
-    let prelimAgent;
-    const allowedProviders = new Set(
-      this.options.req?.app?.locals?.[EModelEndpoint.agents]?.allowedProviders,
-    );
-    try {
-      if (memoryConfig.agent?.id != null && memoryConfig.agent.id !== this.options.agent.id) {
-        prelimAgent = await loadAgent({
-          req: this.options.req,
-          agent_id: memoryConfig.agent.id,
-          endpoint: EModelEndpoint.agents,
-        });
-      } else if (
-        memoryConfig.agent?.id == null &&
-        memoryConfig.agent?.model != null &&
-        memoryConfig.agent?.provider != null
-      ) {
-        prelimAgent = { id: Constants.EPHEMERAL_AGENT_ID, ...memoryConfig.agent };
-      }
-    } catch (error) {
-      logger.error(
-        '[api/server/controllers/agents/client.js #useMemory] Error loading agent for memory',
-        error,
-      );
-    }
-
-    const agent = await initializeAgent({
-      req: this.options.req,
-      res: this.options.res,
-      agent: prelimAgent,
-      allowedProviders,
-    });
-
-    if (!agent) {
-      logger.warn(
-        '[api/server/controllers/agents/client.js #useMemory] No agent found for memory',
-        memoryConfig,
-      );
-      return;
-    }
-
-    const llmConfig = Object.assign(
-      {
-        provider: agent.provider,
-        model: agent.model,
-      },
-      agent.model_parameters,
-    );
-
-    /** @type {import('@librechat/api').MemoryConfig} */
-    const config = {
-      validKeys: memoryConfig.validKeys,
-      instructions: agent.instructions,
-      llmConfig,
-      tokenLimit: memoryConfig.tokenLimit,
-    };
-
-    const userId = this.options.req.user.id + '';
-    const messageId = this.responseMessageId + '';
-    const conversationId = this.conversationId + '';
-    const [withoutKeys, processMemory] = await createMemoryProcessor({
-      userId,
-      config,
-      messageId,
-      conversationId,
-      memoryMethods: {
-        setMemory,
-        deleteMemory,
-        getFormattedMemories,
-      },
-      res: this.options.res,
-    });
-
-    this.processMemory = processMemory;
-    return withoutKeys;
-  }
-
-  /**
-   * @param {BaseMessage[]} messages
-   * @returns {Promise<void | (TAttachment | null)[]>}
-   */
-  async runMemory(messages) {
-    try {
-      if (this.processMemory == null) {
-        return;
-      }
-      /** @type {TCustomConfig['memory']} */
-      const memoryConfig = this.options.req?.app?.locals?.memory;
-      const messageWindowSize = memoryConfig?.messageWindowSize ?? 5;
-
-      let messagesToProcess = [...messages];
-      if (messages.length > messageWindowSize) {
-        for (let i = messages.length - messageWindowSize; i >= 0; i--) {
-          const potentialWindow = messages.slice(i, i + messageWindowSize);
-          if (potentialWindow[0]?.role === 'user') {
-            messagesToProcess = [...potentialWindow];
-            break;
-          }
-        }
-
-        if (messagesToProcess.length === messages.length) {
-          messagesToProcess = [...messages.slice(-messageWindowSize)];
-        }
-      }
-      return await this.processMemory(messagesToProcess);
-    } catch (error) {
-      logger.error('Memory Agent failed to process memory', error);
-    }
-  }
-
   /** @type {sendCompletion} */
   async sendCompletion(payload, opts = {}) {
     await this.chatCompletion({
@@ -667,13 +544,100 @@ class AgentClient extends BaseClient {
     let config;
     /** @type {ReturnType<createRun>} */
     let run;
-    /** @type {Promise<(TAttachment | null)[] | undefined>} */
-    let memoryPromise;
     try {
       if (!abortController) {
         abortController = new AbortController();
       }
 
+      // if (this.options.headers) {
+      //   opts.defaultHeaders = { ...opts.defaultHeaders, ...this.options.headers };
+      // }
+
+      // if (this.options.proxy) {
+      //   opts.httpAgent = new HttpsProxyAgent(this.options.proxy);
+      // }
+
+      // if (this.isVisionModel) {
+      //   modelOptions.max_tokens = 4000;
+      // }
+
+      // /** @type {TAzureConfig | undefined} */
+      // const azureConfig = this.options?.req?.app?.locals?.[EModelEndpoint.azureOpenAI];
+
+      // if (
+      //   (this.azure && 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.defaultHeaders = 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 (this.azure || this.options.azure) {
+      //   /* Azure Bug, extremely short default `max_tokens` response */
+      //   if (!modelOptions.max_tokens && modelOptions.model === 'gpt-4-vision-preview') {
+      //     modelOptions.max_tokens = 4000;
+      //   }
+
+      //   /* Azure does not accept `model` in the body, so we need to remove it. */
+      //   delete modelOptions.model;
+
+      //   opts.baseURL = this.langchainProxy
+      //     ? constructAzureURL({
+      //       baseURL: this.langchainProxy,
+      //       azureOptions: this.azure,
+      //     })
+      //     : this.azureEndpoint.split(/(?<!\/)\/(chat|completion)\//)[0];
+
+      //   opts.defaultQuery = { 'api-version': this.azure.azureOpenAIApiVersion };
+      //   opts.defaultHeaders = { ...opts.defaultHeaders, 'api-key': this.apiKey };
+      // }
+
+      // if (process.env.OPENAI_ORGANIZATION) {
+      //   opts.organization = process.env.OPENAI_ORGANIZATION;
+      // }
+
+      // if (this.options.addParams && typeof this.options.addParams === 'object') {
+      //   modelOptions = {
+      //     ...modelOptions,
+      //     ...this.options.addParams,
+      //   };
+      //   logger.debug('[api/server/controllers/agents/client.js #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('[api/server/controllers/agents/client.js #chatCompletion] dropped params', {
+      //     dropParams: this.options.dropParams,
+      //     modelOptions,
+      //   });
+      // }
+
       /** @type {TCustomConfig['endpoints']['agents']} */
       const agentsEConfig = this.options.req.app.locals[EModelEndpoint.agents];
 
@@ -683,16 +647,13 @@ class AgentClient extends BaseClient {
           last_agent_index: this.agentConfigs?.size ?? 0,
           user_id: this.user ?? this.options.req.user?.id,
           hide_sequential_outputs: this.options.agent.hide_sequential_outputs,
-          user: this.options.req.user,
         },
-        recursionLimit: agentsEConfig?.recursionLimit ?? 25,
+        recursionLimit: agentsEConfig?.recursionLimit,
         signal: abortController.signal,
         streamMode: 'values',
         version: 'v2',
       };
 
-      const getUserMCPAuthMap = await createGetMCPAuthMap();
-
       const toolSet = new Set((this.options.agent.tools ?? []).map((tool) => tool && tool.name));
       let { messages: initialMessages, indexTokenCountMap } = formatAgentMessages(
         payload,
@@ -773,10 +734,6 @@ class AgentClient extends BaseClient {
           messages = addCacheControl(messages);
         }
 
-        if (i === 0) {
-          memoryPromise = this.runMemory(messages);
-        }
-
         run = await createRun({
           agent,
           req: this.options.req,
@@ -812,23 +769,10 @@ class AgentClient extends BaseClient {
           run.Graph.contentData = contentData;
         }
 
-        try {
-          if (getUserMCPAuthMap) {
-            config.configurable.userMCPAuthMap = await getUserMCPAuthMap({
-              tools: agent.tools,
-              userId: this.options.req.user.id,
-            });
-          }
-        } catch (err) {
-          logger.error(
-            `[api/server/controllers/agents/client.js #chatCompletion] Error getting custom user vars for agent ${agent.id}`,
-            err,
-          );
-        }
-
+        const encoding = this.getEncoding();
         await run.processStream({ messages }, config, {
           keepContent: i !== 0,
-          tokenCounter: createTokenCounter(this.getEncoding()),
+          tokenCounter: createTokenCounter(encoding),
           indexTokenCountMap: currentIndexCountMap,
           maxContextTokens: agent.maxContextTokens,
           callbacks: {
@@ -943,12 +887,6 @@ class AgentClient extends BaseClient {
       });
 
       try {
-        if (memoryPromise) {
-          const attachments = await memoryPromise;
-          if (attachments && attachments.length > 0) {
-            this.artifactPromises.push(...attachments);
-          }
-        }
         await this.recordCollectedUsage({ context: 'message' });
       } catch (err) {
         logger.error(
@@ -957,12 +895,6 @@ class AgentClient extends BaseClient {
         );
       }
     } catch (err) {
-      if (memoryPromise) {
-        const attachments = await memoryPromise;
-        if (attachments && attachments.length > 0) {
-          this.artifactPromises.push(...attachments);
-        }
-      }
       logger.error(
         '[api/server/controllers/agents/client.js #sendCompletion] Operation aborted',
         err,
@@ -991,26 +923,23 @@ class AgentClient extends BaseClient {
       throw new Error('Run not initialized');
     }
     const { handleLLMEnd, collected: collectedMetadata } = createMetadataAggregator();
-    const { req, res, agent } = this.options;
-    const endpoint = agent.endpoint;
-
+    const endpoint = this.options.agent.endpoint;
+    const { req, res } = this.options;
     /** @type {import('@librechat/agents').ClientOptions} */
     let clientOptions = {
       maxTokens: 75,
-      model: agent.model_parameters.model,
     };
-
-    const { getOptions, overrideProvider, customEndpointConfig } =
-      await getProviderConfig(endpoint);
-
-    /** @type {TEndpoint | undefined} */
-    const endpointConfig = req.app.locals[endpoint] ?? customEndpointConfig;
+    let endpointConfig = req.app.locals[endpoint];
     if (!endpointConfig) {
-      logger.warn(
-        '[api/server/controllers/agents/client.js #titleConvo] Error getting endpoint config',
-      );
+      try {
+        endpointConfig = await getCustomEndpointConfig(endpoint);
+      } catch (err) {
+        logger.error(
+          '[api/server/controllers/agents/client.js #titleConvo] Error getting custom endpoint config',
+          err,
+        );
+      }
     }
-
     if (
       endpointConfig &&
       endpointConfig.titleModel &&
@@ -1018,50 +947,30 @@ 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 &&
-      options.llmConfig?.azureOpenAIApiInstanceName == null
+      clientOptions.model &&
+      this.options.agent.model_parameters.model !== clientOptions.model
     ) {
-      provider = Providers.OPENAI;
+      clientOptions =
+        (
+          await initOpenAI({
+            req,
+            res,
+            optionsOnly: true,
+            overrideModel: clientOptions.model,
+            overrideEndpoint: endpoint,
+            endpointOption: {
+              model_parameters: clientOptions,
+            },
+          })
+        )?.llmConfig ?? clientOptions;
     }
-
-    /** @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) {
+    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,
@@ -1079,10 +988,8 @@ class AgentClient extends BaseClient {
         let input_tokens, output_tokens;
 
         if (item.usage) {
-          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;
+          input_tokens = item.usage.input_tokens || item.usage.inputTokens;
+          output_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('@librechat/data-schemas');
+const { logger } = require('~/config');
+const getLogStores = require('~/cache/getLogStores');
 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 getLogStores = require('~/cache/getLogStores');
+const { sendResponse } = require('~/server/utils');
 
 /**
  * @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

@@ -0,0 +1,106 @@
+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

@@ -1,5 +1,3 @@
-const { sendEvent } = require('@librechat/api');
-const { logger } = require('@librechat/data-schemas');
 const { Constants } = require('librechat-data-provider');
 const {
   handleAbortError,
@@ -7,7 +5,9 @@ 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 {
@@ -206,7 +206,7 @@ const AgentController = async (req, res, next, initializeClient, addTitle) => {
       // Create a new response object with minimal copies
       const finalResponse = { ...response };
 
-      sendEvent(res, {
+      sendMessage(res, {
         final: true,
         conversation,
         title: conversation.title,
@@ -228,7 +228,7 @@ const AgentController = async (req, res, next, initializeClient, addTitle) => {
     // Save user message if needed
     if (!client.skipSaveUserMessage) {
       await saveMessage(req, userMessage, {
-        context: "api/server/controllers/agents/request.js - don't skip saving user message",
+        context: 'api/server/controllers/agents/request.js - don\'t skip saving user message',
       });
     }
 

api/server/controllers/agents/run.js

@@ -0,0 +1,94 @@
+const { Run, Providers } = require('@librechat/agents');
+const { providerEndpointMap, KnownEndpoints } = require('librechat-data-provider');
+
+/**
+ * @typedef {import('@librechat/agents').t} t
+ * @typedef {import('@librechat/agents').StandardGraphConfig} StandardGraphConfig
+ * @typedef {import('@librechat/agents').StreamEventData} StreamEventData
+ * @typedef {import('@librechat/agents').EventHandler} EventHandler
+ * @typedef {import('@librechat/agents').GraphEvents} GraphEvents
+ * @typedef {import('@librechat/agents').LLMConfig} LLMConfig
+ * @typedef {import('@librechat/agents').IState} IState
+ */
+
+const customProviders = new Set([
+  Providers.XAI,
+  Providers.OLLAMA,
+  Providers.DEEPSEEK,
+  Providers.OPENROUTER,
+]);
+
+/**
+ * Creates a new Run instance with custom handlers and configuration.
+ *
+ * @param {Object} options - The options for creating the Run instance.
+ * @param {ServerRequest} [options.req] - The server request.
+ * @param {string | undefined} [options.runId] - Optional run ID; otherwise, a new run ID will be generated.
+ * @param {Agent} options.agent - The agent for this run.
+ * @param {AbortSignal} options.signal - The signal for this run.
+ * @param {Record<GraphEvents, EventHandler> | undefined} [options.customHandlers] - Custom event handlers.
+ * @param {boolean} [options.streaming=true] - Whether to use streaming.
+ * @param {boolean} [options.streamUsage=true] - Whether to stream usage information.
+ * @returns {Promise<Run<IState>>} A promise that resolves to a new Run instance.
+ */
+async function createRun({
+  runId,
+  agent,
+  signal,
+  customHandlers,
+  streaming = true,
+  streamUsage = true,
+}) {
+  const provider = providerEndpointMap[agent.provider] ?? agent.provider;
+  /** @type {LLMConfig} */
+  const llmConfig = Object.assign(
+    {
+      provider,
+      streaming,
+      streamUsage,
+    },
+    agent.model_parameters,
+  );
+
+  /** Resolves issues with new OpenAI usage field */
+  if (
+    customProviders.has(agent.provider) ||
+    (agent.provider === Providers.OPENAI && agent.endpoint !== agent.provider)
+  ) {
+    llmConfig.streamUsage = false;
+    llmConfig.usage = true;
+  }
+
+  /** @type {'reasoning_content' | 'reasoning'} */
+  let reasoningKey;
+  if (
+    llmConfig.configuration?.baseURL?.includes(KnownEndpoints.openrouter) ||
+    (agent.endpoint && agent.endpoint.toLowerCase().includes(KnownEndpoints.openrouter))
+  ) {
+    reasoningKey = 'reasoning';
+  }
+
+  /** @type {StandardGraphConfig} */
+  const graphConfig = {
+    signal,
+    llmConfig,
+    reasoningKey,
+    tools: agent.tools,
+    instructions: agent.instructions,
+    additional_instructions: agent.additional_instructions,
+    // toolEnd: agent.end_after_tools,
+  };
+
+  // TEMPORARY FOR TESTING
+  if (agent.provider === Providers.ANTHROPIC || agent.provider === Providers.BEDROCK) {
+    graphConfig.streamBuffer = 2000;
+  }
+
+  return Run.create({
+    runId,
+    graphConfig,
+    customHandlers,
+  });
+}
+
+module.exports = { createRun };

api/server/controllers/agents/v1.js

@@ -1,9 +1,9 @@
 const fs = require('fs').promises;
 const { nanoid } = require('nanoid');
-const { logger } = require('@librechat/data-schemas');
 const {
   Tools,
   Constants,
+  FileContext,
   FileSources,
   SystemRoles,
   EToolResources,
@@ -16,16 +16,15 @@ const {
   deleteAgent,
   getListAgents,
 } = require('~/models/Agent');
+const { uploadImageBuffer, filterFile } = require('~/server/services/Files/process');
 const { getStrategyFunctions } = require('~/server/services/Files/strategies');
-const { resizeAvatar } = require('~/server/services/Files/images/avatar');
 const { refreshS3Url } = require('~/server/services/Files/S3/crud');
-const { filterFile } = require('~/server/services/Files/process');
 const { updateAction, getActions } = require('~/models/Action');
-const { getCachedTools } = require('~/server/services/Config');
 const { updateAgentProjects } = require('~/models/Agent');
 const { getProjectByName } = require('~/models/Project');
-const { revertAgentVersion } = require('~/models/Agent');
 const { deleteFileByFilter } = require('~/models/File');
+const { revertAgentVersion } = require('~/models/Agent');
+const { logger } = require('~/config');
 
 const systemTools = {
   [Tools.execute_code]: true,
@@ -47,9 +46,8 @@ const createAgentHandler = async (req, res) => {
 
     agentData.tools = [];
 
-    const availableTools = await getCachedTools({ includeGlobal: true });
     for (const tool of tools) {
-      if (availableTools[tool]) {
+      if (req.app.locals.availableTools[tool]) {
         agentData.tools.push(tool);
       }
 
@@ -170,18 +168,12 @@ const updateAgentHandler = async (req, res) => {
       });
     }
 
-    /** @type {boolean} */
-    const isProjectUpdate = (projectIds?.length ?? 0) > 0 || (removeProjectIds?.length ?? 0) > 0;
-
     let updatedAgent =
       Object.keys(updateData).length > 0
-        ? await updateAgent({ id }, updateData, {
-            updatingUserId: req.user.id,
-            skipVersioning: isProjectUpdate,
-          })
+        ? await updateAgent({ id }, updateData, { updatingUserId: req.user.id })
         : existingAgent;
 
-    if (isProjectUpdate) {
+    if (projectIds || removeProjectIds) {
       updatedAgent = await updateAgentProjects({
         user: req.user,
         agentId: id,
@@ -381,27 +373,12 @@ const uploadAgentAvatarHandler = async (req, res) => {
     }
 
     const buffer = await fs.readFile(req.file.path);
-
-    const fileStrategy = req.app.locals.fileStrategy;
-
-    const resizedBuffer = await resizeAvatar({
-      userId: req.user.id,
-      input: buffer,
+    const image = await uploadImageBuffer({
+      req,
+      context: FileContext.avatar,
+      metadata: { buffer },
     });
 
-    const { processAvatar } = getStrategyFunctions(fileStrategy);
-    const avatarUrl = await processAvatar({
-      buffer: resizedBuffer,
-      userId: req.user.id,
-      manual: 'false',
-      agentId: agent_id,
-    });
-
-    const image = {
-      filepath: avatarUrl,
-      source: fileStrategy,
-    };
-
     let _avatar;
     try {
       const agent = await getAgent({ id: agent_id });
@@ -426,7 +403,7 @@ const uploadAgentAvatarHandler = async (req, res) => {
     const data = {
       avatar: {
         filepath: image.filepath,
-        source: image.source,
+        source: req.app.locals.fileStrategy,
       },
     };
 
@@ -446,7 +423,7 @@ const uploadAgentAvatarHandler = async (req, res) => {
     try {
       await fs.unlink(req.file.path);
       logger.debug('[/:agent_id/avatar] Temp. image upload file deleted');
-    } catch {
+    } catch (error) {
       logger.debug('[/:agent_id/avatar] Temp. image upload file already deleted');
     }
   }

api/server/controllers/assistants/chatV1.js

@@ -1,7 +1,4 @@
 const { v4 } = require('uuid');
-const { sleep } = require('@librechat/agents');
-const { sendEvent } = require('@librechat/api');
-const { logger } = require('@librechat/data-schemas');
 const {
   Time,
   Constants,
@@ -22,20 +19,20 @@ 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 /
@@ -474,7 +471,7 @@ const chatV1 = async (req, res) => {
     await Promise.all(promises);
 
     const sendInitialResponse = () => {
-      sendEvent(res, {
+      sendMessage(res, {
         sync: true,
         conversationId,
         // messages: previousMessages,
@@ -590,7 +587,7 @@ const chatV1 = async (req, res) => {
       iconURL: endpointOption.iconURL,
     };
 
-    sendEvent(res, {
+    sendMessage(res, {
       final: true,
       conversation,
       requestMessage: {

api/server/controllers/assistants/chatV2.js

@@ -1,7 +1,4 @@
 const { v4 } = require('uuid');
-const { sleep } = require('@librechat/agents');
-const { sendEvent } = require('@librechat/api');
-const { logger } = require('@librechat/data-schemas');
 const {
   Time,
   Constants,
@@ -25,14 +22,15 @@ 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 /
@@ -311,7 +309,7 @@ const chatV2 = async (req, res) => {
     await Promise.all(promises);
 
     const sendInitialResponse = () => {
-      sendEvent(res, {
+      sendMessage(res, {
         sync: true,
         conversationId,
         // messages: previousMessages,
@@ -434,7 +432,7 @@ const chatV2 = async (req, res) => {
       iconURL: endpointOption.iconURL,
     };
 
-    sendEvent(res, {
+    sendMessage(res, {
       final: true,
       conversation,
       requestMessage: {

api/server/controllers/assistants/errors.js

@@ -1,10 +1,10 @@
 // errorHandler.js
-const { logger } = require('@librechat/data-schemas');
-const { CacheKeys, ViolationTypes, ContentTypes } = require('librechat-data-provider');
-const { recordUsage, checkMessageGaps } = require('~/server/services/Threads');
-const { sendResponse } = require('~/server/middleware/error');
-const { getConvo } = require('~/models/Conversation');
+const { sendResponse } = require('~/server/utils');
+const { logger } = require('~/config');
 const getLogStores = require('~/cache/getLogStores');
+const { CacheKeys, ViolationTypes, ContentTypes } = require('librechat-data-provider');
+const { getConvo } = require('~/models/Conversation');
+const { recordUsage, checkMessageGaps } = require('~/server/services/Threads');
 
 /**
  * @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/assistants/v1.js

@@ -1,5 +1,4 @@
 const fs = require('fs').promises;
-const { logger } = require('@librechat/data-schemas');
 const { FileContext } = require('librechat-data-provider');
 const { uploadImageBuffer, filterFile } = require('~/server/services/Files/process');
 const validateAuthor = require('~/server/middleware/assistants/validateAuthor');
@@ -7,9 +6,9 @@ const { getStrategyFunctions } = require('~/server/services/Files/strategies');
 const { deleteAssistantActions } = require('~/server/services/ActionService');
 const { updateAssistantDoc, getAssistants } = require('~/models/Assistant');
 const { getOpenAIClient, fetchAssistants } = require('./helpers');
-const { getCachedTools } = require('~/server/services/Config');
 const { manifestToolMap } = require('~/app/clients/tools');
 const { deleteFileByFilter } = require('~/models/File');
+const { logger } = require('~/config');
 
 /**
  * Create an assistant.
@@ -31,20 +30,21 @@ const createAssistant = async (req, res) => {
     delete assistantData.conversation_starters;
     delete assistantData.append_current_datetime;
 
-    const toolDefinitions = await getCachedTools({ includeGlobal: true });
-
     assistantData.tools = tools
       .map((tool) => {
         if (typeof tool !== 'string') {
           return tool;
         }
 
+        const toolDefinitions = req.app.locals.availableTools;
         const toolDef = toolDefinitions[tool];
         if (!toolDef && manifestToolMap[tool] && manifestToolMap[tool].toolkit === true) {
-          return Object.entries(toolDefinitions)
-            .filter(([key]) => key.startsWith(`${tool}_`))
-
-            .map(([_, val]) => val);
+          return (
+            Object.entries(toolDefinitions)
+              .filter(([key]) => key.startsWith(`${tool}_`))
+              // eslint-disable-next-line no-unused-vars
+              .map(([_, val]) => val)
+          );
         }
 
         return toolDef;
@@ -135,21 +135,21 @@ const patchAssistant = async (req, res) => {
       append_current_datetime,
       ...updateData
     } = req.body;
-
-    const toolDefinitions = await getCachedTools({ includeGlobal: true });
-
     updateData.tools = (updateData.tools ?? [])
       .map((tool) => {
         if (typeof tool !== 'string') {
           return tool;
         }
 
+        const toolDefinitions = req.app.locals.availableTools;
         const toolDef = toolDefinitions[tool];
         if (!toolDef && manifestToolMap[tool] && manifestToolMap[tool].toolkit === true) {
-          return Object.entries(toolDefinitions)
-            .filter(([key]) => key.startsWith(`${tool}_`))
-
-            .map(([_, val]) => val);
+          return (
+            Object.entries(toolDefinitions)
+              .filter(([key]) => key.startsWith(`${tool}_`))
+              // eslint-disable-next-line no-unused-vars
+              .map(([_, val]) => val)
+          );
         }
 
         return toolDef;

api/server/controllers/assistants/v2.js

@@ -1,11 +1,10 @@
-const { logger } = require('@librechat/data-schemas');
 const { ToolCallTypes } = require('librechat-data-provider');
 const validateAuthor = require('~/server/middleware/assistants/validateAuthor');
 const { validateAndUpdateTool } = require('~/server/services/ActionService');
-const { getCachedTools } = require('~/server/services/Config');
 const { updateAssistantDoc } = require('~/models/Assistant');
 const { manifestToolMap } = require('~/app/clients/tools');
 const { getOpenAIClient } = require('./helpers');
+const { logger } = require('~/config');
 
 /**
  * Create an assistant.
@@ -28,20 +27,21 @@ const createAssistant = async (req, res) => {
     delete assistantData.conversation_starters;
     delete assistantData.append_current_datetime;
 
-    const toolDefinitions = await getCachedTools({ includeGlobal: true });
-
     assistantData.tools = tools
       .map((tool) => {
         if (typeof tool !== 'string') {
           return tool;
         }
 
+        const toolDefinitions = req.app.locals.availableTools;
         const toolDef = toolDefinitions[tool];
         if (!toolDef && manifestToolMap[tool] && manifestToolMap[tool].toolkit === true) {
-          return Object.entries(toolDefinitions)
-            .filter(([key]) => key.startsWith(`${tool}_`))
-
-            .map(([_, val]) => val);
+          return (
+            Object.entries(toolDefinitions)
+              .filter(([key]) => key.startsWith(`${tool}_`))
+              // eslint-disable-next-line no-unused-vars
+              .map(([_, val]) => val)
+          );
         }
 
         return toolDef;
@@ -125,13 +125,13 @@ const updateAssistant = async ({ req, openai, assistant_id, updateData }) => {
 
   let hasFileSearch = false;
   for (const tool of updateData.tools ?? []) {
-    const toolDefinitions = await getCachedTools({ includeGlobal: true });
+    const toolDefinitions = req.app.locals.availableTools;
     let actualTool = typeof tool === 'string' ? toolDefinitions[tool] : tool;
 
     if (!actualTool && manifestToolMap[tool] && manifestToolMap[tool].toolkit === true) {
       actualTool = Object.entries(toolDefinitions)
         .filter(([key]) => key.startsWith(`${tool}_`))
-
+        // eslint-disable-next-line no-unused-vars
         .map(([_, val]) => val);
     } else if (!actualTool) {
       continue;

api/server/index.js

@@ -1,22 +1,22 @@
 require('dotenv').config();
-const fs = require('fs');
 const path = require('path');
 require('module-alias')({ base: path.resolve(__dirname, '..') });
 const cors = require('cors');
 const axios = require('axios');
 const express = require('express');
-const passport = require('passport');
 const compression = require('compression');
-const cookieParser = require('cookie-parser');
-const { isEnabled } = require('@librechat/api');
-const { logger } = require('@librechat/data-schemas');
+const passport = require('passport');
 const mongoSanitize = require('express-mongo-sanitize');
+const fs = require('fs');
+const cookieParser = require('cookie-parser');
 const { connectDb, indexSync } = require('~/db');
 
+const { jwtLogin, passportLogin } = require('~/strategies');
+const { isEnabled } = require('~/server/utils');
+const { ldapLogin } = require('~/strategies');
+const { logger } = require('~/config');
 const validateImageRequest = require('./middleware/validateImageRequest');
-const { jwtLogin, ldapLogin, passportLogin } = require('~/strategies');
 const errorController = require('./controllers/ErrorController');
-const initializeMCP = require('./services/initializeMCP');
 const configureSocialLogins = require('./socialLogins');
 const AppService = require('./services/AppService');
 const staticCache = require('./utils/staticCache');
@@ -39,9 +39,7 @@ const startServer = async () => {
   await connectDb();
 
   logger.info('Connected to MongoDB');
-  indexSync().catch((err) => {
-    logger.error('[indexSync] Background sync failed:', err);
-  });
+  await indexSync();
 
   app.disable('x-powered-by');
   app.set('trust proxy', trusted_proxy);
@@ -97,6 +95,7 @@ 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);
@@ -117,9 +116,9 @@ const startServer = async () => {
   app.use('/api/roles', routes.roles);
   app.use('/api/agents', routes.agents);
   app.use('/api/banner', routes.banner);
-  app.use('/api/memories', routes.memories);
+  app.use('/api/bedrock', routes.bedrock);
+
   app.use('/api/tags', routes.tags);
-  app.use('/api/mcp', routes.mcp);
 
   app.use((req, res) => {
     res.set({
@@ -143,8 +142,6 @@ const startServer = async () => {
     } else {
       logger.info(`Server listening at http://${host == '0.0.0.0' ? 'localhost' : host}:${port}`);
     }
-
-    initializeMCP(app);
   });
 };
 
@@ -187,5 +184,5 @@ process.on('uncaughtException', (err) => {
   process.exit(1);
 });
 
-/** Export app for easier testing purposes */
+// export app for easier testing purposes
 module.exports = app;

api/server/middleware/abortMiddleware.js

@@ -1,13 +1,13 @@
-const { logger } = require('@librechat/data-schemas');
-const { countTokens, isEnabled, sendEvent } = require('@librechat/api');
+// abortMiddleware.js
 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 sendEvent(res, finalEvent);
+    return sendMessage(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) => {
-    sendEvent(res, { message: userMessage, created: true });
+    sendMessage(res, { message: userMessage, created: true });
 
     const abortKey = userMessage?.conversationId ?? req.user.id;
     getReqData({ abortKey });
@@ -327,7 +327,7 @@ const handleAbortError = async (res, req, error, data) => {
     errorText = `{"type":"${ErrorTypes.INVALID_REQUEST}"}`;
   }
 
-  if (error?.message?.includes("does not support 'system'")) {
+  if (error?.message?.includes('does not support \'system\'')) {
     errorText = `{"type":"${ErrorTypes.NO_SYSTEM_MESSAGES}"}`;
   }
 

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 sendEvent(res, finalEvent);
+    return sendMessage(res, finalEvent);
   }
 
   res.json(finalEvent);

api/server/middleware/buildEndpointOption.js

@@ -1,12 +1,13 @@
-const { logger } = require('@librechat/data-schemas');
 const {
-  EndpointURLs,
+  parseCompactConvo,
   EModelEndpoint,
   isAgentsEndpoint,
-  parseCompactConvo,
+  EndpointURLs,
 } = 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');
@@ -24,6 +25,7 @@ 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,
 };
@@ -34,9 +36,6 @@ 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' });
   }
 
@@ -58,6 +57,15 @@ 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 !== '') {
@@ -69,7 +77,6 @@ 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' });
     }
   }
@@ -77,23 +84,20 @@ async function buildEndpointOption(req, res, next) {
   try {
     const isAgents =
       isAgentsEndpoint(endpoint) || req.baseUrl.startsWith(EndpointURLs[EModelEndpoint.agents]);
-    const builder = isAgents
-      ? (...args) => buildFunction[EModelEndpoint.agents](req, ...args)
-      : buildFunction[endpointType ?? endpoint];
+    const endpointFn = buildFunction[isAgents ? EModelEndpoint.agents : (endpointType ?? endpoint)];
+    const builder = isAgents ? (...args) => endpointFn(req, ...args) : endpointFn;
 
     // 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

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

api/server/middleware/roles/index.js

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

api/server/middleware/setBalanceConfig.js

@@ -1,6 +1,8 @@
+const mongoose = require('mongoose');
 const { logger } = require('@librechat/data-schemas');
 const { getBalanceConfig } = require('~/server/services/Config');
-const { Balance } = require('~/db/models');
+
+const Balance = require('~/db/models').Balance;
 
 /**
  * Middleware to synchronize user balance settings with current balance configuration.

api/server/middleware/validate/convoAccess.js

@@ -1,8 +1,8 @@
-const { isEnabled } = require('@librechat/api');
 const { Constants, ViolationTypes, Time } = require('librechat-data-provider');
 const { searchConversation } = require('~/models/Conversation');
 const denyRequest = require('~/server/middleware/denyRequest');
 const { logViolation, getLogStores } = require('~/cache');
+const { isEnabled } = require('~/server/utils');
 
 const { USE_REDIS, CONVO_ACCESS_VIOLATION_SCORE: score = 0 } = process.env ?? {};
 

api/server/routes/actions.js

@@ -1,10 +1,8 @@
 const express = require('express');
 const jwt = require('jsonwebtoken');
-const { getAccessToken } = require('@librechat/api');
-const { logger } = require('@librechat/data-schemas');
 const { CacheKeys } = require('librechat-data-provider');
-const { findToken, updateToken, createToken } = require('~/models');
-const { getFlowStateManager } = require('~/config');
+const { getAccessToken } = require('~/server/services/TokenService');
+const { logger, getFlowStateManager } = require('~/config');
 const { getLogStores } = require('~/cache');
 
 const router = express.Router();
@@ -30,19 +28,18 @@ router.get('/:action_id/oauth/callback', async (req, res) => {
     try {
       decodedState = jwt.verify(state, JWT_SECRET);
     } catch (err) {
-      logger.error('Error verifying state parameter:', err);
       await flowManager.failFlow(identifier, 'oauth', 'Invalid or expired state parameter');
-      return res.redirect('/oauth/error?error=invalid_state');
+      return res.status(400).send('Invalid or expired state parameter');
     }
 
     if (decodedState.action_id !== action_id) {
       await flowManager.failFlow(identifier, 'oauth', 'Mismatched action ID in state parameter');
-      return res.redirect('/oauth/error?error=invalid_state');
+      return res.status(400).send('Mismatched action ID in state parameter');
     }
 
     if (!decodedState.user) {
       await flowManager.failFlow(identifier, 'oauth', 'Invalid user ID in state parameter');
-      return res.redirect('/oauth/error?error=invalid_state');
+      return res.status(400).send('Invalid user ID in state parameter');
     }
     identifier = `${decodedState.user}:${action_id}`;
     const flowState = await flowManager.getFlowState(identifier, 'oauth');
@@ -50,34 +47,90 @@ router.get('/:action_id/oauth/callback', async (req, res) => {
       throw new Error('OAuth flow not found');
     }
 
-    const tokenData = await getAccessToken(
-      {
-        code,
-        userId: decodedState.user,
-        identifier,
-        client_url: flowState.metadata.client_url,
-        redirect_uri: flowState.metadata.redirect_uri,
-        token_exchange_method: flowState.metadata.token_exchange_method,
-        /** Encrypted values */
-        encrypted_oauth_client_id: flowState.metadata.encrypted_oauth_client_id,
-        encrypted_oauth_client_secret: flowState.metadata.encrypted_oauth_client_secret,
-      },
-      {
-        findToken,
-        updateToken,
-        createToken,
-      },
-    );
+    const tokenData = await getAccessToken({
+      code,
+      userId: decodedState.user,
+      identifier,
+      client_url: flowState.metadata.client_url,
+      redirect_uri: flowState.metadata.redirect_uri,
+      /** Encrypted values */
+      encrypted_oauth_client_id: flowState.metadata.encrypted_oauth_client_id,
+      encrypted_oauth_client_secret: flowState.metadata.encrypted_oauth_client_secret,
+    });
     await flowManager.completeFlow(identifier, 'oauth', tokenData);
-
-    /** Redirect to React success page */
-    const serverName = flowState.metadata?.action_name || `Action ${action_id}`;
-    const redirectUrl = `/oauth/success?serverName=${encodeURIComponent(serverName)}`;
-    res.redirect(redirectUrl);
+    res.send(`
+      <!DOCTYPE html>
+      <html>
+        <head>
+          <title>Authentication Successful</title>
+          <meta name="viewport" content="width=device-width, initial-scale=1.0">
+          <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
+          <style>
+            body {
+              font-family: ui-sans-serif, system-ui, -apple-system, BlinkMacSystemFont;
+              background-color: rgb(249, 250, 251);
+              margin: 0;
+              padding: 2rem;
+              display: flex;
+              justify-content: center;
+              align-items: center;
+              min-height: 100vh;
+            }
+            .card {
+              background-color: white;
+              border-radius: 0.5rem;
+              padding: 2rem;
+              max-width: 28rem;
+              width: 100%;
+              box-shadow: 0 4px 6px -1px rgb(0 0 0 / 0.1), 0 2px 4px -2px rgb(0 0 0 / 0.1);
+              text-align: center;
+            }
+            .heading {
+              color: rgb(17, 24, 39);
+              font-size: 1.875rem;
+              font-weight: 700;
+              margin: 0 0 1rem;
+            }
+            .description {
+              color: rgb(75, 85, 99);
+              font-size: 0.875rem;
+              margin: 0.5rem 0;
+            }
+            .countdown {
+              color: rgb(99, 102, 241);
+              font-weight: 500;
+            }
+          </style>
+        </head>
+        <body>
+          <div class="card">
+            <h1 class="heading">Authentication Successful</h1>
+            <p class="description">
+              Your authentication was successful. This window will close in 
+              <span class="countdown" id="countdown">3</span> seconds.
+            </p>
+          </div>
+          <script>
+            let secondsLeft = 3;
+            const countdownElement = document.getElementById('countdown');
+            
+            const countdown = setInterval(() => {
+              secondsLeft--;
+              countdownElement.textContent = secondsLeft;
+              
+              if (secondsLeft <= 0) {
+                clearInterval(countdown);
+                window.close();
+              }
+            }, 1000);
+          </script>
+        </body>
+      </html>
+    `);
   } catch (error) {
     logger.error('Error in OAuth callback:', error);
     await flowManager.failFlow(identifier, 'oauth', error);
-    res.redirect('/oauth/error?error=callback_failed');
+    res.status(500).send('Authentication failed. Please try again.');
   }
 });
 

api/server/routes/agents/tools.js

@@ -1,5 +1,4 @@
 const express = require('express');
-const { addTool } = require('@librechat/api');
 const { callTool, verifyToolAuth, getToolCalls } = require('~/server/controllers/tools');
 const { getAvailableTools } = require('~/server/controllers/PluginController');
 const { toolCallLimiter } = require('~/server/middleware/limiters');
@@ -37,12 +36,4 @@ router.get('/:toolId/auth', verifyToolAuth);
  */
 router.post('/:toolId/call', toolCallLimiter, callTool);
 
-/**
- * Add a new tool to the system
- * @route POST /agents/tools/add
- * @param {object} req.body - Request body containing tool data
- * @returns {object} Created tool object
- */
-router.post('/add', addTool);
-
 module.exports = router;

api/server/routes/ask/addToCache.js

@@ -0,0 +1,63 @@
+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

@@ -0,0 +1,25 @@
+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;