feat: Add configurable thinking indicator text

- Add thinkingIndicatorText field to interface configuration schema
- Update ContentParts component to use configurable text instead of hardcoded translation
- Update example config with new option
- Maintain backward compatibility by falling back to default translation

.env.example

@@ -15,17 +15,38 @@ HOST=localhost
 PORT=3080
 
 MONGO_URI=mongodb://127.0.0.1:27017/LibreChat
+#The maximum number of connections in the connection pool. */
+MONGO_MAX_POOL_SIZE=
+#The minimum number of connections in the connection pool. */
+MONGO_MIN_POOL_SIZE=
+#The maximum number of connections that may be in the process of being established concurrently by the connection pool. */
+MONGO_MAX_CONNECTING=
+#The maximum number of milliseconds that a connection can remain idle in the pool before being removed and closed. */
+MONGO_MAX_IDLE_TIME_MS=
+#The maximum time in milliseconds that a thread can wait for a connection to become available. */
+MONGO_WAIT_QUEUE_TIMEOUT_MS=
+# Set to false to disable automatic index creation for all models associated with this connection. */
+MONGO_AUTO_INDEX=
+# Set to `false` to disable Mongoose automatically calling `createCollection()` on every model created on this connection. */
+MONGO_AUTO_CREATE=
 
 DOMAIN_CLIENT=http://localhost:3080
 DOMAIN_SERVER=http://localhost:3080
 
 NO_INDEX=true
-# Use the address that is at most n number of hops away from the Express application. 
-# req.socket.remoteAddress is the first hop, and the rest are looked for in the X-Forwarded-For header from right to left. 
+# Use the address that is at most n number of hops away from the Express application.
+# req.socket.remoteAddress is the first hop, and the rest are looked for in the X-Forwarded-For header from right to left.
 # A value of 0 means that the first untrusted address would be req.socket.remoteAddress, i.e. there is no reverse proxy.
 # Defaulted to 1.
 TRUST_PROXY=1
 
+# Minimum password length for user authentication
+# Default: 8
+# Note: When using LDAP authentication, you may want to set this to 1 
+# to bypass local password validation, as LDAP servers handle their own
+# password policies.
+# MIN_PASSWORD_LENGTH=8
+
 #===============#
 # JSON Logging  #
 #===============#
@@ -58,7 +79,7 @@ DEBUG_CONSOLE=false
 #                     Endpoints                     #
 #===================================================#
 
-# ENDPOINTS=openAI,assistants,azureOpenAI,google,gptPlugins,anthropic
+# ENDPOINTS=openAI,assistants,azureOpenAI,google,anthropic
 
 PROXY=
 
@@ -88,7 +109,7 @@ PROXY=
 #============#
 
 ANTHROPIC_API_KEY=user_provided
-# ANTHROPIC_MODELS=claude-3-7-sonnet-latest,claude-3-7-sonnet-20250219,claude-3-5-haiku-20241022,claude-3-5-sonnet-20241022,claude-3-5-sonnet-latest,claude-3-5-sonnet-20240620,claude-3-opus-20240229,claude-3-sonnet-20240229,claude-3-haiku-20240307,claude-2.1,claude-2,claude-1.2,claude-1,claude-1-100k,claude-instant-1,claude-instant-1-100k
+# ANTHROPIC_MODELS=claude-opus-4-20250514,claude-sonnet-4-20250514,claude-3-7-sonnet-20250219,claude-3-5-sonnet-20241022,claude-3-5-haiku-20241022,claude-3-opus-20240229,claude-3-sonnet-20240229,claude-3-haiku-20240307
 # ANTHROPIC_REVERSE_PROXY=
 
 #============#
@@ -142,12 +163,12 @@ GOOGLE_KEY=user_provided
 # GOOGLE_AUTH_HEADER=true
 
 # Gemini API (AI Studio)
-# GOOGLE_MODELS=gemini-2.5-pro-exp-03-25,gemini-2.0-flash-exp,gemini-2.0-flash-thinking-exp-1219,gemini-exp-1121,gemini-exp-1114,gemini-1.5-flash-latest,gemini-1.0-pro,gemini-1.0-pro-001,gemini-1.0-pro-latest,gemini-1.0-pro-vision-latest,gemini-1.5-pro-latest,gemini-pro,gemini-pro-vision
+# GOOGLE_MODELS=gemini-2.5-pro,gemini-2.5-flash,gemini-2.5-flash-lite,gemini-2.0-flash,gemini-2.0-flash-lite
 
 # Vertex AI
-# GOOGLE_MODELS=gemini-1.5-flash-preview-0514,gemini-1.5-pro-preview-0514,gemini-1.0-pro-vision-001,gemini-1.0-pro-002,gemini-1.0-pro-001,gemini-pro-vision,gemini-1.0-pro
+# GOOGLE_MODELS=gemini-2.5-pro,gemini-2.5-flash,gemini-2.5-flash-lite,gemini-2.0-flash-001,gemini-2.0-flash-lite-001
 
-# GOOGLE_TITLE_MODEL=gemini-pro
+# GOOGLE_TITLE_MODEL=gemini-2.0-flash-lite-001
 
 # GOOGLE_LOC=us-central1
 
@@ -349,6 +370,11 @@ REGISTRATION_VIOLATION_SCORE=1
 CONCURRENT_VIOLATION_SCORE=1
 MESSAGE_VIOLATION_SCORE=1
 NON_BROWSER_VIOLATION_SCORE=20
+TTS_VIOLATION_SCORE=0
+STT_VIOLATION_SCORE=0
+FORK_VIOLATION_SCORE=0
+IMPORT_VIOLATION_SCORE=0
+FILE_UPLOAD_VIOLATION_SCORE=0
 
 LOGIN_MAX=7
 LOGIN_WINDOW=5
@@ -437,12 +463,85 @@ OPENID_REQUIRED_ROLE_PARAMETER_PATH=
 OPENID_USERNAME_CLAIM=
 # Set to determine which user info property returned from OpenID Provider to store as the User's name
 OPENID_NAME_CLAIM=
+# Optional audience parameter for OpenID authorization requests
+OPENID_AUDIENCE=
 
 OPENID_BUTTON_LABEL=
 OPENID_IMAGE_URL=
 # Set to true to automatically redirect to the OpenID provider when a user visits the login page
 # This will bypass the login form completely for users, only use this if OpenID is your only authentication method
 OPENID_AUTO_REDIRECT=false
+# Set to true to use PKCE (Proof Key for Code Exchange) for OpenID authentication
+OPENID_USE_PKCE=false
+#Set to true to reuse openid tokens for authentication management instead of using the mongodb session and the custom refresh token.
+OPENID_REUSE_TOKENS=
+#By default, signing key verification results are cached in order to prevent excessive HTTP requests to the JWKS endpoint.
+#If a signing key matching the kid is found, this will be cached and the next time this kid is requested the signing key will be served from the cache.
+#Default is true.
+OPENID_JWKS_URL_CACHE_ENABLED=
+OPENID_JWKS_URL_CACHE_TIME= # 600000 ms eq to 10 minutes leave empty to disable caching
+#Set to true to trigger token exchange flow to acquire access token for the userinfo endpoint.
+OPENID_ON_BEHALF_FLOW_FOR_USERINFO_REQUIRED=
+OPENID_ON_BEHALF_FLOW_USERINFO_SCOPE="user.read" # example for Scope Needed for Microsoft Graph API
+# Set to true to use the OpenID Connect end session endpoint for logout
+OPENID_USE_END_SESSION_ENDPOINT=
+
+#========================#
+# SharePoint Integration #
+#========================#
+# Requires Entra ID (OpenID) authentication to be configured
+
+# Enable SharePoint file picker in chat and agent panels
+# ENABLE_SHAREPOINT_FILEPICKER=true
+
+# SharePoint tenant base URL (e.g., https://yourtenant.sharepoint.com)
+# SHAREPOINT_BASE_URL=https://yourtenant.sharepoint.com
+
+# Microsoft Graph API And SharePoint scopes for file picker
+# SHAREPOINT_PICKER_SHAREPOINT_SCOPE==https://yourtenant.sharepoint.com/AllSites.Read
+# SHAREPOINT_PICKER_GRAPH_SCOPE=Files.Read.All
+#========================#
+
+# SAML
+# Note: If OpenID is enabled, SAML authentication will be automatically disabled.
+SAML_ENTRY_POINT=
+SAML_ISSUER=
+SAML_CERT=
+SAML_CALLBACK_URL=/oauth/saml/callback
+SAML_SESSION_SECRET=
+
+# Attribute mappings (optional)
+SAML_EMAIL_CLAIM=
+SAML_USERNAME_CLAIM=
+SAML_GIVEN_NAME_CLAIM=
+SAML_FAMILY_NAME_CLAIM=
+SAML_PICTURE_CLAIM=
+SAML_NAME_CLAIM=
+
+# Logint buttion settings (optional)
+SAML_BUTTON_LABEL=
+SAML_IMAGE_URL=
+
+# Whether the SAML Response should be signed.
+# - If "true", the entire `SAML Response` will be signed.
+# - If "false" or unset, only the `SAML Assertion` will be signed (default behavior).
+# SAML_USE_AUTHN_RESPONSE_SIGNED=
+
+
+#===============================================#
+# Microsoft Graph API / Entra ID Integration  #
+#===============================================#
+
+# Enable Entra ID people search integration in permissions/sharing system
+# When enabled, the people picker will search both local database and Entra ID
+USE_ENTRA_ID_FOR_PEOPLE_SEARCH=false
+
+# When enabled, entra id groups owners will be considered as members of the group
+ENTRA_ID_INCLUDE_OWNERS_AS_MEMBERS=false
+
+# Microsoft Graph API scopes needed for people/group search
+# Default scopes provide access to user profiles and group memberships
+OPENID_GRAPH_SCOPES=User.Read,People.Read,GroupMember.Read.All
 
 # LDAP
 LDAP_URL=
@@ -474,6 +573,18 @@ 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           #
 #========================#
@@ -522,6 +633,10 @@ ALLOW_SHARED_LINKS_PUBLIC=true
 # If you have another service in front of your LibreChat doing compression, disable express based compression here
 # DISABLE_COMPRESSION=true
 
+# If you have gzipped version of uploaded image images in the same folder, this will enable gzip scan and serving of these images
+# Note: The images folder will be scanned on startup and a ma kept in memory. Be careful for large number of images.
+# ENABLE_IMAGE_OUTPUT_GZIP_SCAN=true
+
 #===================================================#
 #                        UI                         #
 #===================================================#
@@ -539,11 +654,44 @@ HELP_AND_FAQ_URL=https://librechat.ai
 # REDIS Options #
 #===============#
 
-# REDIS_URI=10.10.10.10:6379
+# Enable Redis for caching and session storage
 # USE_REDIS=true
 
-# USE_REDIS_CLUSTER=true
-# REDIS_CA=/path/to/ca.crt
+# Single Redis instance
+# REDIS_URI=redis://127.0.0.1:6379
+
+# Redis cluster (multiple nodes)
+# REDIS_URI=redis://127.0.0.1:7001,redis://127.0.0.1:7002,redis://127.0.0.1:7003
+
+# Redis with TLS/SSL encryption and CA certificate
+# REDIS_URI=rediss://127.0.0.1:6380
+# REDIS_CA=/path/to/ca-cert.pem
+
+# Elasticache may need to use an alternate dnsLookup for TLS connections.  see "Special Note: Aws Elasticache Clusters with TLS" on this webpage: https://www.npmjs.com/package/ioredis
+# Enable alternative dnsLookup for redis
+# REDIS_USE_ALTERNATIVE_DNS_LOOKUP=true
+
+# Redis authentication (if required)
+# REDIS_USERNAME=your_redis_username
+# REDIS_PASSWORD=your_redis_password
+
+# Redis key prefix configuration
+# Use environment variable name for dynamic prefix (recommended for cloud deployments)
+# REDIS_KEY_PREFIX_VAR=K_REVISION
+# Or use static prefix directly
+# REDIS_KEY_PREFIX=librechat
+
+# Redis connection limits
+# REDIS_MAX_LISTENERS=40
+
+# Redis ping interval in seconds (0 = disabled, >0 = enabled)
+# When set to a positive integer, Redis clients will ping the server at this interval to keep connections alive
+# When unset or 0, no pinging is performed (recommended for most use cases)
+# REDIS_PING_INTERVAL=300
+
+# Force specific cache namespaces to use in-memory storage even when Redis is enabled
+# Comma-separated list of CacheKeys (e.g., ROLES,MESSAGES)
+# FORCED_IN_MEMORY_CACHE_NAMESPACES=ROLES,MESSAGES
 
 #==================================================#
 #                      Others                      #
@@ -563,9 +711,9 @@ HELP_AND_FAQ_URL=https://librechat.ai
 #   users always get the latest version. Customize    #
 #   only if you understand caching implications.      #
 
-# INDEX_HTML_CACHE_CONTROL=no-cache, no-store, must-revalidate
-# INDEX_HTML_PRAGMA=no-cache
-# INDEX_HTML_EXPIRES=0
+# INDEX_CACHE_CONTROL=no-cache, no-store, must-revalidate
+# INDEX_PRAGMA=no-cache
+# INDEX_EXPIRES=0
 
 # no-cache: Forces validation with server before using cached version
 # no-store: Prevents storing the response entirely
@@ -575,3 +723,46 @@ HELP_AND_FAQ_URL=https://librechat.ai
 #                  OpenWeather                        #
 #=====================================================#
 OPENWEATHER_API_KEY=
+
+#====================================#
+# LibreChat Code Interpreter API     #
+#====================================#
+
+# https://code.librechat.ai
+# LIBRECHAT_CODE_API_KEY=your-key
+
+#======================#
+# Web Search           #
+#======================#
+
+# Note: All of the following variable names can be customized.
+# Omit values to allow user to provide them.
+
+# For more information on configuration values, see:
+# https://librechat.ai/docs/features/web_search
+
+# Search Provider (Required)
+# SERPER_API_KEY=your_serper_api_key
+
+# Scraper (Required)
+# FIRECRAWL_API_KEY=your_firecrawl_api_key
+# Optional: Custom Firecrawl API URL
+# FIRECRAWL_API_URL=your_firecrawl_api_url
+
+# Reranker (Required)
+# JINA_API_KEY=your_jina_api_key
+# or
+# COHERE_API_KEY=your_cohere_api_key
+
+#======================#
+# MCP Configuration    #
+#======================#
+
+# Treat 401/403 responses as OAuth requirement when no oauth metadata found
+# MCP_OAUTH_ON_AUTH_ERROR=true
+
+# Timeout for OAuth detection requests in milliseconds
+# MCP_OAUTH_DETECTION_TIMEOUT=5000
+
+# Cache connection status checks for this many milliseconds to avoid expensive verification
+# MCP_CONNECTION_CHECK_TTL=60000

.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 MCP: `npm run build:mcp`.
-6. Build data schemas: `npm run build:data-schemas`.
+5. Build data schemas: `npm run build:data-schemas`.
+6. Build API methods: `npm run build:api`.
 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`.
@@ -147,7 +147,7 @@ Apply the following naming conventions to branches, labels, and other Git-relate
 ## 8. Module Import Conventions
 
 - `npm` packages first, 
-     - from shortest line (top) to longest (bottom)
+     - from longest line (top) to shortest (bottom)
 
 - Followed by typescript types (pertains to data-provider and client workspaces)
      - longest line (top) to shortest (bottom)
@@ -157,6 +157,8 @@ Apply the following naming conventions to branches, labels, and other Git-relate
      - longest line (top) to shortest (bottom)
      - imports with alias `~` treated the same as relative import with respect to line length
 
+**Note:** ESLint will automatically enforce these import conventions when you run `npm run lint --fix` or through pre-commit hooks.
+
 ---
 
 Please ensure that you adapt this summary to fit the specific context and nuances of your project.

.github/workflows/backend-review.yml

@@ -7,6 +7,7 @@ on:
       - release/*
     paths:
       - 'api/**'
+      - 'packages/**'
 jobs:
   tests_Backend:
     name: Run Backend unit tests
@@ -36,12 +37,12 @@ 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
@@ -66,5 +67,8 @@ jobs:
       - name: Run librechat-data-provider unit tests
         run: cd packages/data-provider && npm run test:ci
 
-      - name: Run librechat-mcp unit tests
-        run: cd packages/mcp && 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

.github/workflows/client.yml

@@ -0,0 +1,58 @@
+name: Publish `@librechat/client` to NPM
+
+on:
+  push:
+    branches:
+      - main
+    paths:
+      - 'packages/client/package.json'
+  workflow_dispatch:
+    inputs:
+      reason:
+        description: 'Reason for manual trigger'
+        required: false
+        default: 'Manual publish requested'
+
+jobs:
+  build-and-publish:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      
+      - name: Use Node.js
+        uses: actions/setup-node@v4
+        with:
+          node-version: '20.x'
+        
+      - name: Install client dependencies
+        run: cd packages/client && npm ci
+        
+      - name: Build client
+        run: cd packages/client && npm run build
+        
+      - name: Set up npm authentication
+        run: echo "//registry.npmjs.org/:_authToken=${{ secrets.PUBLISH_NPM_TOKEN }}" > ~/.npmrc
+        
+      - name: Check version change
+        id: check
+        working-directory: packages/client
+        run: |
+          PACKAGE_VERSION=$(node -p "require('./package.json').version")
+          PUBLISHED_VERSION=$(npm view @librechat/client version 2>/dev/null || echo "0.0.0")
+          if [ "$PACKAGE_VERSION" = "$PUBLISHED_VERSION" ]; then
+            echo "No version change, skipping publish"
+            echo "skip=true" >> $GITHUB_OUTPUT
+          else
+            echo "Version changed, proceeding with publish"
+            echo "skip=false" >> $GITHUB_OUTPUT
+          fi
+            
+      - name: Pack package
+        if: steps.check.outputs.skip != 'true'
+        working-directory: packages/client
+        run: npm pack
+        
+      - name: Publish
+        if: steps.check.outputs.skip != 'true'
+        working-directory: packages/client
+        run: npm publish *.tgz --access public

.github/workflows/data-provider.yml

@@ -1,4 +1,4 @@
-name: Node.js Package
+name: Publish `librechat-data-provider` to NPM
 
 on:
   push:
@@ -6,6 +6,12 @@ on:
       - main
     paths:
       - 'packages/data-provider/package.json'
+  workflow_dispatch:
+    inputs:
+      reason:
+        description: 'Reason for manual trigger'
+        required: false
+        default: 'Manual publish requested'
 
 jobs:
   build:
@@ -14,7 +20,7 @@ jobs:
       - uses: actions/checkout@v4
       - uses: actions/setup-node@v4
         with:
-          node-version: 16
+          node-version: 20
       - run: cd packages/data-provider && npm ci
       - run: cd packages/data-provider && npm run build
 
@@ -25,7 +31,7 @@ jobs:
       - uses: actions/checkout@v4
       - uses: actions/setup-node@v4
         with:
-          node-version: 16
+          node-version: 20
           registry-url: 'https://registry.npmjs.org'
       - run: cd packages/data-provider && npm ci
       - run: cd packages/data-provider && npm run build

.github/workflows/data-schemas.yml

@@ -22,7 +22,7 @@ jobs:
       - name: Use Node.js
         uses: actions/setup-node@v4
         with:
-          node-version: '18.x'
+          node-version: '20.x'
           
       - name: Install dependencies
         run: cd packages/data-schemas && npm ci

.github/workflows/deploy-dev.yml

@@ -2,7 +2,7 @@ name: Update Test Server
 
 on:
   workflow_run:
-    workflows: ["Docker Dev Images Build"]
+    workflows: ["Docker Dev Branch Images Build"]
     types:
       - completed
   workflow_dispatch:
@@ -12,7 +12,8 @@ jobs:
     runs-on: ubuntu-latest
     if: |
       github.repository == 'danny-avila/LibreChat' &&
-      (github.event_name == 'workflow_dispatch' || github.event.workflow_run.conclusion == 'success')
+      (github.event_name == 'workflow_dispatch' || 
+       (github.event.workflow_run.conclusion == 'success' && github.event.workflow_run.head_branch == 'dev'))
     steps:
     - name: Checkout repository
       uses: actions/checkout@v4
@@ -29,13 +30,17 @@ 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 && \
-        npm run update:deployed && \
+        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 && \
         git checkout do-deploy && \
-        git rebase main && \
-        npm run start:deployed && \
+        git rebase dev && \
+        sudo npm run start:deployed && \
         echo "Update completed. Application should be running now."
         EEOF
         EOF

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

@@ -0,0 +1,72 @@
+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/frontend-review.yml

@@ -8,7 +8,7 @@ on:
       - release/*
     paths:
       - 'client/**'
-      - 'packages/**'
+      - 'packages/data-provider/**'
 
 jobs:
   tests_frontend_ubuntu:

.github/workflows/generate-release-changelog-pr.yml

@@ -1,94 +0,0 @@
-name: Generate Release Changelog PR
-
-on:
-  push:
-    tags:
-      - 'v*.*.*'
-
-jobs:
-  generate-release-changelog-pr:
-    permissions:
-      contents: write    # Needed for pushing commits and creating branches.
-      pull-requests: write
-    runs-on: ubuntu-latest
-    steps:
-      # 1. Checkout the repository (with full history).
-      - name: Checkout Repository
-        uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-
-      # 2. Generate the release changelog using our custom configuration.
-      - name: Generate Release Changelog
-        id: generate_release
-        uses: mikepenz/release-changelog-builder-action@v5.1.0
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-        with:
-          configuration: ".github/configuration-release.json"
-          owner: ${{ github.repository_owner }}
-          repo: ${{ github.event.repository.name }}
-          outputFile: CHANGELOG-release.md
-
-      # 3. Update the main CHANGELOG.md:
-      #    - If it doesn't exist, create it with a basic header.
-      #    - Remove the "Unreleased" section (if present).
-      #    - Prepend the new release changelog above previous releases.
-      #    - Remove all temporary files before committing.
-      - name: Update CHANGELOG.md
-        run: |
-          # Determine the release tag, e.g. "v1.2.3"
-          TAG=${GITHUB_REF##*/}
-          echo "Using release tag: $TAG"
-          
-          # Ensure CHANGELOG.md exists; if not, create a basic header.
-          if [ ! -f CHANGELOG.md ]; then
-            echo "# Changelog" > CHANGELOG.md
-            echo "" >> CHANGELOG.md
-            echo "All notable changes to this project will be documented in this file." >> CHANGELOG.md
-            echo "" >> CHANGELOG.md
-          fi
-
-          echo "Updating CHANGELOG.md…"
-          
-          # Remove the "Unreleased" section (from "## [Unreleased]" until the first occurrence of '---') if it exists.
-          if grep -q "^## \[Unreleased\]" CHANGELOG.md; then
-            awk '/^## \[Unreleased\]/{flag=1} flag && /^---/{flag=0; next} !flag' CHANGELOG.md > CHANGELOG.cleaned
-          else
-            cp CHANGELOG.md CHANGELOG.cleaned
-          fi
-
-          # Split the cleaned file into:
-          #   - header.md: content before the first release header ("## [v...").
-          #   - tail.md: content from the first release header onward.
-          awk '/^## \[v/{exit} {print}' CHANGELOG.cleaned > header.md
-          awk 'f{print} /^## \[v/{f=1; print}' CHANGELOG.cleaned > tail.md
-
-          # Combine header, the new release changelog, and the tail.
-          echo "Combining updated changelog parts..."
-          cat header.md CHANGELOG-release.md > CHANGELOG.md.new
-          echo "" >> CHANGELOG.md.new
-          cat tail.md >> CHANGELOG.md.new
-
-          mv CHANGELOG.md.new CHANGELOG.md
-
-          # Remove temporary files.
-          rm -f CHANGELOG.cleaned header.md tail.md CHANGELOG-release.md
-
-          echo "Final CHANGELOG.md content:"
-          cat CHANGELOG.md
-
-      # 4. Create (or update) the Pull Request with the updated CHANGELOG.md.
-      - name: Create Pull Request
-        uses: peter-evans/create-pull-request@v7
-        with:
-          token: ${{ secrets.GITHUB_TOKEN }}
-          sign-commits: true
-          commit-message: "chore: update CHANGELOG for release ${{ github.ref_name }}"
-          base: main
-          branch: "changelog/${{ github.ref_name }}"
-          reviewers: danny-avila
-          title: "chore: update CHANGELOG for release ${{ github.ref_name }}"
-          body: |
-            **Description**:
-            - This PR updates the CHANGELOG.md by removing the "Unreleased" section and adding new release notes for release ${{ github.ref_name }} above previous releases.

.github/workflows/generate-unreleased-changelog-pr.yml

@@ -1,107 +0,0 @@
-name: Generate Unreleased Changelog PR
-
-on:
-  schedule:
-    - cron: "0 0 * * 1"  # Runs every Monday at 00:00 UTC
-  workflow_dispatch:
-
-jobs:
-  generate-unreleased-changelog-pr:
-    permissions:
-      contents: write    # Needed for pushing commits and creating branches.
-      pull-requests: write
-    runs-on: ubuntu-latest
-    steps:
-      # 1. Checkout the repository on main.
-      - name: Checkout Repository on Main
-        uses: actions/checkout@v4
-        with:
-          ref: main
-          fetch-depth: 0
-
-      # 4. Get the latest version tag.
-      - name: Get Latest Tag
-        id: get_latest_tag
-        run: |
-          LATEST_TAG=$(git describe --tags $(git rev-list --tags --max-count=1) || echo "none")
-          echo "Latest tag: $LATEST_TAG"
-          echo "tag=$LATEST_TAG" >> $GITHUB_OUTPUT
-
-      # 5. Generate the Unreleased changelog.
-      - name: Generate Unreleased Changelog
-        id: generate_unreleased
-        uses: mikepenz/release-changelog-builder-action@v5.1.0
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-        with:
-          configuration: ".github/configuration-unreleased.json"
-          owner: ${{ github.repository_owner }}
-          repo: ${{ github.event.repository.name }}
-          outputFile: CHANGELOG-unreleased.md
-          fromTag: ${{ steps.get_latest_tag.outputs.tag }}
-          toTag: main
-
-      # 7. Update CHANGELOG.md with the new Unreleased section.
-      - name: Update CHANGELOG.md
-        id: update_changelog
-        run: |
-          # Create CHANGELOG.md if it doesn't exist.
-          if [ ! -f CHANGELOG.md ]; then
-            echo "# Changelog" > CHANGELOG.md
-            echo "" >> CHANGELOG.md
-            echo "All notable changes to this project will be documented in this file." >> CHANGELOG.md
-            echo "" >> CHANGELOG.md
-          fi
-          
-          echo "Updating CHANGELOG.md…"
-          
-          # Extract content before the "## [Unreleased]" (or first version header if missing).
-          if grep -q "^## \[Unreleased\]" CHANGELOG.md; then
-            awk '/^## \[Unreleased\]/{exit} {print}' CHANGELOG.md > CHANGELOG_TMP.md
-          else
-            awk '/^## \[v/{exit} {print}' CHANGELOG.md > CHANGELOG_TMP.md
-          fi
-          
-          # Append the generated Unreleased changelog.
-          echo "" >> CHANGELOG_TMP.md
-          cat CHANGELOG-unreleased.md >> CHANGELOG_TMP.md
-          echo "" >> CHANGELOG_TMP.md
-          
-          # Append the remainder of the original changelog (starting from the first version header).
-          awk 'f{print} /^## \[v/{f=1; print}' CHANGELOG.md >> CHANGELOG_TMP.md
-          
-          # Replace the old file with the updated file.
-          mv CHANGELOG_TMP.md CHANGELOG.md
-          
-          # Remove the temporary generated file.
-          rm -f CHANGELOG-unreleased.md
-          
-          echo "Final CHANGELOG.md:"
-          cat CHANGELOG.md
-
-      # 8. Check if CHANGELOG.md has any updates.
-      - name: Check for CHANGELOG.md changes
-        id: changelog_changes
-        run: |
-          if git diff --quiet CHANGELOG.md; then
-            echo "has_changes=false" >> $GITHUB_OUTPUT
-          else
-            echo "has_changes=true" >> $GITHUB_OUTPUT
-          fi
-
-      # 9. Create (or update) the Pull Request only if there are changes.
-      - name: Create Pull Request
-        if: steps.changelog_changes.outputs.has_changes == 'true'
-        uses: peter-evans/create-pull-request@v7
-        with:
-          token: ${{ secrets.GITHUB_TOKEN }}
-          base: main
-          branch: "changelog/unreleased-update"
-          sign-commits: true
-          commit-message: "action: update Unreleased changelog"
-          title: "action: update Unreleased changelog"
-          body: |
-            **Description**:
-            - This PR updates the Unreleased section in CHANGELOG.md.
-            - It compares the current main branch with the latest version tag (determined as ${{ steps.get_latest_tag.outputs.tag }}),
-              regenerates the Unreleased changelog, removes any old Unreleased block, and inserts the new content.

.github/workflows/helmcharts.yml

@@ -4,12 +4,13 @@ name: Build Helm Charts on Tag
 on:
   push:
     tags:
-      - "*"
+      - "chart-*"
 
 jobs:
   release:
     permissions:
       contents: write
+      packages: write
     runs-on: ubuntu-latest
     steps:
       - name: Checkout
@@ -26,8 +27,49 @@ jobs:
         uses: azure/setup-helm@v4
         env:
           GITHUB_TOKEN: "${{ secrets.GITHUB_TOKEN }}"
+      
+      - name: Build Subchart Deps
+        run: |
+          cd helm/librechat
+          helm dependency build  
+          cd ../librechat-rag-api
+          helm dependency build      
 
-      - name: Run chart-releaser
-        uses: helm/chart-releaser-action@v1.6.0
-        env:
-          CR_TOKEN: "${{ secrets.GITHUB_TOKEN }}"
+      - name: Get Chart Version
+        id: chart-version
+        run: |
+          CHART_VERSION=$(echo "${{ github.ref_name }}" | cut -d'-' -f2)
+          echo "CHART_VERSION=${CHART_VERSION}" >> "$GITHUB_OUTPUT"
+
+      # Log in to GitHub Container Registry
+      - name: Log in to GitHub Container Registry
+        uses: docker/login-action@v3
+        with:
+          registry: ghcr.io
+          username: ${{ github.actor }}
+          password: ${{ secrets.GITHUB_TOKEN }}
+
+      # Run Helm OCI Charts Releaser
+      # This is for the librechat chart
+      - name: Release Helm OCI Charts for librechat
+        uses: appany/helm-oci-chart-releaser@v0.4.2
+        with:
+          name: librechat
+          repository: ${{ github.actor }}/librechat-chart
+          tag: ${{ steps.chart-version.outputs.CHART_VERSION }}
+          path: helm/librechat
+          registry: ghcr.io
+          registry_username: ${{ github.actor }}
+          registry_password: ${{ secrets.GITHUB_TOKEN }}
+      
+      # this is for the librechat-rag-api chart
+      - name: Release Helm OCI Charts for librechat-rag-api
+        uses: appany/helm-oci-chart-releaser@v0.4.2
+        with:
+          name: librechat-rag-api
+          repository: ${{ github.actor }}/librechat-chart
+          tag: ${{ steps.chart-version.outputs.CHART_VERSION }}
+          path: helm/librechat-rag-api
+          registry: ghcr.io
+          registry_username: ${{ github.actor }}
+          registry_password: ${{ secrets.GITHUB_TOKEN }}

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

@@ -1,16 +1,24 @@
 name: Detect Unused i18next Strings
 
+# This workflow checks for unused i18n keys in translation files.
+# It has special handling for:
+# - com_ui_special_var_* keys that are dynamically constructed
+# - com_agents_category_* keys that are stored in the database and used dynamically
+
 on:
   pull_request:
     paths:
       - "client/src/**"
       - "api/**"
+      - "packages/data-provider/src/**"
+      - "packages/client/**"
+      - "packages/data-schemas/src/**"
 
 jobs:
   detect-unused-i18n-keys:
     runs-on: ubuntu-latest
     permissions:
-      pull-requests: write  # Required for posting PR comments
+      pull-requests: write
     steps:
       - name: Checkout repository
         uses: actions/checkout@v3
@@ -22,7 +30,7 @@ jobs:
 
           # Define paths
           I18N_FILE="client/src/locales/en/translation.json"
-          SOURCE_DIRS=("client/src" "api")
+          SOURCE_DIRS=("client/src" "api" "packages/data-provider/src" "packages/client" "packages/data-schemas/src")
 
           # Check if translation file exists
           if [[ ! -f "$I18N_FILE" ]]; then
@@ -50,6 +58,31 @@ jobs:
                 fi
               done
               
+              # Also check if the key is directly used somewhere
+              if [[ "$FOUND" == false ]]; then
+                for DIR in "${SOURCE_DIRS[@]}"; do
+                  if grep -r --include=\*.{js,jsx,ts,tsx} -q "$KEY" "$DIR"; then
+                    FOUND=true
+                    break
+                  fi
+                done
+              fi
+            # Special case for agent category keys that are dynamically used from database
+            elif [[ "$KEY" == com_agents_category_* ]]; then
+              # Check if agent category localization is being used
+              for DIR in "${SOURCE_DIRS[@]}"; do
+                # Check for dynamic category label/description usage
+                if grep -r --include=\*.{js,jsx,ts,tsx} -E "category\.(label|description).*startsWith.*['\"]com_" "$DIR" > /dev/null 2>&1 || \
+                   # Check for the method that defines these keys
+                   grep -r --include=\*.{js,jsx,ts,tsx} "ensureDefaultCategories" "$DIR" > /dev/null 2>&1 || \
+                   # Check for direct usage in agentCategory.ts
+                   grep -r --include=\*.ts -E "label:.*['\"]$KEY['\"]" "$DIR" > /dev/null 2>&1 || \
+                   grep -r --include=\*.ts -E "description:.*['\"]$KEY['\"]" "$DIR" > /dev/null 2>&1; then
+                  FOUND=true
+                  break
+                fi
+              done
+              
               # Also check if the key is directly used somewhere
               if [[ "$FOUND" == false ]]; then
                 for DIR in "${SOURCE_DIRS[@]}"; do

.github/workflows/locize-i18n-sync.yml

@@ -48,7 +48,7 @@ jobs:
 
       # 2. Download translation files from locize.
       - name: Download Translations from locize
-        uses: locize/download@v1
+        uses: locize/download@v2
         with:
           project-id: ${{ secrets.LOCIZE_PROJECT_ID }}
           path: "client/src/locales"

.github/workflows/unused-packages.yml

@@ -7,6 +7,7 @@ on:
       - 'package-lock.json'
       - 'client/**'
       - 'api/**'
+      - 'packages/client/**'
 
 jobs:
   detect-unused-packages:
@@ -28,7 +29,7 @@ jobs:
 
       - name: Validate JSON files
         run: |
-          for FILE in package.json client/package.json api/package.json; do
+          for FILE in package.json client/package.json api/package.json packages/client/package.json; do
             if [[ -f "$FILE" ]]; then
               jq empty "$FILE" || (echo "::error title=Invalid JSON::$FILE is invalid" && exit 1)
             fi
@@ -63,12 +64,31 @@ jobs:
             local folder=$1
             local output_file=$2
             if [[ -d "$folder" ]]; then
-              grep -rEho "require\\(['\"]([a-zA-Z0-9@/._-]+)['\"]\\)" "$folder" --include=\*.{js,ts,mjs,cjs} | \
+              # Extract require() statements
+              grep -rEho "require\\(['\"]([a-zA-Z0-9@/._-]+)['\"]\\)" "$folder" --include=\*.{js,ts,tsx,jsx,mjs,cjs} | \
                 sed -E "s/require\\(['\"]([a-zA-Z0-9@/._-]+)['\"]\\)/\1/" > "$output_file"
           
-              grep -rEho "import .* from ['\"]([a-zA-Z0-9@/._-]+)['\"]" "$folder" --include=\*.{js,ts,mjs,cjs} | \
+              # Extract ES6 imports - various patterns
+              # import x from 'module'
+              grep -rEho "import .* from ['\"]([a-zA-Z0-9@/._-]+)['\"]" "$folder" --include=\*.{js,ts,tsx,jsx,mjs,cjs} | \
                 sed -E "s/import .* from ['\"]([a-zA-Z0-9@/._-]+)['\"]/\1/" >> "$output_file"
+              
+              # import 'module' (side-effect imports)
+              grep -rEho "import ['\"]([a-zA-Z0-9@/._-]+)['\"]" "$folder" --include=\*.{js,ts,tsx,jsx,mjs,cjs} | \
+                sed -E "s/import ['\"]([a-zA-Z0-9@/._-]+)['\"]/\1/" >> "$output_file"
+              
+              # export { x } from 'module' or export * from 'module'
+              grep -rEho "export .* from ['\"]([a-zA-Z0-9@/._-]+)['\"]" "$folder" --include=\*.{js,ts,tsx,jsx,mjs,cjs} | \
+                sed -E "s/export .* from ['\"]([a-zA-Z0-9@/._-]+)['\"]/\1/" >> "$output_file"
+              
+              # import type { x } from 'module' (TypeScript)
+              grep -rEho "import type .* from ['\"]([a-zA-Z0-9@/._-]+)['\"]" "$folder" --include=\*.{ts,tsx} | \
+                sed -E "s/import type .* from ['\"]([a-zA-Z0-9@/._-]+)['\"]/\1/" >> "$output_file"
           
+              # Remove subpath imports but keep the base package
+              # e.g., '@tanstack/react-query/devtools' becomes '@tanstack/react-query'
+              sed -i -E 's|^(@?[a-zA-Z0-9-]+(/[a-zA-Z0-9-]+)?)/.*|\1|' "$output_file"
+              
               sort -u "$output_file" -o "$output_file"
             else
               touch "$output_file"
@@ -78,13 +98,80 @@ jobs:
           extract_deps_from_code "." root_used_code.txt
           extract_deps_from_code "client" client_used_code.txt
           extract_deps_from_code "api" api_used_code.txt
+          
+          # Extract dependencies used by @librechat/client package
+          extract_deps_from_code "packages/client" packages_client_used_code.txt
+
+      - name: Get @librechat/client dependencies
+        id: get-librechat-client-deps
+        run: |
+          if [[ -f "packages/client/package.json" ]]; then
+            # Get all dependencies from @librechat/client (dependencies, devDependencies, and peerDependencies)
+            DEPS=$(jq -r '.dependencies // {} | keys[]' packages/client/package.json 2>/dev/null || echo "")
+            DEV_DEPS=$(jq -r '.devDependencies // {} | keys[]' packages/client/package.json 2>/dev/null || echo "")
+            PEER_DEPS=$(jq -r '.peerDependencies // {} | keys[]' packages/client/package.json 2>/dev/null || echo "")
+            
+            # Combine all dependencies
+            echo "$DEPS" > librechat_client_deps.txt
+            echo "$DEV_DEPS" >> librechat_client_deps.txt
+            echo "$PEER_DEPS" >> librechat_client_deps.txt
+            
+            # Also include dependencies that are imported in packages/client
+            cat packages_client_used_code.txt >> librechat_client_deps.txt
+            
+            # Remove empty lines and sort
+            grep -v '^$' librechat_client_deps.txt | sort -u > temp_deps.txt
+            mv temp_deps.txt librechat_client_deps.txt
+          else
+            touch librechat_client_deps.txt
+          fi
+
+      - name: Extract Workspace Dependencies
+        id: extract-workspace-deps
+        run: |
+          # Function to get dependencies from a workspace package that are used by another package
+          get_workspace_package_deps() {
+            local package_json=$1
+            local output_file=$2
+            
+            # Get all workspace dependencies (starting with @librechat/)
+            if [[ -f "$package_json" ]]; then
+              local workspace_deps=$(jq -r '.dependencies // {} | to_entries[] | select(.key | startswith("@librechat/")) | .key' "$package_json" 2>/dev/null || echo "")
+              
+              # For each workspace dependency, get its dependencies
+              for dep in $workspace_deps; do
+                # Convert @librechat/api to packages/api
+                local workspace_path=$(echo "$dep" | sed 's/@librechat\//packages\//')
+                local workspace_package_json="${workspace_path}/package.json"
+                
+                if [[ -f "$workspace_package_json" ]]; then
+                  # Extract all dependencies from the workspace package
+                  jq -r '.dependencies // {} | keys[]' "$workspace_package_json" 2>/dev/null >> "$output_file"
+                  # Also extract peerDependencies
+                  jq -r '.peerDependencies // {} | keys[]' "$workspace_package_json" 2>/dev/null >> "$output_file"
+                fi
+              done
+            fi
+            
+            if [[ -f "$output_file" ]]; then
+              sort -u "$output_file" -o "$output_file"
+            else
+              touch "$output_file"
+            fi
+          }
+          
+          # Get workspace dependencies for each package
+          get_workspace_package_deps "package.json" root_workspace_deps.txt
+          get_workspace_package_deps "client/package.json" client_workspace_deps.txt
+          get_workspace_package_deps "api/package.json" api_workspace_deps.txt
 
       - name: Run depcheck for root package.json
         id: check-root
         run: |
           if [[ -f "package.json" ]]; then
             UNUSED=$(depcheck --json | jq -r '.dependencies | join("\n")' || echo "")
-            UNUSED=$(comm -23 <(echo "$UNUSED" | sort) <(cat root_used_deps.txt root_used_code.txt | sort) || echo "")
+            # Exclude dependencies used in scripts, code, and workspace packages
+            UNUSED=$(comm -23 <(echo "$UNUSED" | sort) <(cat root_used_deps.txt root_used_code.txt root_workspace_deps.txt | sort) || echo "")
             echo "ROOT_UNUSED<<EOF" >> $GITHUB_ENV
             echo "$UNUSED" >> $GITHUB_ENV
             echo "EOF" >> $GITHUB_ENV
@@ -97,7 +184,10 @@ jobs:
             chmod -R 755 client
             cd client
             UNUSED=$(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 "")
+            # Exclude dependencies used in scripts, code, and workspace packages
+            UNUSED=$(comm -23 <(echo "$UNUSED" | sort) <(cat ../client_used_deps.txt ../client_used_code.txt ../client_workspace_deps.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" >> $GITHUB_ENV
@@ -111,7 +201,8 @@ jobs:
             chmod -R 755 api
             cd api
             UNUSED=$(depcheck --json | jq -r '.dependencies | join("\n")' || echo "")
-            UNUSED=$(comm -23 <(echo "$UNUSED" | sort) <(cat ../api_used_deps.txt ../api_used_code.txt | sort) || echo "")
+            # Exclude dependencies used in scripts, code, and workspace packages
+            UNUSED=$(comm -23 <(echo "$UNUSED" | sort) <(cat ../api_used_deps.txt ../api_used_code.txt ../api_workspace_deps.txt | sort) || echo "")
             echo "API_UNUSED<<EOF" >> $GITHUB_ENV
             echo "$UNUSED" >> $GITHUB_ENV
             echo "EOF" >> $GITHUB_ENV

.gitignore

@@ -13,6 +13,9 @@ pids
 *.seed
 .git
 
+# CI/CD data
+test-image*
+
 # Directory for instrumented libs generated by jscoverage/JSCover
 lib-cov
 
@@ -52,8 +55,10 @@ bower_components/
 *.d.ts
 !vite-env.d.ts
 
-# Cline
+# AI
 .clineignore
+.cursor
+.aider*
 
 # Floobits
 .floo
@@ -113,4 +118,23 @@ uploads/
 
 # owner
 release/
+
+# Helm
+helm/librechat/Chart.lock
+helm/**/charts/
+helm/**/.values.yaml
+
 !/client/src/@types/i18next.d.ts
+
+# SAML Idp cert
+*.cert
+
+# AI Assistants
+/.claude/
+/.cursor/
+/.copilot/
+/.aider/
+/.openai/
+/.tabnine/
+/.codeium
+*.local.md

.vscode/launch.json

@@ -8,7 +8,8 @@
       "skipFiles": ["<node_internals>/**"],
       "program": "${workspaceFolder}/api/server/index.js",
       "env": {
-        "NODE_ENV": "production"
+        "NODE_ENV": "production",
+        "NODE_TLS_REJECT_UNAUTHORIZED": "0"
       },
       "console": "integratedTerminal",
       "envFile": "${workspaceFolder}/.env"

CHANGELOG.md

@@ -3,10 +3,99 @@
 All notable changes to this project will be documented in this file.
 
 
+
+
+
+
 ## [Unreleased]
 
 ### ✨ New Features
 
+- ✨ feat: implement search parameter updates by **@mawburn** in [#7151](https://github.com/danny-avila/LibreChat/pull/7151)
+- 🎏 feat: Add MCP support for Streamable HTTP Transport by **@benverhees** in [#7353](https://github.com/danny-avila/LibreChat/pull/7353)
+- 🔒 feat: Add Content Security Policy using Helmet middleware by **@rubentalstra** in [#7377](https://github.com/danny-avila/LibreChat/pull/7377)
+- ✨ feat: Add Normalization for MCP Server Names by **@danny-avila** in [#7421](https://github.com/danny-avila/LibreChat/pull/7421)
+- 📊 feat: Improve Helm Chart by **@hofq** in [#3638](https://github.com/danny-avila/LibreChat/pull/3638)
+- 🦾 feat: Claude-4 Support by **@danny-avila** in [#7509](https://github.com/danny-avila/LibreChat/pull/7509)
+- 🪨 feat: Bedrock Support for Claude-4 Reasoning by **@danny-avila** in [#7517](https://github.com/danny-avila/LibreChat/pull/7517)
+
+### 🌍 Internationalization
+
+- 🌍 i18n: Add `Danish` and `Czech` and `Catalan` localization support by **@rubentalstra** in [#7373](https://github.com/danny-avila/LibreChat/pull/7373)
+- 🌍 i18n: Update translation.json with latest translations by **@github-actions[bot]** in [#7375](https://github.com/danny-avila/LibreChat/pull/7375)
+- 🌍 i18n: Update translation.json with latest translations by **@github-actions[bot]** in [#7468](https://github.com/danny-avila/LibreChat/pull/7468)
+
+### 🔧 Fixes
+
+- 💬 fix: update aria-label for accessibility in ConvoLink component by **@berry-13** in [#7320](https://github.com/danny-avila/LibreChat/pull/7320)
+- 🔑 fix: use `apiKey` instead of `openAIApiKey` in OpenAI-like Config by **@danny-avila** in [#7337](https://github.com/danny-avila/LibreChat/pull/7337)
+- 🔄 fix: update navigation logic in `useFocusChatEffect` to ensure correct search parameters are used by **@mawburn** in [#7340](https://github.com/danny-avila/LibreChat/pull/7340)
+- 🔄 fix: Improve MCP Connection Cleanup by **@danny-avila** in [#7400](https://github.com/danny-avila/LibreChat/pull/7400)
+- 🛡️ fix: Preset and Validation Logic for URL Query Params by **@danny-avila** in [#7407](https://github.com/danny-avila/LibreChat/pull/7407)
+- 🌘 fix: artifact of preview text is illegible in dark mode by **@nhtruong** in [#7405](https://github.com/danny-avila/LibreChat/pull/7405)
+- 🛡️ fix: Temporarily Remove CSP until Configurable by **@danny-avila** in [#7419](https://github.com/danny-avila/LibreChat/pull/7419)
+- 💽 fix: Exclude index page `/` from static cache settings by **@sbruel** in [#7382](https://github.com/danny-avila/LibreChat/pull/7382)
+
+### ⚙️ Other Changes
+
+- 📜 docs: CHANGELOG for release v0.7.8 by **@github-actions[bot]** in [#7290](https://github.com/danny-avila/LibreChat/pull/7290)
+- 📦 chore: Update API Package Dependencies by **@danny-avila** in [#7359](https://github.com/danny-avila/LibreChat/pull/7359)
+- 📜 docs: Unreleased Changelog by **@github-actions[bot]** in [#7321](https://github.com/danny-avila/LibreChat/pull/7321)
+- 📜 docs: Unreleased Changelog by **@github-actions[bot]** in [#7434](https://github.com/danny-avila/LibreChat/pull/7434)
+- 🛡️ chore: `multer` v2.0.0 for CVE-2025-47935 and CVE-2025-47944 by **@danny-avila** in [#7454](https://github.com/danny-avila/LibreChat/pull/7454)
+- 📂 refactor: Improve `FileAttachment` & File Form Deletion by **@danny-avila** in [#7471](https://github.com/danny-avila/LibreChat/pull/7471)
+- 📊 chore: Remove Old Helm Chart by **@hofq** in [#7512](https://github.com/danny-avila/LibreChat/pull/7512)
+- 🪖 chore: bump helm app version to v0.7.8 by **@austin-barrington** in [#7524](https://github.com/danny-avila/LibreChat/pull/7524)
+
+
+
+---
+## [v0.7.8] - 
+
+Changes from v0.7.8-rc1 to v0.7.8.
+
+### ✨ New Features
+
+- ✨ feat: Enhance form submission for touch screens by **@berry-13** in [#7198](https://github.com/danny-avila/LibreChat/pull/7198)
+- 🔍 feat: Additional Tavily API Tool Parameters by **@glowforge-opensource** in [#7232](https://github.com/danny-avila/LibreChat/pull/7232)
+- 🐋 feat: Add python to Dockerfile for increased MCP compatibility by **@technicalpickles** in [#7270](https://github.com/danny-avila/LibreChat/pull/7270)
+
+### 🔧 Fixes
+
+- 🔧 fix: Google Gemma Support & OpenAI Reasoning Instructions by **@danny-avila** in [#7196](https://github.com/danny-avila/LibreChat/pull/7196)
+- 🛠️ fix: Conversation Navigation State by **@danny-avila** in [#7210](https://github.com/danny-avila/LibreChat/pull/7210)
+- 🔄 fix: o-Series Model Regex for System Messages by **@danny-avila** in [#7245](https://github.com/danny-avila/LibreChat/pull/7245)
+- 🔖 fix: Custom Headers for Initial MCP SSE Connection by **@danny-avila** in [#7246](https://github.com/danny-avila/LibreChat/pull/7246)
+- 🛡️ fix: Deep Clone `MCPOptions` for User MCP Connections by **@danny-avila** in [#7247](https://github.com/danny-avila/LibreChat/pull/7247)
+- 🔄 fix: URL Param Race Condition and File Draft Persistence by **@danny-avila** in [#7257](https://github.com/danny-avila/LibreChat/pull/7257)
+- 🔄 fix: Assistants Endpoint & Minor Issues by **@danny-avila** in [#7274](https://github.com/danny-avila/LibreChat/pull/7274)
+- 🔄 fix: Ollama Think Tag Edge Case with Tools by **@danny-avila** in [#7275](https://github.com/danny-avila/LibreChat/pull/7275)
+
+### ⚙️ Other Changes
+
+- 📜 docs: CHANGELOG for release v0.7.8-rc1 by **@github-actions[bot]** in [#7153](https://github.com/danny-avila/LibreChat/pull/7153)
+- 🔄 refactor: Artifact Visibility Management by **@danny-avila** in [#7181](https://github.com/danny-avila/LibreChat/pull/7181)
+- 📦 chore: Bump Package Security by **@danny-avila** in [#7183](https://github.com/danny-avila/LibreChat/pull/7183)
+- 🌿 refactor: Unmount Fork Popover on Hide for Better Performance by **@danny-avila** in [#7189](https://github.com/danny-avila/LibreChat/pull/7189)
+- 🧰 chore: ESLint configuration to enforce Prettier formatting rules by **@mawburn** in [#7186](https://github.com/danny-avila/LibreChat/pull/7186)
+- 🎨 style: Improve KaTeX Rendering for LaTeX Equations by **@andresgit** in [#7223](https://github.com/danny-avila/LibreChat/pull/7223)
+- 📝 docs: Update `.env.example` Google models by **@marlonka** in [#7254](https://github.com/danny-avila/LibreChat/pull/7254)
+- 💬 refactor: MCP Chat Visibility Option, Google Rates, Remove OpenAPI Plugins by **@danny-avila** in [#7286](https://github.com/danny-avila/LibreChat/pull/7286)
+- 📜 docs: Unreleased Changelog by **@github-actions[bot]** in [#7214](https://github.com/danny-avila/LibreChat/pull/7214)
+
+
+
+[See full release details][release-v0.7.8]
+
+[release-v0.7.8]: https://github.com/danny-avila/LibreChat/releases/tag/v0.7.8
+
+---
+## [v0.7.8-rc1] - 
+
+Changes from v0.7.7 to v0.7.8-rc1.
+
+### ✨ New Features
+
 - 🔍 feat: Mistral OCR API / Upload Files as Text by **@danny-avila** in [#6274](https://github.com/danny-avila/LibreChat/pull/6274)
 - 🤖 feat: Support OpenAI Web Search models by **@danny-avila** in [#6313](https://github.com/danny-avila/LibreChat/pull/6313)
 - 🔗 feat: Agent Chain (Mixture-of-Agents) by **@danny-avila** in [#6374](https://github.com/danny-avila/LibreChat/pull/6374)
@@ -136,7 +225,12 @@ All notable changes to this project will be documented in this file.
 - 🧭 refactor: Modernize Nav/Header by **@danny-avila** in [#7094](https://github.com/danny-avila/LibreChat/pull/7094)
 - 🪶 refactor: Chat Input Focus for Conversation Navigations & ChatForm Optimizations by **@danny-avila** in [#7100](https://github.com/danny-avila/LibreChat/pull/7100)
 - 🔃 refactor: Streamline Navigation, Message Loading UX by **@danny-avila** in [#7118](https://github.com/danny-avila/LibreChat/pull/7118)
+- 📜 docs: Unreleased changelog by **@github-actions[bot]** in [#6265](https://github.com/danny-avila/LibreChat/pull/6265)
 
 
 
+[See full release details][release-v0.7.8-rc1]
+
+[release-v0.7.8-rc1]: https://github.com/danny-avila/LibreChat/releases/tag/v0.7.8-rc1
+
 ---

Dockerfile

@@ -1,10 +1,11 @@
-# v0.7.8-rc1
+# v0.8.0
 
 # Base node image
 FROM node:20-alpine AS node
 
 # Install jemalloc
 RUN apk add --no-cache jemalloc
+RUN apk add --no-cache python3 py3-pip uv
 
 # Set environment variable to use jemalloc
 ENV LD_PRELOAD=/usr/lib/libjemalloc.so.2
@@ -18,24 +19,31 @@ WORKDIR /app
 
 USER node
 
-COPY --chown=node:node . .
+COPY --chown=node:node package.json package-lock.json ./
+COPY --chown=node:node api/package.json ./api/package.json
+COPY --chown=node:node client/package.json ./client/package.json
+COPY --chown=node:node packages/data-provider/package.json ./packages/data-provider/package.json
+COPY --chown=node:node packages/data-schemas/package.json ./packages/data-schemas/package.json
+COPY --chown=node:node packages/api/package.json ./packages/api/package.json
 
 RUN \
     # Allow mounting of these files, which have no default
     touch .env ; \
     # Create directories for the volumes to inherit the correct permissions
-    mkdir -p /app/client/public/images /app/api/logs ; \
+    mkdir -p /app/client/public/images /app/api/logs /app/uploads ; \
     npm config set fetch-retry-maxtimeout 600000 ; \
     npm config set fetch-retries 5 ; \
     npm config set fetch-retry-mintimeout 15000 ; \
-    npm install --no-audit; \
+    npm ci --no-audit
+
+COPY --chown=node:node . .
+
+RUN \
     # React client build
     NODE_OPTIONS="--max-old-space-size=2048" npm run frontend; \
     npm prune --production; \
     npm cache clean --force
 
-RUN mkdir -p /app/client/public/images /app/api/logs
-
 # Node API setup
 EXPOSE 3080
 ENV HOST=0.0.0.0
@@ -46,4 +54,4 @@ CMD ["npm", "run", "backend"]
 # WORKDIR /usr/share/nginx/html
 # COPY --from=node /app/client/dist /usr/share/nginx/html
 # COPY client/nginx.conf /etc/nginx/conf.d/default.conf
-# ENTRYPOINT ["nginx", "-g", "daemon off;"]
+# ENTRYPOINT ["nginx", "-g", "daemon off;"]

Dockerfile.multi

@@ -1,5 +1,5 @@
 # Dockerfile.multi
-# v0.7.8-rc1
+# v0.8.0
 
 # Base for all builds
 FROM node:20-alpine AS base-min
@@ -14,8 +14,9 @@ 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/mcp/package*.json ./packages/mcp/
+COPY packages/api/package*.json ./packages/api/
 COPY packages/data-schemas/package*.json ./packages/data-schemas/
+COPY packages/client/package*.json ./packages/client/
 COPY client/package*.json ./client/
 COPY api/package*.json ./api/
 
@@ -24,31 +25,40 @@ FROM base-min AS base
 WORKDIR /app
 RUN npm ci
 
-# Build data-provider
+# Build `data-provider` package
 FROM base AS data-provider-build
 WORKDIR /app/packages/data-provider
 COPY packages/data-provider ./
 RUN npm run build
 
-# 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
+# Build `data-schemas` package
 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
+
+# Build `client` package
+FROM base AS client-package-build
+WORKDIR /app/packages/client
+COPY packages/client ./
+RUN npm run build
+
 # Client build
 FROM base AS client-build
 WORKDIR /app/client
 COPY client ./
 COPY --from=data-provider-build /app/packages/data-provider/dist /app/packages/data-provider/dist
+COPY --from=client-package-build /app/packages/client/dist /app/packages/client/dist
+COPY --from=client-package-build /app/packages/client/src /app/packages/client/src
 ENV NODE_OPTIONS="--max-old-space-size=2048"
 RUN npm run build
 
@@ -63,8 +73,8 @@ 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

@@ -52,7 +52,7 @@
 - 🖥️ **UI & Experience** inspired by ChatGPT with enhanced design and features
 
 - 🤖 **AI Model Selection**:  
-  - Anthropic (Claude), AWS Bedrock, OpenAI, Azure OpenAI, Google, Vertex AI, OpenAI Assistants API (incl. Azure)
+  - Anthropic (Claude), AWS Bedrock, OpenAI, Azure OpenAI, Google, Vertex AI, OpenAI Responses API (incl. Azure)
   - [Custom Endpoints](https://www.librechat.ai/docs/quick_start/custom_endpoints): Use any OpenAI-compatible API with LibreChat, no proxy required
   - Compatible with [Local & Remote AI Providers](https://www.librechat.ai/docs/configuration/librechat_yaml/ai_endpoints):
     - Ollama, groq, Cohere, Mistral AI, Apple MLX, koboldcpp, together.ai,
@@ -65,11 +65,18 @@
 
 - 🔦 **Agents & Tools Integration**:  
   - **[LibreChat Agents](https://www.librechat.ai/docs/features/agents)**:
-    - No-Code Custom Assistants: Build specialized, AI-driven helpers without coding  
-    - Flexible & Extensible: Attach tools like DALL-E-3, file search, code execution, and more  
-    - Compatible with Custom Endpoints, OpenAI, Azure, Anthropic, AWS Bedrock, and more
+    - No-Code Custom Assistants: Build specialized, AI-driven helpers
+    - Agent Marketplace: Discover and deploy community-built agents
+    - Collaborative Sharing: Share agents with specific users and groups
+    - Flexible & Extensible: Use MCP Servers, tools, file search, code execution, and more
+    - Compatible with Custom Endpoints, OpenAI, Azure, Anthropic, AWS Bedrock, Google, Vertex AI, Responses API, and more
     - [Model Context Protocol (MCP) Support](https://modelcontextprotocol.io/clients#librechat) for Tools
-  - Use LibreChat Agents and OpenAI Assistants with Files, Code Interpreter, Tools, and API Actions
+
+- 🔍 **Web Search**:  
+  - Search the internet and retrieve relevant information to enhance your AI context
+  - Combines search providers, content scrapers, and result rerankers for optimal results
+  - **Customizable Jina Reranking**: Configure custom Jina API URLs for reranking services
+  - **[Learn More →](https://www.librechat.ai/docs/features/web_search)**
 
 - 🪄 **Generative UI with Code Artifacts**:  
   - [Code Artifacts](https://youtu.be/GfTj7O4gmd0?si=WJbdnemZpJzBrJo3) allow creation of React, HTML, and Mermaid diagrams directly in chat
@@ -83,15 +90,18 @@
   - Create, Save, & Share Custom Presets  
   - Switch between AI Endpoints and Presets mid-chat
   - Edit, Resubmit, and Continue Messages with Conversation branching  
+  - Create and share prompts with specific users and groups
   - [Fork Messages & Conversations](https://www.librechat.ai/docs/features/fork) for Advanced Context control
 
 - 💬 **Multimodal & File Interactions**:  
   - Upload and analyze images with Claude 3, GPT-4.5, GPT-4o, o1, Llama-Vision, and Gemini 📸  
   - Chat with Files using Custom Endpoints, OpenAI, Azure, Anthropic, AWS Bedrock, & Google 🗃️
 
-- 🌎 **Multilingual UI**:  
-  - English, 中文, Deutsch, Español, Français, Italiano, Polski, Português Brasileiro
-  - Русский, 日本語, Svenska, 한국어, Tiếng Việt, 繁體中文, العربية, Türkçe, Nederlands, עברית
+- 🌎 **Multilingual UI**:
+  - English, 中文 (简体), 中文 (繁體), العربية, Deutsch, Español, Français, Italiano
+  - Polski, Português (PT), Português (BR), Русский, 日本語, Svenska, 한국어, Tiếng Việt
+  - Türkçe, Nederlands, עברית, Català, Čeština, Dansk, Eesti, فارسی
+  - Suomi, Magyar, Հայերեն, Bahasa Indonesia, ქართული, Latviešu, ไทย, ئۇيغۇرچە
 
 - 🧠 **Reasoning UI**:  
   - Dynamic Reasoning UI for Chain-of-Thought/Reasoning AI models like DeepSeek-R1
@@ -145,8 +155,8 @@ Click on the thumbnail to open the video☝️
 
 **Other:**
   - **Website:** [librechat.ai](https://librechat.ai)
-  - **Documentation:** [docs.librechat.ai](https://docs.librechat.ai)
-  - **Blog:** [blog.librechat.ai](https://blog.librechat.ai)
+  - **Documentation:** [librechat.ai/docs](https://librechat.ai/docs)
+  - **Blog:** [librechat.ai/blog](https://librechat.ai/blog)
 
 ---
 

api/app/clients/AnthropicClient.js

@@ -1,4 +1,5 @@
 const Anthropic = require('@anthropic-ai/sdk');
+const { logger } = require('@librechat/data-schemas');
 const { HttpsProxyAgent } = require('https-proxy-agent');
 const {
   Constants,
@@ -9,7 +10,18 @@ const {
   getResponseSender,
   validateVisionModel,
 } = require('librechat-data-provider');
-const { SplitStreamHandler: _Handler } = require('@librechat/agents');
+const { sleep, SplitStreamHandler: _Handler } = require('@librechat/agents');
+const {
+  Tokenizer,
+  createFetch,
+  matchModelName,
+  getClaudeHeaders,
+  getModelMaxTokens,
+  configureReasoning,
+  checkPromptCacheSupport,
+  getModelMaxOutputTokens,
+  createStreamEventHandlers,
+} = require('@librechat/api');
 const {
   truncateText,
   formatMessage,
@@ -18,19 +30,9 @@ const {
   parseParamFromPrompt,
   createContextHandlers,
 } = require('./prompts');
-const {
-  getClaudeHeaders,
-  configureReasoning,
-  checkPromptCacheSupport,
-} = require('~/server/services/Endpoints/anthropic/helpers');
-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');
 
 const HUMAN_PROMPT = '\n\nHuman:';
 const AI_PROMPT = '\n\nAssistant:';
@@ -70,13 +72,10 @@ class AnthropicClient extends BaseClient {
     this.message_delta;
     /** Whether the model is part of the Claude 3 Family
      * @type {boolean} */
-    this.isClaude3;
+    this.isClaudeLatest;
     /** Whether to use Messages API or Completions API
      * @type {boolean} */
     this.useMessages;
-    /** Whether or not the model is limited to the legacy amount of output tokens
-     * @type {boolean} */
-    this.isLegacyOutput;
     /** Whether or not the model supports Prompt Caching
      * @type {boolean} */
     this.supportsCacheControl;
@@ -116,21 +115,25 @@ class AnthropicClient extends BaseClient {
     );
 
     const modelMatch = matchModelName(this.modelOptions.model, EModelEndpoint.anthropic);
-    this.isClaude3 = modelMatch.includes('claude-3');
-    this.isLegacyOutput = !(
-      /claude-3[-.]5-sonnet/.test(modelMatch) || /claude-3[-.]7/.test(modelMatch)
+    this.isClaudeLatest =
+      /claude-[3-9]/.test(modelMatch) || /claude-(?:sonnet|opus|haiku)-[4-9]/.test(modelMatch);
+    const isLegacyOutput = !(
+      /claude-3[-.]5-sonnet/.test(modelMatch) ||
+      /claude-3[-.]7/.test(modelMatch) ||
+      /claude-(?:sonnet|opus|haiku)-[4-9]/.test(modelMatch) ||
+      /claude-[4-9]/.test(modelMatch)
     );
     this.supportsCacheControl = this.options.promptCache && checkPromptCacheSupport(modelMatch);
 
     if (
-      this.isLegacyOutput &&
+      isLegacyOutput &&
       this.modelOptions.maxOutputTokens &&
       this.modelOptions.maxOutputTokens > legacy.maxOutputTokens.default
     ) {
       this.modelOptions.maxOutputTokens = legacy.maxOutputTokens.default;
     }
 
-    this.useMessages = this.isClaude3 || !!this.options.attachments;
+    this.useMessages = this.isClaudeLatest || !!this.options.attachments;
 
     this.defaultVisionModel = this.options.visionModel ?? 'claude-3-sonnet-20240229';
     this.options.attachments?.then((attachments) => this.checkVisionRequest(attachments));
@@ -190,10 +193,11 @@ class AnthropicClient extends BaseClient {
         reverseProxyUrl: this.options.reverseProxyUrl,
       }),
       apiKey: this.apiKey,
+      fetchOptions: {},
     };
 
     if (this.options.proxy) {
-      options.httpAgent = new HttpsProxyAgent(this.options.proxy);
+      options.fetchOptions.agent = new HttpsProxyAgent(this.options.proxy);
     }
 
     if (this.options.reverseProxyUrl) {
@@ -396,13 +400,13 @@ class AnthropicClient extends BaseClient {
     const formattedMessages = orderedMessages.map((message, i) => {
       const formattedMessage = this.useMessages
         ? formatMessage({
-          message,
-          endpoint: EModelEndpoint.anthropic,
-        })
+            message,
+            endpoint: EModelEndpoint.anthropic,
+          })
         : {
-          author: message.isCreatedByUser ? this.userLabel : this.assistantLabel,
-          content: message?.content ?? message.text,
-        };
+            author: message.isCreatedByUser ? this.userLabel : this.assistantLabel,
+            content: message?.content ?? message.text,
+          };
 
       const needsTokenCount = this.contextStrategy && !orderedMessages[i].tokenCount;
       /* If tokens were never counted, or, is a Vision request and the message has files, count again */
@@ -654,7 +658,10 @@ class AnthropicClient extends BaseClient {
       );
     };
 
-    if (this.modelOptions.model.includes('claude-3')) {
+    if (
+      /claude-[3-9]/.test(this.modelOptions.model) ||
+      /claude-(?:sonnet|opus|haiku)-[4-9]/.test(this.modelOptions.model)
+    ) {
       await buildMessagesPayload();
       processTokens();
       return {
@@ -680,7 +687,7 @@ class AnthropicClient extends BaseClient {
   }
 
   getCompletion() {
-    logger.debug('AnthropicClient doesn\'t use getCompletion (all handled in sendCompletion)');
+    logger.debug("AnthropicClient doesn't use getCompletion (all handled in sendCompletion)");
   }
 
   /**
@@ -888,7 +895,7 @@ class AnthropicClient extends BaseClient {
   }
 
   getBuildMessagesOptions() {
-    logger.debug('AnthropicClient doesn\'t use getBuildMessagesOptions');
+    logger.debug("AnthropicClient doesn't use getBuildMessagesOptions");
   }
 
   getEncoding() {

api/app/clients/BaseClient.js

@@ -1,5 +1,7 @@
 const crypto = require('crypto');
 const fetch = require('node-fetch');
+const { logger } = require('@librechat/data-schemas');
+const { getBalanceConfig } = require('@librechat/api');
 const {
   supportsBalanceCheck,
   isAgentsEndpoint,
@@ -13,10 +15,8 @@ const {
 const { getMessages, saveMessage, updateMessage, saveConvo, getConvo } = require('~/models');
 const { checkBalance } = require('~/models/balanceMethods');
 const { truncateToolCallOutputs } = require('./prompts');
-const { addSpaceIfNeeded } = require('~/server/utils');
 const { getFiles } = require('~/models/File');
 const TextStream = require('./TextStream');
-const { logger } = require('~/config');
 
 class BaseClient {
   constructor(apiKey, options = {}) {
@@ -38,6 +38,8 @@ class BaseClient {
     this.conversationId;
     /** @type {string} */
     this.responseMessageId;
+    /** @type {string} */
+    this.parentMessageId;
     /** @type {TAttachment[]} */
     this.attachments;
     /** The key for the usage object's input tokens
@@ -63,15 +65,15 @@ class BaseClient {
   }
 
   setOptions() {
-    throw new Error('Method \'setOptions\' must be implemented.');
+    throw new Error("Method 'setOptions' must be implemented.");
   }
 
   async getCompletion() {
-    throw new Error('Method \'getCompletion\' must be implemented.');
+    throw new Error("Method 'getCompletion' must be implemented.");
   }
 
   async sendCompletion() {
-    throw new Error('Method \'sendCompletion\' must be implemented.');
+    throw new Error("Method 'sendCompletion' must be implemented.");
   }
 
   getSaveOptions() {
@@ -109,12 +111,17 @@ class BaseClient {
   /**
    * Abstract method to record token usage. Subclasses must implement this method.
    * If a correction to the token usage is needed, the method should return an object with the corrected token counts.
+   * Should only be used if `recordCollectedUsage` was not used instead.
+   * @param {string} [model]
+   * @param {AppConfig['balance']} [balance]
    * @param {number} promptTokens
    * @param {number} completionTokens
    * @returns {Promise<void>}
    */
-  async recordTokenUsage({ promptTokens, completionTokens }) {
+  async recordTokenUsage({ model, balance, promptTokens, completionTokens }) {
     logger.debug('[BaseClient] `recordTokenUsage` not implemented.', {
+      model,
+      balance,
       promptTokens,
       completionTokens,
     });
@@ -183,7 +190,8 @@ class BaseClient {
     this.user = user;
     const saveOptions = this.getSaveOptions();
     this.abortController = opts.abortController ?? new AbortController();
-    const conversationId = overrideConvoId ?? opts.conversationId ?? crypto.randomUUID();
+    const requestConvoId = overrideConvoId ?? opts.conversationId;
+    const conversationId = requestConvoId ?? crypto.randomUUID();
     const parentMessageId = opts.parentMessageId ?? Constants.NO_PARENT;
     const userMessageId =
       overrideUserMessageId ?? opts.overrideParentMessageId ?? crypto.randomUUID();
@@ -198,17 +206,22 @@ class BaseClient {
       this.currentMessages[this.currentMessages.length - 1].messageId = head;
     }
 
+    if (opts.isRegenerate && responseMessageId.endsWith('_')) {
+      responseMessageId = crypto.randomUUID();
+    }
+
     this.responseMessageId = responseMessageId;
 
     return {
       ...opts,
       user,
       head,
+      saveOptions,
+      userMessageId,
+      requestConvoId,
       conversationId,
       parentMessageId,
-      userMessageId,
       responseMessageId,
-      saveOptions,
     };
   }
 
@@ -227,21 +240,22 @@ class BaseClient {
     const {
       user,
       head,
+      saveOptions,
+      userMessageId,
+      requestConvoId,
       conversationId,
       parentMessageId,
-      userMessageId,
       responseMessageId,
-      saveOptions,
     } = await this.setMessageOptions(opts);
 
     const userMessage = opts.isEdited
       ? this.currentMessages[this.currentMessages.length - 2]
       : this.createUserMessage({
-        messageId: userMessageId,
-        parentMessageId,
-        conversationId,
-        text: message,
-      });
+          messageId: userMessageId,
+          parentMessageId,
+          conversationId,
+          text: message,
+        });
 
     if (typeof opts?.getReqData === 'function') {
       opts.getReqData({
@@ -253,7 +267,8 @@ class BaseClient {
     }
 
     if (typeof opts?.onStart === 'function') {
-      opts.onStart(userMessage, responseMessageId);
+      const isNewConvo = !requestConvoId && parentMessageId === Constants.NO_PARENT;
+      opts.onStart(userMessage, responseMessageId, isNewConvo);
     }
 
     return {
@@ -559,6 +574,7 @@ class BaseClient {
   }
 
   async sendMessage(message, opts = {}) {
+    const appConfig = this.options.req?.config;
     /** @type {Promise<TMessage>} */
     let userMessagePromise;
     const { user, head, isEdited, conversationId, responseMessageId, saveOptions, userMessage } =
@@ -572,7 +588,7 @@ class BaseClient {
       });
     }
 
-    const { generation = '' } = opts;
+    const { editedContent } = opts;
 
     // It's not necessary to push to currentMessages
     // depending on subclass implementation of handling messages
@@ -587,26 +603,40 @@ class BaseClient {
           isCreatedByUser: false,
           model: this.modelOptions?.model ?? this.model,
           sender: this.sender,
-          text: generation,
         };
         this.currentMessages.push(userMessage, latestMessage);
-      } else {
-        latestMessage.text = generation;
+      } else if (editedContent != null) {
+        // Handle editedContent for content parts
+        if (editedContent && latestMessage.content && Array.isArray(latestMessage.content)) {
+          const { index, text, type } = editedContent;
+          if (index >= 0 && index < latestMessage.content.length) {
+            const contentPart = latestMessage.content[index];
+            if (type === ContentTypes.THINK && contentPart.type === ContentTypes.THINK) {
+              contentPart[ContentTypes.THINK] = text;
+            } else if (type === ContentTypes.TEXT && contentPart.type === ContentTypes.TEXT) {
+              contentPart[ContentTypes.TEXT] = text;
+            }
+          }
+        }
       }
       this.continued = true;
     } else {
       this.currentMessages.push(userMessage);
     }
 
+    /**
+     * When the userMessage is pushed to currentMessages, the parentMessage is the userMessageId.
+     * this only matters when buildMessages is utilizing the parentMessageId, and may vary on implementation
+     */
+    const parentMessageId = isEdited ? head : userMessage.messageId;
+    this.parentMessageId = parentMessageId;
     let {
       prompt: payload,
       tokenCountMap,
       promptTokens,
     } = await this.buildMessages(
       this.currentMessages,
-      // When the userMessage is pushed to currentMessages, the parentMessage is the userMessageId.
-      // this only matters when buildMessages is utilizing the parentMessageId, and may vary on implementation
-      isEdited ? head : userMessage.messageId,
+      parentMessageId,
       this.getBuildMessagesOptions(opts),
       opts,
     );
@@ -631,9 +661,9 @@ class BaseClient {
       }
     }
 
-    const balance = this.options.req?.app?.locals?.balance;
+    const balanceConfig = getBalanceConfig(appConfig);
     if (
-      balance?.enabled &&
+      balanceConfig?.enabled &&
       supportsBalanceCheck[this.options.endpointType ?? this.options.endpoint]
     ) {
       await checkBalance({
@@ -672,16 +702,32 @@ class BaseClient {
     };
 
     if (typeof completion === 'string') {
-      responseMessage.text = addSpaceIfNeeded(generation) + completion;
+      responseMessage.text = completion;
     } else if (
       Array.isArray(completion) &&
       (this.clientName === EModelEndpoint.agents ||
         isParamEndpoint(this.options.endpoint, this.options.endpointType))
     ) {
       responseMessage.text = '';
-      responseMessage.content = completion;
+
+      if (!opts.editedContent || this.currentMessages.length === 0) {
+        responseMessage.content = completion;
+      } else {
+        const latestMessage = this.currentMessages[this.currentMessages.length - 1];
+        if (!latestMessage?.content) {
+          responseMessage.content = completion;
+        } else {
+          const existingContent = [...latestMessage.content];
+          const { type: editedType } = opts.editedContent;
+          responseMessage.content = this.mergeEditedContent(
+            existingContent,
+            completion,
+            editedType,
+          );
+        }
+      }
     } else if (Array.isArray(completion)) {
-      responseMessage.text = addSpaceIfNeeded(generation) + completion.join('');
+      responseMessage.text = completion.join('');
     }
 
     if (
@@ -712,9 +758,14 @@ class BaseClient {
       } else {
         responseMessage.tokenCount = this.getTokenCountForResponse(responseMessage);
         completionTokens = responseMessage.tokenCount;
+        await this.recordTokenUsage({
+          usage,
+          promptTokens,
+          completionTokens,
+          balance: balanceConfig,
+          model: responseMessage.model,
+        });
       }
-
-      await this.recordTokenUsage({ promptTokens, completionTokens, usage });
     }
 
     if (userMessagePromise) {
@@ -792,7 +843,8 @@ class BaseClient {
 
     userMessage.tokenCount = userMessageTokenCount;
     /*
-      Note: `AskController` saves the user message, so we update the count of its `userMessage` reference
+      Note: `AgentController` saves the user message if not saved here
+      (noted by `savedMessageIds`), so we update the count of its `userMessage` reference
     */
     if (typeof opts?.getReqData === 'function') {
       opts.getReqData({
@@ -801,7 +853,8 @@ class BaseClient {
     }
     /*
       Note: we update the user message to be sure it gets the calculated token count;
-      though `AskController` saves the user message, EditController does not
+      though `AgentController` saves the user message if not saved here
+      (noted by `savedMessageIds`), EditController does not
     */
     await userMessagePromise;
     await this.updateMessageInDatabase({
@@ -1093,6 +1146,50 @@ class BaseClient {
     return numTokens;
   }
 
+  /**
+   * Merges completion content with existing content when editing TEXT or THINK types
+   * @param {Array} existingContent - The existing content array
+   * @param {Array} newCompletion - The new completion content
+   * @param {string} editedType - The type of content being edited
+   * @returns {Array} The merged content array
+   */
+  mergeEditedContent(existingContent, newCompletion, editedType) {
+    if (!newCompletion.length) {
+      return existingContent.concat(newCompletion);
+    }
+
+    if (editedType !== ContentTypes.TEXT && editedType !== ContentTypes.THINK) {
+      return existingContent.concat(newCompletion);
+    }
+
+    const lastIndex = existingContent.length - 1;
+    const lastExisting = existingContent[lastIndex];
+    const firstNew = newCompletion[0];
+
+    if (lastExisting?.type !== firstNew?.type || firstNew?.type !== editedType) {
+      return existingContent.concat(newCompletion);
+    }
+
+    const mergedContent = [...existingContent];
+    if (editedType === ContentTypes.TEXT) {
+      mergedContent[lastIndex] = {
+        ...mergedContent[lastIndex],
+        [ContentTypes.TEXT]:
+          (mergedContent[lastIndex][ContentTypes.TEXT] || '') + (firstNew[ContentTypes.TEXT] || ''),
+      };
+    } else {
+      mergedContent[lastIndex] = {
+        ...mergedContent[lastIndex],
+        [ContentTypes.THINK]:
+          (mergedContent[lastIndex][ContentTypes.THINK] || '') +
+          (firstNew[ContentTypes.THINK] || ''),
+      };
+    }
+
+    // Add remaining completion items
+    return mergedContent.concat(newCompletion.slice(1));
+  }
+
   async sendPayload(payload, opts = {}) {
     if (opts && typeof opts === 'object') {
       this.setOptions(opts);

api/app/clients/ChatGPTClient.js

@@ -1,804 +0,0 @@
-const { Keyv } = require('keyv');
-const crypto = require('crypto');
-const { CohereClient } = require('cohere-ai');
-const { fetchEventSource } = require('@waylaidwanderer/fetch-event-source');
-const { 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

@@ -1,6 +1,10 @@
 const { google } = require('googleapis');
+const { sleep } = require('@librechat/agents');
+const { logger } = require('@librechat/data-schemas');
+const { getModelMaxTokens } = require('@librechat/api');
 const { concat } = require('@langchain/core/utils/stream');
 const { ChatVertexAI } = require('@langchain/google-vertexai');
+const { Tokenizer, getSafetySettings } = require('@librechat/api');
 const { ChatGoogleGenerativeAI } = require('@langchain/google-genai');
 const { GoogleGenerativeAI: GenAI } = require('@google/generative-ai');
 const { HumanMessage, SystemMessage } = require('@langchain/core/messages');
@@ -11,19 +15,15 @@ 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');
-const { logger } = require('~/config');
 const {
   formatMessage,
   createContextHandlers,
@@ -34,7 +34,8 @@ const BaseClient = require('./BaseClient');
 
 const loc = process.env.GOOGLE_LOC || 'us-central1';
 const publisher = 'google';
-const endpointPrefix = `${loc}-aiplatform.googleapis.com`;
+const endpointPrefix =
+  loc === 'global' ? 'aiplatform.googleapis.com' : `${loc}-aiplatform.googleapis.com`;
 
 const settings = endpointSettings[EModelEndpoint.google];
 const EXCLUDED_GENAI_MODELS = /gemini-(?:1\.0|1-0|pro)/;
@@ -140,8 +141,7 @@ class GoogleClient extends BaseClient {
     this.options.attachments?.then((attachments) => this.checkVisionRequest(attachments));
 
     /** @type {boolean} Whether using a "GenerativeAI" Model */
-    this.isGenerativeModel =
-      this.modelOptions.model.includes('gemini') || this.modelOptions.model.includes('learnlm');
+    this.isGenerativeModel = /gemini|learnlm|gemma/.test(this.modelOptions.model);
 
     this.maxContextTokens =
       this.options.maxContextTokens ??
@@ -166,6 +166,16 @@ class GoogleClient extends BaseClient {
       );
     }
 
+    // Add thinking configuration
+    this.modelOptions.thinkingConfig = {
+      thinkingBudget:
+        (this.modelOptions.thinking ?? googleSettings.thinking.default)
+          ? this.modelOptions.thinkingBudget
+          : 0,
+    };
+    delete this.modelOptions.thinking;
+    delete this.modelOptions.thinkingBudget;
+
     this.sender =
       this.options.sender ??
       getResponseSender({
@@ -237,11 +247,11 @@ 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

@@ -1,10 +1,11 @@
 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, logAxiosError } = require('~/utils');
-const { sleep } = require('~/server/utils');
-const { logger } = require('~/config');
+const { deriveBaseURL } = require('~/utils');
 
 const ollamaPayloadSchema = z.object({
   mirostat: z.number().optional(),
@@ -67,7 +68,7 @@ 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

@@ -1,13 +1,23 @@
-const { OllamaClient } = require('./OllamaClient');
+const { logger } = require('@librechat/data-schemas');
 const { HttpsProxyAgent } = require('https-proxy-agent');
-const { SplitStreamHandler, CustomOpenAIClient: OpenAI } = require('@librechat/agents');
+const { sleep, SplitStreamHandler, CustomOpenAIClient: OpenAI } = require('@librechat/agents');
+const {
+  isEnabled,
+  Tokenizer,
+  createFetch,
+  resolveHeaders,
+  constructAzureURL,
+  getModelMaxTokens,
+  genAzureChatCompletion,
+  getModelMaxOutputTokens,
+  createStreamEventHandlers,
+} = require('@librechat/api');
 const {
   Constants,
   ImageDetail,
   ContentTypes,
   parseTextParts,
   EModelEndpoint,
-  resolveHeaders,
   KnownEndpoints,
   openAISettings,
   ImageDetailCost,
@@ -16,13 +26,6 @@ const {
   validateVisionModel,
   mapModelToAzureConfig,
 } = require('librechat-data-provider');
-const {
-  extractBaseURL,
-  constructAzureURL,
-  getModelMaxTokens,
-  genAzureChatCompletion,
-  getModelMaxOutputTokens,
-} = require('~/utils');
 const {
   truncateText,
   formatMessage,
@@ -31,28 +34,20 @@ const {
   createContextHandlers,
 } = require('./prompts');
 const { encodeAndFormat } = require('~/server/services/Files/images/encode');
-const { createFetch, createStreamEventHandlers } = require('./generators');
-const { addSpaceIfNeeded, isEnabled, sleep } = require('~/server/utils');
-const Tokenizer = require('~/server/services/Tokenizer');
 const { spendTokens } = require('~/models/spendTokens');
+const { addSpaceIfNeeded } = require('~/server/utils');
 const { handleOpenAIErrors } = require('./tools/util');
-const { createLLM, RunManager } = require('./llm');
-const ChatGPTClient = require('./ChatGPTClient');
+const { OllamaClient } = require('./OllamaClient');
 const { summaryBuffer } = require('./memory');
 const { runTitleChain } = require('./chains');
+const { extractBaseURL } = require('~/utils');
 const { tokenSplit } = require('./document');
 const BaseClient = require('./BaseClient');
-const { logger } = require('~/config');
+const { createLLM } = require('./llm');
 
 class OpenAIClient extends BaseClient {
   constructor(apiKey, options = {}) {
     super(apiKey, options);
-    this.ChatGPTClient = new ChatGPTClient();
-    this.buildPrompt = this.ChatGPTClient.buildPrompt.bind(this);
-    /** @type {getCompletion} */
-    this.getCompletion = this.ChatGPTClient.getCompletion.bind(this);
-    /** @type {cohereChatCompletion} */
-    this.cohereChatCompletion = this.ChatGPTClient.cohereChatCompletion.bind(this);
     this.contextStrategy = options.contextStrategy
       ? options.contextStrategy.toLowerCase()
       : 'discard';
@@ -379,23 +374,12 @@ class OpenAIClient extends BaseClient {
     return files;
   }
 
-  async buildMessages(
-    messages,
-    parentMessageId,
-    { isChatCompletion = false, promptPrefix = null },
-    opts,
-  ) {
+  async buildMessages(messages, parentMessageId, { promptPrefix = null }, opts) {
     let orderedMessages = this.constructor.getMessagesForConversation({
       messages,
       parentMessageId,
       summary: this.shouldSummarize,
     });
-    if (!isChatCompletion) {
-      return await this.buildPrompt(orderedMessages, {
-        isChatGptModel: isChatCompletion,
-        promptPrefix,
-      });
-    }
 
     let payload;
     let instructions;
@@ -475,7 +459,9 @@ class OpenAIClient extends BaseClient {
       promptPrefix = this.augmentedPrompt + promptPrefix;
     }
 
-    if (promptPrefix && this.isOmni !== true) {
+    const noSystemModelRegex = /\b(o1-preview|o1-mini)\b/i.test(this.modelOptions.model);
+
+    if (promptPrefix && !noSystemModelRegex) {
       promptPrefix = `Instructions:\n${promptPrefix.trim()}`;
       instructions = {
         role: 'system',
@@ -503,7 +489,7 @@ class OpenAIClient extends BaseClient {
     };
 
     /** EXPERIMENTAL */
-    if (promptPrefix && this.isOmni === true) {
+    if (promptPrefix && noSystemModelRegex) {
       const lastUserMessageIndex = payload.findLastIndex((message) => message.role === 'user');
       if (lastUserMessageIndex !== -1) {
         if (Array.isArray(payload[lastUserMessageIndex].content)) {
@@ -634,10 +620,6 @@ class OpenAIClient extends BaseClient {
     temperature = 0.2,
     max_tokens,
     streaming,
-    context,
-    tokenBuffer,
-    initialMessageCount,
-    conversationId,
   }) {
     const modelOptions = {
       modelName: modelName ?? model,
@@ -669,8 +651,10 @@ class OpenAIClient extends BaseClient {
     if (headers && typeof headers === 'object' && !Array.isArray(headers)) {
       configOptions.baseOptions = {
         headers: resolveHeaders({
-          ...headers,
-          ...configOptions?.baseOptions?.headers,
+          headers: {
+            ...headers,
+            ...configOptions?.baseOptions?.headers,
+          },
         }),
       };
     }
@@ -680,22 +664,12 @@ class OpenAIClient extends BaseClient {
       configOptions.httpsAgent = new HttpsProxyAgent(this.options.proxy);
     }
 
-    const { req, res, debug } = this.options;
-    const runManager = new RunManager({ req, res, debug, abortController: this.abortController });
-    this.runManager = runManager;
-
     const llm = createLLM({
       modelOptions,
       configOptions,
       openAIApiKey: this.apiKey,
       azure: this.azure,
       streaming,
-      callbacks: runManager.createCallbacks({
-        context,
-        tokenBuffer,
-        conversationId: this.conversationId ?? conversationId,
-        initialMessageCount,
-      }),
     });
 
     return llm;
@@ -716,6 +690,7 @@ class OpenAIClient extends BaseClient {
    *                            In case of failure, it will return the default title, "New Chat".
    */
   async titleConvo({ text, conversationId, responseText = '' }) {
+    const appConfig = this.options.req?.config;
     this.conversationId = conversationId;
 
     if (this.options.attachments) {
@@ -744,8 +719,7 @@ class OpenAIClient extends BaseClient {
       max_tokens: 16,
     };
 
-    /** @type {TAzureConfig | undefined} */
-    const azureConfig = this.options?.req?.app?.locals?.[EModelEndpoint.azureOpenAI];
+    const azureConfig = appConfig?.endpoints?.[EModelEndpoint.azureOpenAI];
 
     const resetTitleOptions = !!(
       (this.azure && azureConfig) ||
@@ -765,7 +739,7 @@ class OpenAIClient extends BaseClient {
         groupMap,
       });
 
-      this.options.headers = resolveHeaders(headers);
+      this.options.headers = resolveHeaders({ headers });
       this.options.reverseProxyUrl = baseURL ?? null;
       this.langchainProxy = extractBaseURL(this.options.reverseProxyUrl);
       this.apiKey = azureOptions.azureOpenAIApiKey;
@@ -1134,6 +1108,7 @@ ${convo}
   }
 
   async chatCompletion({ payload, onProgress, abortController = null }) {
+    const appConfig = this.options.req?.config;
     let error = null;
     let intermediateReply = [];
     const errorCallback = (err) => (error = err);
@@ -1157,6 +1132,7 @@ ${convo}
       logger.debug('[OpenAIClient] chatCompletion', { baseURL, modelOptions });
       const opts = {
         baseURL,
+        fetchOptions: {},
       };
 
       if (this.useOpenRouter) {
@@ -1175,11 +1151,10 @@ ${convo}
       }
 
       if (this.options.proxy) {
-        opts.httpAgent = new HttpsProxyAgent(this.options.proxy);
+        opts.fetchOptions.agent = new HttpsProxyAgent(this.options.proxy);
       }
 
-      /** @type {TAzureConfig | undefined} */
-      const azureConfig = this.options?.req?.app?.locals?.[EModelEndpoint.azureOpenAI];
+      const azureConfig = appConfig?.endpoints?.[EModelEndpoint.azureOpenAI];
 
       if (
         (this.azure && this.isVisionModel && azureConfig) ||
@@ -1196,7 +1171,7 @@ ${convo}
           modelGroupMap,
           groupMap,
         });
-        opts.defaultHeaders = resolveHeaders(headers);
+        opts.defaultHeaders = resolveHeaders({ headers });
         this.langchainProxy = extractBaseURL(baseURL);
         this.apiKey = azureOptions.azureOpenAIApiKey;
 
@@ -1227,9 +1202,9 @@ ${convo}
 
         opts.baseURL = this.langchainProxy
           ? constructAzureURL({
-            baseURL: this.langchainProxy,
-            azureOptions: this.azure,
-          })
+              baseURL: this.langchainProxy,
+              azureOptions: this.azure,
+            })
           : this.azureEndpoint.split(/(?<!\/)\/(chat|completion)\//)[0];
 
         opts.defaultQuery = { 'api-version': this.azure.azureOpenAIApiVersion };
@@ -1237,7 +1212,9 @@ ${convo}
       }
 
       if (this.isOmni === true && modelOptions.max_tokens != null) {
-        modelOptions.max_completion_tokens = modelOptions.max_tokens;
+        const paramName =
+          modelOptions.useResponsesApi === true ? 'max_output_tokens' : 'max_completion_tokens';
+        modelOptions[paramName] = modelOptions.max_tokens;
         delete modelOptions.max_tokens;
       }
       if (this.isOmni === true && modelOptions.temperature != null) {
@@ -1283,6 +1260,14 @@ ${convo}
         modelOptions.messages[0].role = 'user';
       }
 
+      if (
+        (this.options.endpoint === EModelEndpoint.openAI ||
+          this.options.endpoint === EModelEndpoint.azureOpenAI) &&
+        modelOptions.stream === true
+      ) {
+        modelOptions.stream_options = { include_usage: true };
+      }
+
       if (this.options.addParams && typeof this.options.addParams === 'object') {
         const addParams = { ...this.options.addParams };
         modelOptions = {
@@ -1385,13 +1370,7 @@ ${convo}
           ...modelOptions,
           stream: true,
         };
-        if (
-          this.options.endpoint === EModelEndpoint.openAI ||
-          this.options.endpoint === EModelEndpoint.azureOpenAI
-        ) {
-          params.stream_options = { include_usage: true };
-        }
-        const stream = await openai.beta.chat.completions
+        const stream = await openai.chat.completions
           .stream(params)
           .on('abort', () => {
             /* Do nothing here */

api/app/clients/PluginsClient.js

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

api/app/clients/TextStream.js

@@ -1,5 +1,5 @@
 const { Readable } = require('stream');
-const { logger } = require('~/config');
+const { logger } = require('@librechat/data-schemas');
 
 class TextStream extends Readable {
   constructor(text, options = {}) {

api/app/clients/agents/CustomAgent/outputParser.js

@@ -1,5 +1,5 @@
+const { logger } = require('@librechat/data-schemas');
 const { ZeroShotAgentOutputParser } = require('langchain/agents');
-const { logger } = require('~/config');
 
 class CustomOutputParser extends ZeroShotAgentOutputParser {
   constructor(fields) {

api/app/clients/callbacks/createStartHandler.js

@@ -1,95 +0,0 @@
-const { promptTokensEstimate } = require('openai-chat-tokens');
-const { EModelEndpoint, supportsBalanceCheck } = require('librechat-data-provider');
-const { formatFromLangChain } = require('~/app/clients/prompts');
-const { getBalanceConfig } = require('~/server/services/Config');
-const { checkBalance } = require('~/models/balanceMethods');
-const { logger } = require('~/config');
-
-const createStartHandler = ({
-  context,
-  conversationId,
-  tokenBuffer = 0,
-  initialMessageCount,
-  manager,
-}) => {
-  return async (_llm, _messages, runId, parentRunId, extraParams) => {
-    const { invocation_params } = extraParams;
-    const { model, functions, function_call } = invocation_params;
-    const messages = _messages[0].map(formatFromLangChain);
-
-    logger.debug(`[createStartHandler] handleChatModelStart: ${context}`, {
-      model,
-      function_call,
-    });
-
-    if (context !== 'title') {
-      logger.debug(`[createStartHandler] handleChatModelStart: ${context}`, {
-        functions,
-      });
-    }
-
-    const payload = { messages };
-    let prelimPromptTokens = 1;
-
-    if (functions) {
-      payload.functions = functions;
-      prelimPromptTokens += 2;
-    }
-
-    if (function_call) {
-      payload.function_call = function_call;
-      prelimPromptTokens -= 5;
-    }
-
-    prelimPromptTokens += promptTokensEstimate(payload);
-    logger.debug('[createStartHandler]', {
-      prelimPromptTokens,
-      tokenBuffer,
-    });
-    prelimPromptTokens += tokenBuffer;
-
-    try {
-      const balance = await getBalanceConfig();
-      if (balance?.enabled && supportsBalanceCheck[EModelEndpoint.openAI]) {
-        const generations =
-          initialMessageCount && messages.length > initialMessageCount
-            ? messages.slice(initialMessageCount)
-            : null;
-        await checkBalance({
-          req: manager.req,
-          res: manager.res,
-          txData: {
-            user: manager.user,
-            tokenType: 'prompt',
-            amount: prelimPromptTokens,
-            debug: manager.debug,
-            generations,
-            model,
-            endpoint: EModelEndpoint.openAI,
-          },
-        });
-      }
-    } catch (err) {
-      logger.error(`[createStartHandler][${context}] checkBalance error`, err);
-      manager.abortController.abort();
-      if (context === 'summary' || context === 'plugins') {
-        manager.addRun(runId, { conversationId, error: err.message });
-        throw new Error(err);
-      }
-      return;
-    }
-
-    manager.addRun(runId, {
-      model,
-      messages,
-      functions,
-      function_call,
-      runId,
-      parentRunId,
-      conversationId,
-      prelimPromptTokens,
-    });
-  };
-};
-
-module.exports = createStartHandler;

api/app/clients/callbacks/index.js

@@ -1,5 +0,0 @@
-const createStartHandler = require('./createStartHandler');
-
-module.exports = {
-  createStartHandler,
-};

api/app/clients/chains/runTitleChain.js

@@ -1,7 +1,7 @@
 const { z } = require('zod');
+const { logger } = require('@librechat/data-schemas');
 const { langPrompt, createTitlePrompt, escapeBraces, getSnippet } = require('../prompts');
 const { createStructuredOutputChainFromZod } = require('langchain/chains/openai_functions');
-const { logger } = require('~/config');
 
 const langSchema = z.object({
   language: z.string().describe('The language of the input text (full noun, no abbreviations).'),

api/app/clients/generators.js

@@ -1,71 +0,0 @@
-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

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

api/app/clients/llm/RunManager.js

@@ -1,105 +0,0 @@
-const { createStartHandler } = require('~/app/clients/callbacks');
-const { spendTokens } = require('~/models/spendTokens');
-const { logger } = require('~/config');
-
-class RunManager {
-  constructor(fields) {
-    const { req, res, abortController, debug } = fields;
-    this.abortController = abortController;
-    this.user = req.user.id;
-    this.req = req;
-    this.res = res;
-    this.debug = debug;
-    this.runs = new Map();
-    this.convos = new Map();
-  }
-
-  addRun(runId, runData) {
-    if (!this.runs.has(runId)) {
-      this.runs.set(runId, runData);
-      if (runData.conversationId) {
-        this.convos.set(runData.conversationId, runId);
-      }
-      return runData;
-    } else {
-      const existingData = this.runs.get(runId);
-      const update = { ...existingData, ...runData };
-      this.runs.set(runId, update);
-      if (update.conversationId) {
-        this.convos.set(update.conversationId, runId);
-      }
-      return update;
-    }
-  }
-
-  removeRun(runId) {
-    if (this.runs.has(runId)) {
-      this.runs.delete(runId);
-    } else {
-      logger.error(`[api/app/clients/llm/RunManager] Run with ID ${runId} does not exist.`);
-    }
-  }
-
-  getAllRuns() {
-    return Array.from(this.runs.values());
-  }
-
-  getRunById(runId) {
-    return this.runs.get(runId);
-  }
-
-  getRunByConversationId(conversationId) {
-    const runId = this.convos.get(conversationId);
-    return { run: this.runs.get(runId), runId };
-  }
-
-  createCallbacks(metadata) {
-    return [
-      {
-        handleChatModelStart: createStartHandler({ ...metadata, manager: this }),
-        handleLLMEnd: async (output, runId, _parentRunId) => {
-          const { llmOutput, ..._output } = output;
-          logger.debug(`[RunManager] handleLLMEnd: ${JSON.stringify(metadata)}`, {
-            runId,
-            _parentRunId,
-            llmOutput,
-          });
-
-          if (metadata.context !== 'title') {
-            logger.debug('[RunManager] handleLLMEnd:', {
-              output: _output,
-            });
-          }
-
-          const { tokenUsage } = output.llmOutput;
-          const run = this.getRunById(runId);
-          this.removeRun(runId);
-
-          const txData = {
-            user: this.user,
-            model: run?.model ?? 'gpt-3.5-turbo',
-            ...metadata,
-          };
-
-          await spendTokens(txData, tokenUsage);
-        },
-        handleLLMError: async (err) => {
-          logger.error(`[RunManager] handleLLMError: ${JSON.stringify(metadata)}`, err);
-          if (metadata.context === 'title') {
-            return;
-          } else if (metadata.context === 'plugins') {
-            throw new Error(err);
-          }
-          const { conversationId } = metadata;
-          const { run } = this.getRunByConversationId(conversationId);
-          if (run && run.error) {
-            const { error } = run;
-            throw new Error(error);
-          }
-        },
-      },
-    ];
-  }
-}
-
-module.exports = RunManager;

api/app/clients/llm/createLLM.js

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

api/app/clients/llm/index.js

@@ -1,9 +1,7 @@
 const createLLM = require('./createLLM');
-const RunManager = require('./RunManager');
 const createCoherePayload = require('./createCoherePayload');
 
 module.exports = {
   createLLM,
-  RunManager,
   createCoherePayload,
 };

api/app/clients/memory/summaryBuffer.js

@@ -1,7 +1,7 @@
+const { logger } = require('@librechat/data-schemas');
 const { ConversationSummaryBufferMemory, ChatMessageHistory } = require('langchain/memory');
 const { formatLangChainMessages, SUMMARY_PROMPT } = require('../prompts');
 const { predictNewSummary } = require('../chains');
-const { logger } = require('~/config');
 
 const createSummaryBufferMemory = ({ llm, prompt, messages, ...rest }) => {
   const chatHistory = new ChatMessageHistory(messages);

api/app/clients/output_parsers/addImages.js

@@ -1,4 +1,4 @@
-const { logger } = require('~/config');
+const { logger } = require('@librechat/data-schemas');
 
 /**
  * The `addImages` function corrects any erroneous image URLs in the `responseMessage.text`

api/app/clients/prompts/createContextHandlers.js

@@ -1,6 +1,6 @@
 const axios = require('axios');
-const { isEnabled } = require('~/server/utils');
-const { logger } = require('~/config');
+const { logger } = require('@librechat/data-schemas');
+const { isEnabled, generateShortLivedToken } = require('@librechat/api');
 
 const footer = `Use the context as your learned knowledge to better answer the user.
 
@@ -18,7 +18,7 @@ function createContextHandlers(req, userMessageContent) {
   const queryPromises = [];
   const processedFiles = [];
   const processedIds = new Set();
-  const jwtToken = req.headers.authorization.split(' ')[1];
+  const jwtToken = generateShortLivedToken(req.user.id);
   const useFullContext = isEnabled(process.env.RAG_USE_FULL_CONTEXT);
 
   const query = async (file) => {
@@ -96,35 +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/prompts/formatMessages.js

@@ -237,41 +237,9 @@ const formatAgentMessages = (payload) => {
   return messages;
 };
 
-/**
- * Formats an array of messages for LangChain, making sure all content fields are strings
- * @param {Array<(HumanMessage|AIMessage|SystemMessage|ToolMessage)>} payload - The array of messages to format.
- * @returns {Array<(HumanMessage|AIMessage|SystemMessage|ToolMessage)>} - The array of formatted LangChain messages, including ToolMessages for tool calls.
- */
-const formatContentStrings = (payload) => {
-  const messages = [];
-
-  for (const message of payload) {
-    if (typeof message.content === 'string') {
-      continue;
-    }
-
-    if (!Array.isArray(message.content)) {
-      continue;
-    }
-
-    // Reduce text types to a single string, ignore all other types
-    const content = message.content.reduce((acc, curr) => {
-      if (curr.type === ContentTypes.TEXT) {
-        return `${acc}${curr[ContentTypes.TEXT]}\n`;
-      }
-      return acc;
-    }, '');
-
-    message.content = content.trim();
-  }
-
-  return messages;
-};
-
 module.exports = {
   formatMessage,
   formatFromLangChain,
   formatAgentMessages,
-  formatContentStrings,
   formatLangChainMessages,
 };

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

@@ -15,7 +15,7 @@ describe('AnthropicClient', () => {
     {
       role: 'user',
       isCreatedByUser: true,
-      text: 'What\'s up',
+      text: "What's up",
       messageId: '3',
       parentMessageId: '2',
     },
@@ -170,7 +170,7 @@ describe('AnthropicClient', () => {
       client.options.modelLabel = 'Claude-2';
       const result = await client.buildMessages(messages, parentMessageId);
       const { prompt } = result;
-      expect(prompt).toContain('Human\'s name: John');
+      expect(prompt).toContain("Human's name: John");
       expect(prompt).toContain('You are Claude-2');
     });
   });
@@ -244,6 +244,70 @@ describe('AnthropicClient', () => {
       );
     });
 
+    describe('Claude 4 model headers', () => {
+      it('should add "prompt-caching" and "context-1m" beta headers for claude-sonnet-4 model', () => {
+        const client = new AnthropicClient('test-api-key');
+        const modelOptions = {
+          model: 'claude-sonnet-4-20250514',
+        };
+        client.setOptions({ modelOptions, promptCache: true });
+        const anthropicClient = client.getClient(modelOptions);
+        expect(anthropicClient._options.defaultHeaders).toBeDefined();
+        expect(anthropicClient._options.defaultHeaders).toHaveProperty('anthropic-beta');
+        expect(anthropicClient._options.defaultHeaders['anthropic-beta']).toBe(
+          'prompt-caching-2024-07-31,context-1m-2025-08-07',
+        );
+      });
+
+      it('should add "prompt-caching" and "context-1m" beta headers for claude-sonnet-4 model formats', () => {
+        const client = new AnthropicClient('test-api-key');
+        const modelVariations = [
+          'claude-sonnet-4-20250514',
+          'claude-sonnet-4-latest',
+          'anthropic/claude-sonnet-4-20250514',
+        ];
+
+        modelVariations.forEach((model) => {
+          const modelOptions = { model };
+          client.setOptions({ modelOptions, promptCache: true });
+          const anthropicClient = client.getClient(modelOptions);
+          expect(anthropicClient._options.defaultHeaders).toBeDefined();
+          expect(anthropicClient._options.defaultHeaders).toHaveProperty('anthropic-beta');
+          expect(anthropicClient._options.defaultHeaders['anthropic-beta']).toBe(
+            'prompt-caching-2024-07-31,context-1m-2025-08-07',
+          );
+        });
+      });
+
+      it('should add "prompt-caching" beta header for claude-opus-4 model', () => {
+        const client = new AnthropicClient('test-api-key');
+        const modelOptions = {
+          model: 'claude-opus-4-20250514',
+        };
+        client.setOptions({ modelOptions, promptCache: true });
+        const anthropicClient = client.getClient(modelOptions);
+        expect(anthropicClient._options.defaultHeaders).toBeDefined();
+        expect(anthropicClient._options.defaultHeaders).toHaveProperty('anthropic-beta');
+        expect(anthropicClient._options.defaultHeaders['anthropic-beta']).toBe(
+          'prompt-caching-2024-07-31',
+        );
+      });
+
+      it('should add "prompt-caching" beta header for claude-4-opus model', () => {
+        const client = new AnthropicClient('test-api-key');
+        const modelOptions = {
+          model: 'claude-4-opus-20250514',
+        };
+        client.setOptions({ modelOptions, promptCache: true });
+        const anthropicClient = client.getClient(modelOptions);
+        expect(anthropicClient._options.defaultHeaders).toBeDefined();
+        expect(anthropicClient._options.defaultHeaders).toHaveProperty('anthropic-beta');
+        expect(anthropicClient._options.defaultHeaders['anthropic-beta']).toBe(
+          'prompt-caching-2024-07-31',
+        );
+      });
+    });
+
     it('should not add beta header for claude-3-5-sonnet-latest model', () => {
       const client = new AnthropicClient('test-api-key');
       const modelOptions = {
@@ -251,7 +315,7 @@ describe('AnthropicClient', () => {
       };
       client.setOptions({ modelOptions, promptCache: true });
       const anthropicClient = client.getClient(modelOptions);
-      expect(anthropicClient.defaultHeaders).not.toHaveProperty('anthropic-beta');
+      expect(anthropicClient._options.defaultHeaders).toBeUndefined();
     });
 
     it('should not add beta header for other models', () => {
@@ -262,7 +326,7 @@ describe('AnthropicClient', () => {
         },
       });
       const anthropicClient = client.getClient();
-      expect(anthropicClient.defaultHeaders).not.toHaveProperty('anthropic-beta');
+      expect(anthropicClient._options.defaultHeaders).toBeUndefined();
     });
   });
 
@@ -456,6 +520,34 @@ describe('AnthropicClient', () => {
       expect(client.modelOptions.maxOutputTokens).toBe(highTokenValue);
     });
 
+    it('should not cap maxOutputTokens for Claude 4 Sonnet models', () => {
+      const client = new AnthropicClient('test-api-key');
+      const highTokenValue = anthropicSettings.legacy.maxOutputTokens.default * 10; // 40,960 tokens
+
+      client.setOptions({
+        modelOptions: {
+          model: 'claude-sonnet-4-20250514',
+          maxOutputTokens: highTokenValue,
+        },
+      });
+
+      expect(client.modelOptions.maxOutputTokens).toBe(highTokenValue);
+    });
+
+    it('should not cap maxOutputTokens for Claude 4 Opus models', () => {
+      const client = new AnthropicClient('test-api-key');
+      const highTokenValue = anthropicSettings.legacy.maxOutputTokens.default * 6; // 24,576 tokens (under 32K limit)
+
+      client.setOptions({
+        modelOptions: {
+          model: 'claude-opus-4-20250514',
+          maxOutputTokens: highTokenValue,
+        },
+      });
+
+      expect(client.modelOptions.maxOutputTokens).toBe(highTokenValue);
+    });
+
     it('should cap maxOutputTokens for Claude 3.5 Haiku models', () => {
       const client = new AnthropicClient('test-api-key');
       const highTokenValue = anthropicSettings.legacy.maxOutputTokens.default * 2;
@@ -729,4 +821,223 @@ describe('AnthropicClient', () => {
     expect(capturedOptions).toHaveProperty('topK', 10);
     expect(capturedOptions).toHaveProperty('topP', 0.9);
   });
+
+  describe('isClaudeLatest', () => {
+    it('should set isClaudeLatest to true for claude-3 models', () => {
+      const client = new AnthropicClient('test-api-key');
+      client.setOptions({
+        modelOptions: {
+          model: 'claude-3-sonnet-20240229',
+        },
+      });
+      expect(client.isClaudeLatest).toBe(true);
+    });
+
+    it('should set isClaudeLatest to true for claude-3.5 models', () => {
+      const client = new AnthropicClient('test-api-key');
+      client.setOptions({
+        modelOptions: {
+          model: 'claude-3.5-sonnet-20240229',
+        },
+      });
+      expect(client.isClaudeLatest).toBe(true);
+    });
+
+    it('should set isClaudeLatest to true for claude-sonnet-4 models', () => {
+      const client = new AnthropicClient('test-api-key');
+      client.setOptions({
+        modelOptions: {
+          model: 'claude-sonnet-4-20240229',
+        },
+      });
+      expect(client.isClaudeLatest).toBe(true);
+    });
+
+    it('should set isClaudeLatest to true for claude-opus-4 models', () => {
+      const client = new AnthropicClient('test-api-key');
+      client.setOptions({
+        modelOptions: {
+          model: 'claude-opus-4-20240229',
+        },
+      });
+      expect(client.isClaudeLatest).toBe(true);
+    });
+
+    it('should set isClaudeLatest to true for claude-3.5-haiku models', () => {
+      const client = new AnthropicClient('test-api-key');
+      client.setOptions({
+        modelOptions: {
+          model: 'claude-3.5-haiku-20240229',
+        },
+      });
+      expect(client.isClaudeLatest).toBe(true);
+    });
+
+    it('should set isClaudeLatest to false for claude-2 models', () => {
+      const client = new AnthropicClient('test-api-key');
+      client.setOptions({
+        modelOptions: {
+          model: 'claude-2',
+        },
+      });
+      expect(client.isClaudeLatest).toBe(false);
+    });
+
+    it('should set isClaudeLatest to false for claude-instant models', () => {
+      const client = new AnthropicClient('test-api-key');
+      client.setOptions({
+        modelOptions: {
+          model: 'claude-instant',
+        },
+      });
+      expect(client.isClaudeLatest).toBe(false);
+    });
+
+    it('should set isClaudeLatest to false for claude-sonnet-3 models', () => {
+      const client = new AnthropicClient('test-api-key');
+      client.setOptions({
+        modelOptions: {
+          model: 'claude-sonnet-3-20240229',
+        },
+      });
+      expect(client.isClaudeLatest).toBe(false);
+    });
+
+    it('should set isClaudeLatest to false for claude-opus-3 models', () => {
+      const client = new AnthropicClient('test-api-key');
+      client.setOptions({
+        modelOptions: {
+          model: 'claude-opus-3-20240229',
+        },
+      });
+      expect(client.isClaudeLatest).toBe(false);
+    });
+
+    it('should set isClaudeLatest to false for claude-haiku-3 models', () => {
+      const client = new AnthropicClient('test-api-key');
+      client.setOptions({
+        modelOptions: {
+          model: 'claude-haiku-3-20240229',
+        },
+      });
+      expect(client.isClaudeLatest).toBe(false);
+    });
+  });
+
+  describe('configureReasoning', () => {
+    it('should enable thinking for claude-opus-4 and claude-sonnet-4 models', async () => {
+      const client = new AnthropicClient('test-api-key');
+      // Create a mock async generator function
+      async function* mockAsyncGenerator() {
+        yield { type: 'message_start', message: { usage: {} } };
+        yield { delta: { text: 'Test response' } };
+        yield { type: 'message_delta', usage: {} };
+      }
+
+      // Mock createResponse to return the async generator
+      jest.spyOn(client, 'createResponse').mockImplementation(() => {
+        return mockAsyncGenerator();
+      });
+
+      // Test claude-opus-4
+      client.setOptions({
+        modelOptions: {
+          model: 'claude-opus-4-20250514',
+        },
+        thinking: true,
+        thinkingBudget: 2000,
+      });
+
+      let capturedOptions = null;
+      jest.spyOn(client, 'getClient').mockImplementation((options) => {
+        capturedOptions = options;
+        return {};
+      });
+
+      const payload = [{ role: 'user', content: 'Test message' }];
+      await client.sendCompletion(payload, {});
+
+      expect(capturedOptions).toHaveProperty('thinking');
+      expect(capturedOptions.thinking).toEqual({
+        type: 'enabled',
+        budget_tokens: 2000,
+      });
+
+      // Test claude-sonnet-4
+      client.setOptions({
+        modelOptions: {
+          model: 'claude-sonnet-4-20250514',
+        },
+        thinking: true,
+        thinkingBudget: 2000,
+      });
+
+      await client.sendCompletion(payload, {});
+
+      expect(capturedOptions).toHaveProperty('thinking');
+      expect(capturedOptions.thinking).toEqual({
+        type: 'enabled',
+        budget_tokens: 2000,
+      });
+    });
+  });
+});
+
+describe('Claude Model Tests', () => {
+  it('should handle Claude 3 and 4 series models correctly', () => {
+    const client = new AnthropicClient('test-key');
+    // Claude 3 series models
+    const claude3Models = [
+      'claude-3-opus-20240229',
+      'claude-3-sonnet-20240229',
+      'claude-3-haiku-20240307',
+      'claude-3-5-sonnet-20240620',
+      'claude-3-5-haiku-20240620',
+      'claude-3.5-sonnet-20240620',
+      'claude-3.5-haiku-20240620',
+      'claude-3.7-sonnet-20240620',
+      'claude-3.7-haiku-20240620',
+      'anthropic/claude-3-opus-20240229',
+      'claude-3-opus-20240229/anthropic',
+    ];
+
+    // Claude 4 series models
+    const claude4Models = [
+      'claude-sonnet-4-20250514',
+      'claude-opus-4-20250514',
+      'claude-4-sonnet-20250514',
+      'claude-4-opus-20250514',
+      'anthropic/claude-sonnet-4-20250514',
+      'claude-sonnet-4-20250514/anthropic',
+    ];
+
+    // Test Claude 3 series
+    claude3Models.forEach((model) => {
+      client.setOptions({ modelOptions: { model } });
+      expect(
+        /claude-[3-9]/.test(client.modelOptions.model) ||
+          /claude-(?:sonnet|opus|haiku)-[4-9]/.test(client.modelOptions.model),
+      ).toBe(true);
+    });
+
+    // Test Claude 4 series
+    claude4Models.forEach((model) => {
+      client.setOptions({ modelOptions: { model } });
+      expect(
+        /claude-[3-9]/.test(client.modelOptions.model) ||
+          /claude-(?:sonnet|opus|haiku)-[4-9]/.test(client.modelOptions.model),
+      ).toBe(true);
+    });
+
+    // Test non-Claude 3/4 models
+    const nonClaudeModels = ['claude-2', 'claude-instant', 'gpt-4', 'gpt-3.5-turbo'];
+
+    nonClaudeModels.forEach((model) => {
+      client.setOptions({ modelOptions: { model } });
+      expect(
+        /claude-[3-9]/.test(client.modelOptions.model) ||
+          /claude-(?:sonnet|opus|haiku)-[4-9]/.test(client.modelOptions.model),
+      ).toBe(false);
+    });
+  });
 });

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

@@ -1,7 +1,15 @@
 const { Constants } = require('librechat-data-provider');
 const { initializeFakeClient } = require('./FakeClient');
 
-jest.mock('~/lib/db/connectDb');
+jest.mock('~/db/connect');
+jest.mock('~/server/services/Config', () => ({
+  getAppConfig: jest.fn().mockResolvedValue({
+    // Default app config for tests
+    paths: { uploads: '/tmp' },
+    fileStrategy: 'local',
+    memory: { disabled: false },
+  }),
+}));
 jest.mock('~/models', () => ({
   User: jest.fn(),
   Key: jest.fn(),
@@ -33,7 +41,9 @@ 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 {};
     }),
@@ -52,7 +62,7 @@ const messageHistory = [
   {
     role: 'user',
     isCreatedByUser: true,
-    text: 'What\'s up',
+    text: "What's up",
     messageId: '3',
     parentMessageId: '2',
   },
@@ -420,6 +430,46 @@ describe('BaseClient', () => {
       expect(response).toEqual(expectedResult);
     });
 
+    test('should replace responseMessageId with new UUID when isRegenerate is true and messageId ends with underscore', async () => {
+      const mockCrypto = require('crypto');
+      const newUUID = 'new-uuid-1234';
+      jest.spyOn(mockCrypto, 'randomUUID').mockReturnValue(newUUID);
+
+      const opts = {
+        isRegenerate: true,
+        responseMessageId: 'existing-message-id_',
+      };
+
+      await TestClient.setMessageOptions(opts);
+
+      expect(TestClient.responseMessageId).toBe(newUUID);
+      expect(TestClient.responseMessageId).not.toBe('existing-message-id_');
+
+      mockCrypto.randomUUID.mockRestore();
+    });
+
+    test('should not replace responseMessageId when isRegenerate is false', async () => {
+      const opts = {
+        isRegenerate: false,
+        responseMessageId: 'existing-message-id_',
+      };
+
+      await TestClient.setMessageOptions(opts);
+
+      expect(TestClient.responseMessageId).toBe('existing-message-id_');
+    });
+
+    test('should not replace responseMessageId when it does not end with underscore', async () => {
+      const opts = {
+        isRegenerate: true,
+        responseMessageId: 'existing-message-id',
+      };
+
+      await TestClient.setMessageOptions(opts);
+
+      expect(TestClient.responseMessageId).toBe('existing-message-id');
+    });
+
     test('sendMessage should work with provided conversationId and parentMessageId', async () => {
       const userMessage = 'Second message in the conversation';
       const opts = {
@@ -456,7 +506,7 @@ describe('BaseClient', () => {
 
       const chatMessages2 = await TestClient.loadHistory(conversationId, '3');
       expect(TestClient.currentMessages).toHaveLength(3);
-      expect(chatMessages2[chatMessages2.length - 1].text).toEqual('What\'s up');
+      expect(chatMessages2[chatMessages2.length - 1].text).toEqual("What's up");
     });
 
     /* Most of the new sendMessage logic revolving around edited/continued AI messages
@@ -537,6 +587,8 @@ describe('BaseClient', () => {
       expect(onStart).toHaveBeenCalledWith(
         expect.objectContaining({ text: 'Hello, world!' }),
         expect.any(String),
+        /** `isNewConvo` */
+        true,
       );
     });
 

api/app/clients/specs/FakeClient.js

@@ -1,5 +1,5 @@
+const { getModelMaxTokens } = require('@librechat/api');
 const BaseClient = require('../BaseClient');
-const { getModelMaxTokens } = require('../../../utils');
 
 class FakeClient extends BaseClient {
   constructor(apiKey, options = {}) {

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

@@ -5,7 +5,7 @@ const getLogStores = require('~/cache/getLogStores');
 const OpenAIClient = require('../OpenAIClient');
 jest.mock('meilisearch');
 
-jest.mock('~/lib/db/connectDb');
+jest.mock('~/db/connect');
 jest.mock('~/models', () => ({
   User: jest.fn(),
   Key: jest.fn(),
@@ -462,17 +462,17 @@ describe('OpenAIClient', () => {
         role: 'system',
         name: 'example_user',
         content:
-          'Let\'s circle back when we have more bandwidth to touch base on opportunities for increased leverage.',
+          "Let's circle back when we have more bandwidth to touch base on opportunities for increased leverage.",
       },
       {
         role: 'system',
         name: 'example_assistant',
-        content: 'Let\'s talk later when we\'re less busy about how to do better.',
+        content: "Let's talk later when we're less busy about how to do better.",
       },
       {
         role: 'user',
         content:
-          'This late pivot means we don\'t have time to boil the ocean for the client deliverable.',
+          "This late pivot means we don't have time to boil the ocean for the client deliverable.",
       },
     ];
 
@@ -531,44 +531,6 @@ describe('OpenAIClient', () => {
     });
   });
 
-  describe('sendMessage/getCompletion/chatCompletion', () => {
-    afterEach(() => {
-      delete process.env.AZURE_OPENAI_DEFAULT_MODEL;
-      delete process.env.AZURE_USE_MODEL_AS_DEPLOYMENT_NAME;
-    });
-
-    it('should call getCompletion and fetchEventSource when using a text/instruct model', async () => {
-      const model = 'text-davinci-003';
-      const onProgress = jest.fn().mockImplementation(() => ({}));
-
-      const testClient = new OpenAIClient('test-api-key', {
-        ...defaultOptions,
-        modelOptions: { model },
-      });
-
-      const getCompletion = jest.spyOn(testClient, 'getCompletion');
-      await testClient.sendMessage('Hi mom!', { onProgress });
-
-      expect(getCompletion).toHaveBeenCalled();
-      expect(getCompletion.mock.calls.length).toBe(1);
-
-      expect(getCompletion.mock.calls[0][0]).toBe('||>User:\nHi mom!\n||>Assistant:\n');
-
-      expect(fetchEventSource).toHaveBeenCalled();
-      expect(fetchEventSource.mock.calls.length).toBe(1);
-
-      // Check if the first argument (url) is correct
-      const firstCallArgs = fetchEventSource.mock.calls[0];
-
-      const expectedURL = 'https://api.openai.com/v1/completions';
-      expect(firstCallArgs[0]).toBe(expectedURL);
-
-      const requestBody = JSON.parse(firstCallArgs[1].body);
-      expect(requestBody).toHaveProperty('model');
-      expect(requestBody.model).toBe(model);
-    });
-  });
-
   describe('checkVisionRequest functionality', () => {
     let client;
     const attachments = [{ type: 'image/png' }];

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

@@ -1,314 +0,0 @@
-const crypto = require('crypto');
-const { Constants } = require('librechat-data-provider');
-const { HumanMessage, AIMessage } = require('@langchain/core/messages');
-const PluginsClient = require('../PluginsClient');
-
-jest.mock('~/lib/db/connectDb');
-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

@@ -1,184 +0,0 @@
-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

@@ -1,72 +0,0 @@
-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/index.js

@@ -1,4 +1,4 @@
-const availableTools = require('./manifest.json');
+const manifest = require('./manifest');
 
 // Structured Tools
 const DALLE3 = require('./structured/DALLE3');
@@ -13,23 +13,8 @@ const TraversaalSearch = require('./structured/TraversaalSearch');
 const createOpenAIImageTools = require('./structured/OpenAIImageTools');
 const TavilySearchResults = require('./structured/TavilySearchResults');
 
-/** @type {Record<string, TPlugin | undefined>} */
-const manifestToolMap = {};
-
-/** @type {Array<TPlugin>} */
-const toolkits = [];
-
-availableTools.forEach((tool) => {
-  manifestToolMap[tool.pluginKey] = tool;
-  if (tool.toolkit === true) {
-    toolkits.push(tool);
-  }
-});
-
 module.exports = {
-  toolkits,
-  availableTools,
-  manifestToolMap,
+  ...manifest,
   // Structured Tools
   DALLE3,
   FluxAPI,

api/app/clients/tools/manifest.js

@@ -0,0 +1,20 @@
+const availableTools = require('./manifest.json');
+
+/** @type {Record<string, TPlugin | undefined>} */
+const manifestToolMap = {};
+
+/** @type {Array<TPlugin>} */
+const toolkits = [];
+
+availableTools.forEach((tool) => {
+  manifestToolMap[tool.pluginKey] = tool;
+  if (tool.toolkit === true) {
+    toolkits.push(tool);
+  }
+});
+
+module.exports = {
+  toolkits,
+  availableTools,
+  manifestToolMap,
+};

api/app/clients/tools/manifest.json

@@ -49,7 +49,7 @@
     "pluginKey": "image_gen_oai",
     "toolkit": true,
     "description": "Image Generation and Editing using OpenAI's latest state-of-the-art models",
-    "icon": "/assets/image_gen_oai.png",
+    "icon": "assets/image_gen_oai.png",
     "authConfig": [
       {
         "authField": "IMAGE_GEN_OAI_API_KEY",
@@ -75,7 +75,7 @@
     "name": "Browser",
     "pluginKey": "web-browser",
     "description": "Scrape and summarize webpage data",
-    "icon": "/assets/web-browser.svg",
+    "icon": "assets/web-browser.svg",
     "authConfig": [
       {
         "authField": "OPENAI_API_KEY",
@@ -170,7 +170,7 @@
     "name": "OpenWeather",
     "pluginKey": "open_weather",
     "description": "Get weather forecasts and historical data from the OpenWeather API",
-    "icon": "/assets/openweather.png",
+    "icon": "assets/openweather.png",
     "authConfig": [
       {
         "authField": "OPENWEATHER_API_KEY",

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

@@ -1,7 +1,7 @@
 const { z } = require('zod');
 const { Tool } = require('@langchain/core/tools');
+const { logger } = require('@librechat/data-schemas');
 const { SearchClient, AzureKeyCredential } = require('@azure/search-documents');
-const { logger } = require('~/config');
 
 class AzureAISearch extends Tool {
   // Constants for default values
@@ -18,7 +18,7 @@ class AzureAISearch extends Tool {
     super();
     this.name = 'azure-ai-search';
     this.description =
-      'Use the \'azure-ai-search\' tool to retrieve search results relevant to your input';
+      "Use the 'azure-ai-search' tool to retrieve search results relevant to your input";
     /* Used to initialize the Tool without necessary variables. */
     this.override = fields.override ?? false;
 

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

@@ -3,15 +3,15 @@ const path = require('path');
 const OpenAI = require('openai');
 const fetch = require('node-fetch');
 const { v4: uuidv4 } = require('uuid');
+const { ProxyAgent } = require('undici');
 const { Tool } = require('@langchain/core/tools');
-const { HttpsProxyAgent } = require('https-proxy-agent');
+const { logger } = require('@librechat/data-schemas');
+const { getImageBasename } = require('@librechat/api');
 const { FileContext, ContentTypes } = require('librechat-data-provider');
-const { getImageBasename } = require('~/server/services/Files/images');
 const extractBaseURL = require('~/utils/extractBaseURL');
-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();
@@ -46,7 +46,10 @@ class DALLE3 extends Tool {
     }
 
     if (process.env.PROXY) {
-      config.httpAgent = new HttpsProxyAgent(process.env.PROXY);
+      const proxyAgent = new ProxyAgent(process.env.PROXY);
+      config.fetchOptions = {
+        dispatcher: proxyAgent,
+      };
     }
 
     /** @type {OpenAI} */
@@ -163,7 +166,8 @@ Error Message: ${error.message}`);
     if (this.isAgent) {
       let fetchOptions = {};
       if (process.env.PROXY) {
-        fetchOptions.agent = new HttpsProxyAgent(process.env.PROXY);
+        const proxyAgent = new ProxyAgent(process.env.PROXY);
+        fetchOptions.dispatcher = proxyAgent;
       }
       const imageResponse = await fetch(theImageUrl, fetchOptions);
       const arrayBuffer = await imageResponse.arrayBuffer();

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

@@ -3,12 +3,12 @@ const axios = require('axios');
 const fetch = require('node-fetch');
 const { v4: uuidv4 } = require('uuid');
 const { Tool } = require('@langchain/core/tools');
+const { logger } = require('@librechat/data-schemas');
 const { HttpsProxyAgent } = require('https-proxy-agent');
 const { FileContext, ContentTypes } = require('librechat-data-provider');
-const { logger } = require('~/config');
 
 const displayMessage =
-  'Flux 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.';
+  "Flux 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.";
 
 /**
  * FluxAPI - A tool for generating high-quality images from text prompts using the Flux API.

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

@@ -1,70 +1,18 @@
-const { z } = require('zod');
 const axios = require('axios');
 const { v4 } = require('uuid');
 const OpenAI = require('openai');
 const FormData = require('form-data');
+const { ProxyAgent } = require('undici');
 const { tool } = require('@langchain/core/tools');
-const { HttpsProxyAgent } = require('https-proxy-agent');
+const { logger } = require('@librechat/data-schemas');
+const { logAxiosError, oaiToolkit } = require('@librechat/api');
 const { ContentTypes, EImageOutputType } = require('librechat-data-provider');
 const { getStrategyFunctions } = require('~/server/services/Files/strategies');
-const { logAxiosError, extractBaseURL } = require('~/utils');
+const extractBaseURL = require('~/utils/extractBaseURL');
 const { getFiles } = require('~/models/File');
-const { logger } = require('~/config');
-
-/** Default descriptions for image generation tool  */
-const DEFAULT_IMAGE_GEN_DESCRIPTION = `
-Generates high-quality, original images based solely on text, not using any uploaded reference images.
-
-When to use \`image_gen_oai\`:
-- To create entirely new images from detailed text descriptions that do NOT reference any image files.
-
-When NOT to use \`image_gen_oai\`:
-- If the user has uploaded any images and requests modifications, enhancements, or remixing based on those uploads → use \`image_edit_oai\` instead.
-
-Generated image IDs will be returned in the response, so you can refer to them in future requests made to \`image_edit_oai\`.
-`.trim();
-
-/** Default description for image editing tool  */
-const DEFAULT_IMAGE_EDIT_DESCRIPTION =
-  `Generates high-quality, original images based on text and one or more uploaded/referenced images.
-
-When to use \`image_edit_oai\`:
-- The user wants to modify, extend, or remix one **or more** uploaded images, either:
-  - Previously generated, or in the current request (both to be included in the \`image_ids\` array).
-- Always when the user refers to uploaded images for editing, enhancement, remixing, style transfer, or combining elements.
-- Any current or existing images are to be used as visual guides.
-- If there are any files in the current request, they are more likely than not expected as references for image edit requests.
-
-When NOT to use \`image_edit_oai\`:
-- Brand-new generations that do not rely on an existing image → use \`image_gen_oai\` instead.
-
-Both generated and referenced image IDs will be returned in the response, so you can refer to them in future requests made to \`image_edit_oai\`.
-`.trim();
-
-/** Default prompt descriptions  */
-const DEFAULT_IMAGE_GEN_PROMPT_DESCRIPTION = `Describe the image you want in detail. 
-      Be highly specific—break your idea into layers: 
-      (1) main concept and subject,
-      (2) composition and position,
-      (3) lighting and mood,
-      (4) style, medium, or camera details,
-      (5) important features (age, expression, clothing, etc.),
-      (6) background.
-      Use positive, descriptive language and specify what should be included, not what to avoid. 
-      List number and characteristics of people/objects, and mention style/technical requirements (e.g., "DSLR photo, 85mm lens, golden hour").
-      Do not reference any uploaded images—use for new image creation from text only.`;
-
-const DEFAULT_IMAGE_EDIT_PROMPT_DESCRIPTION = `Describe the changes, enhancements, or new ideas to apply to the uploaded image(s).
-      Be highly specific—break your request into layers: 
-      (1) main concept or transformation,
-      (2) specific edits/replacements or composition guidance,
-      (3) desired style, mood, or technique,
-      (4) features/items to keep, change, or add (such as objects, people, clothing, lighting, etc.).
-      Use positive, descriptive language and clarify what should be included or changed, not what to avoid.
-      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
@@ -90,21 +38,11 @@ function returnValue(value) {
   return value;
 }
 
-const getImageGenDescription = () => {
-  return process.env.IMAGE_GEN_OAI_DESCRIPTION || DEFAULT_IMAGE_GEN_DESCRIPTION;
-};
-
-const getImageEditDescription = () => {
-  return process.env.IMAGE_EDIT_OAI_DESCRIPTION || DEFAULT_IMAGE_EDIT_DESCRIPTION;
-};
-
-const getImageGenPromptDescription = () => {
-  return process.env.IMAGE_GEN_OAI_PROMPT_DESCRIPTION || DEFAULT_IMAGE_GEN_PROMPT_DESCRIPTION;
-};
-
-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)
@@ -114,7 +52,9 @@ const getImageEditPromptDescription = () => {
  * @param {string} fields.IMAGE_GEN_OAI_API_KEY - The OpenAI API key
  * @param {boolean} [fields.override] - Whether to override the API key check, necessary for app initialization
  * @param {MongoFile[]} [fields.imageFiles] - The images to be used for editing
- * @returns {Array} - Array of image tools
+ * @param {string} [fields.imageOutputType] - The image output type configuration
+ * @param {string} [fields.fileStrategy] - The file storage strategy
+ * @returns {Array<ReturnType<tool>>} - Array of image tools
  */
 function createOpenAIImageTools(fields = {}) {
   /** @type {boolean} Used to initialize the Tool without necessary variables. */
@@ -124,8 +64,8 @@ function createOpenAIImageTools(fields = {}) {
     throw new Error('This tool is only available for agents.');
   }
   const { req } = fields;
-  const imageOutputType = req?.app.locals.imageOutputType || EImageOutputType.PNG;
-  const appFileStrategy = req?.app.locals.fileStrategy;
+  const imageOutputType = fields.imageOutputType || EImageOutputType.PNG;
+  const appFileStrategy = fields.fileStrategy;
 
   const getApiKey = () => {
     const apiKey = process.env.IMAGE_GEN_OAI_API_KEY ?? '';
@@ -182,7 +122,10 @@ function createOpenAIImageTools(fields = {}) {
       }
       const clientConfig = { ...closureConfig };
       if (process.env.PROXY) {
-        clientConfig.httpAgent = new HttpsProxyAgent(process.env.PROXY);
+        const proxyAgent = new ProxyAgent(process.env.PROXY);
+        clientConfig.fetchOptions = {
+          dispatcher: proxyAgent,
+        };
       }
 
       /** @type {OpenAI} */
@@ -200,10 +143,18 @@ function createOpenAIImageTools(fields = {}) {
       }
 
       let resp;
+      /** @type {AbortSignal} */
+      let derivedSignal = null;
+      /** @type {() => void} */
+      let abortHandler = null;
+
       try {
-        const derivedSignal = runnableConfig?.signal
-          ? AbortSignal.any([runnableConfig.signal])
-          : undefined;
+        if (runnableConfig?.signal) {
+          derivedSignal = AbortSignal.any([runnableConfig.signal]);
+          abortHandler = createAbortHandler();
+          derivedSignal.addEventListener('abort', abortHandler, { once: true });
+        }
+
         resp = await openai.images.generate(
           {
             model: 'gpt-image-1',
@@ -227,6 +178,10 @@ function createOpenAIImageTools(fields = {}) {
         logAxiosError({ error, message });
         return returnValue(`Something went wrong when trying to generate the image. The OpenAI API may be unavailable:
 Error Message: ${error.message}`);
+      } finally {
+        if (abortHandler && derivedSignal) {
+          derivedSignal.removeEventListener('abort', abortHandler);
+        }
       }
 
       if (!resp) {
@@ -263,46 +218,7 @@ Error Message: ${error.message}`);
       ];
       return [response, { content, file_ids }];
     },
-    {
-      name: 'image_gen_oai',
-      description: getImageGenDescription(),
-      schema: z.object({
-        prompt: z.string().max(32000).describe(getImageGenPromptDescription()),
-        background: z
-          .enum(['transparent', 'opaque', 'auto'])
-          .optional()
-          .describe(
-            'Sets transparency for the background. Must be one of transparent, opaque or auto (default). When transparent, the output format should be png or webp.',
-          ),
-        /*
-        n: z
-          .number()
-          .int()
-          .min(1)
-          .max(10)
-          .optional()
-          .describe('The number of images to generate. Must be between 1 and 10.'),
-        output_compression: z
-          .number()
-          .int()
-          .min(0)
-          .max(100)
-          .optional()
-          .describe('The compression level (0-100%) for webp or jpeg formats. Defaults to 100.'),
-           */
-        quality: z
-          .enum(['auto', 'high', 'medium', 'low'])
-          .optional()
-          .describe('The quality of the image. One of auto (default), high, medium, or low.'),
-        size: z
-          .enum(['auto', '1024x1024', '1536x1024', '1024x1536'])
-          .optional()
-          .describe(
-            'The size of the generated image. One of 1024x1024, 1536x1024 (landscape), 1024x1536 (portrait), or auto (default).',
-          ),
-      }),
-      responseFormat: 'content_and_artifact',
-    },
+    oaiToolkit.image_gen_oai,
   );
 
   /**
@@ -316,7 +232,10 @@ Error Message: ${error.message}`);
 
       const clientConfig = { ...closureConfig };
       if (process.env.PROXY) {
-        clientConfig.httpAgent = new HttpsProxyAgent(process.env.PROXY);
+        const proxyAgent = new ProxyAgent(process.env.PROXY);
+        clientConfig.fetchOptions = {
+          dispatcher: proxyAgent,
+        };
       }
 
       const formData = new FormData();
@@ -408,10 +327,17 @@ Error Message: ${error.message}`);
         headers['Authorization'] = `Bearer ${apiKey}`;
       }
 
+      /** @type {AbortSignal} */
+      let derivedSignal = null;
+      /** @type {() => void} */
+      let abortHandler = null;
+
       try {
-        const derivedSignal = runnableConfig?.signal
-          ? AbortSignal.any([runnableConfig.signal])
-          : undefined;
+        if (runnableConfig?.signal) {
+          derivedSignal = AbortSignal.any([runnableConfig.signal]);
+          abortHandler = createAbortHandler();
+          derivedSignal.addEventListener('abort', abortHandler, { once: true });
+        }
 
         /** @type {import('axios').AxiosRequestConfig} */
         const axiosConfig = {
@@ -421,6 +347,19 @@ Error Message: ${error.message}`);
           baseURL,
         };
 
+        if (process.env.PROXY) {
+          try {
+            const url = new URL(process.env.PROXY);
+            axiosConfig.proxy = {
+              host: url.hostname.replace(/^\[|\]$/g, ''),
+              port: url.port ? parseInt(url.port, 10) : undefined,
+              protocol: url.protocol.replace(':', ''),
+            };
+          } catch (error) {
+            logger.error('Error parsing proxy URL:', error);
+          }
+        }
+
         if (process.env.IMAGE_GEN_OAI_AZURE_API_VERSION && process.env.IMAGE_GEN_OAI_BASEURL) {
           axiosConfig.params = {
             'api-version': process.env.IMAGE_GEN_OAI_AZURE_API_VERSION,
@@ -466,50 +405,13 @@ 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);
+        }
       }
     },
-    {
-      name: 'image_edit_oai',
-      description: getImageEditDescription(),
-      schema: z.object({
-        image_ids: z
-          .array(z.string())
-          .min(1)
-          .describe(
-            `
-IDs (image ID strings) of previously generated or uploaded images that should guide the edit.
-
-Guidelines:
-- If the user's request depends on any prior image(s), copy their image IDs into the \`image_ids\` array (in the same order the user refers to them).  
-- Never invent or hallucinate IDs; only use IDs that are still visible in the conversation context.
-- If no earlier image is relevant, omit the field entirely.
-`.trim(),
-          ),
-        prompt: z.string().max(32000).describe(getImageEditPromptDescription()),
-        /*
-        n: z
-          .number()
-          .int()
-          .min(1)
-          .max(10)
-          .optional()
-          .describe('The number of images to generate. Must be between 1 and 10. Defaults to 1.'),
-        */
-        quality: z
-          .enum(['auto', 'high', 'medium', 'low'])
-          .optional()
-          .describe(
-            'The quality of the image. One of auto (default), high, medium, or low. High/medium/low only supported for gpt-image-1.',
-          ),
-        size: z
-          .enum(['auto', '1024x1024', '1536x1024', '1024x1536', '256x256', '512x512'])
-          .optional()
-          .describe(
-            'The size of the generated images. For gpt-image-1: auto (default), 1024x1024, 1536x1024, 1024x1536. For dall-e-2: 256x256, 512x512, 1024x1024.',
-          ),
-      }),
-      responseFormat: 'content_and_artifact',
-    },
+    oaiToolkit.image_edit_oai,
   );
 
   return [imageGenTool, imageEditTool];

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

@@ -6,19 +6,19 @@ const axios = require('axios');
 const sharp = require('sharp');
 const { v4: uuidv4 } = require('uuid');
 const { Tool } = require('@langchain/core/tools');
+const { logger } = require('@librechat/data-schemas');
 const { FileContext, ContentTypes } = require('librechat-data-provider');
 const paths = require('~/config/paths');
-const { logger } = require('~/config');
 
 const displayMessage =
-  'Stable Diffusion 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.';
+  "Stable Diffusion 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 StableDiffusionAPI extends Tool {
   constructor(fields) {
     super();
     /** @type {string} User ID */
     this.userId = fields.userId;
-    /** @type {Express.Request | undefined} Express Request object, only provided by ToolService */
+    /** @type {ServerRequest | undefined} Express Request object, only provided by ToolService */
     this.req = fields.req;
     /** @type {boolean} Used to initialize the Tool without necessary variables. */
     this.override = fields.override ?? false;
@@ -44,7 +44,7 @@ class StableDiffusionAPI extends Tool {
 // "negative_prompt":"semi-realistic, cgi, 3d, render, sketch, cartoon, drawing, anime, out of frame, low quality, ugly, mutation, deformed"
 // - Generate images only once per human query unless explicitly requested by the user`;
     this.description =
-      'You can generate images using text with \'stable-diffusion\'. This tool is exclusively for visual content.';
+      "You can generate images using text with 'stable-diffusion'. This tool is exclusively for visual content.";
     this.schema = z.object({
       prompt: z
         .string()

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

@@ -43,9 +43,39 @@ class TavilySearchResults extends Tool {
         .boolean()
         .optional()
         .describe('Whether to include answers in the search results. Default is False.'),
-      // include_raw_content: z.boolean().optional().describe('Whether to include raw content in the search results. Default is False.'),
-      // include_domains: z.array(z.string()).optional().describe('A list of domains to specifically include in the search results.'),
-      // exclude_domains: z.array(z.string()).optional().describe('A list of domains to specifically exclude from the search results.'),
+      include_raw_content: z
+        .boolean()
+        .optional()
+        .describe('Whether to include raw content in the search results. Default is False.'),
+      include_domains: z
+        .array(z.string())
+        .optional()
+        .describe('A list of domains to specifically include in the search results.'),
+      exclude_domains: z
+        .array(z.string())
+        .optional()
+        .describe('A list of domains to specifically exclude from the search results.'),
+      topic: z
+        .enum(['general', 'news', 'finance'])
+        .optional()
+        .describe(
+          'The category of the search. Use news ONLY if query SPECIFCALLY mentions the word "news".',
+        ),
+      time_range: z
+        .enum(['day', 'week', 'month', 'year', 'd', 'w', 'm', 'y'])
+        .optional()
+        .describe('The time range back from the current date to filter results.'),
+      days: z
+        .number()
+        .min(1)
+        .optional()
+        .describe('Number of days back from the current date to include. Only if topic is news.'),
+      include_image_descriptions: z
+        .boolean()
+        .optional()
+        .describe(
+          'When include_images is true, also add a descriptive text for each image. Default is false.',
+        ),
     });
   }
 

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

@@ -1,7 +1,7 @@
 const { z } = require('zod');
 const { Tool } = require('@langchain/core/tools');
+const { logger } = require('@librechat/data-schemas');
 const { getEnvironmentVariable } = require('@langchain/core/utils/env');
-const { logger } = require('~/config');
 
 /**
  * Tool for the Traversaal AI search API, Ares.
@@ -21,7 +21,7 @@ class TraversaalSearch extends Tool {
       query: z
         .string()
         .describe(
-          'A properly written sentence to be interpreted by an AI to search the web according to the user\'s request.',
+          "A properly written sentence to be interpreted by an AI to search the web according to the user's request.",
         ),
     });
 
@@ -38,7 +38,6 @@ class TraversaalSearch extends Tool {
     return apiKey;
   }
 
-  // eslint-disable-next-line no-unused-vars
   async _call({ query }, _runManager) {
     const body = {
       query: [query],

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

@@ -1,8 +1,8 @@
 /* eslint-disable no-useless-escape */
-const axios = require('axios');
 const { z } = require('zod');
+const axios = require('axios');
 const { Tool } = require('@langchain/core/tools');
-const { logger } = require('~/config');
+const { logger } = require('@librechat/data-schemas');
 
 class WolframAlphaAPI extends Tool {
   constructor(fields) {

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

@@ -1,9 +1,9 @@
-const { z } = require('zod');
+const { ytToolkit } = require('@librechat/api');
 const { tool } = require('@langchain/core/tools');
 const { youtube } = require('@googleapis/youtube');
+const { logger } = require('@librechat/data-schemas');
 const { YoutubeTranscript } = require('youtube-transcript');
 const { getApiKey } = require('./credentials');
-const { logger } = require('~/config');
 
 function extractVideoId(url) {
   const rawIdRegex = /^[a-zA-Z0-9_-]{11}$/;
@@ -29,7 +29,7 @@ function parseTranscript(transcriptResponse) {
     .map((entry) => entry.text.trim())
     .filter((text) => text)
     .join(' ')
-    .replaceAll(''', '\'');
+    .replaceAll(''', "'");
 }
 
 function createYouTubeTools(fields = {}) {
@@ -42,160 +42,94 @@ function createYouTubeTools(fields = {}) {
     auth: apiKey,
   });
 
-  const searchTool = tool(
-    async ({ query, maxResults = 5 }) => {
-      const response = await youtubeClient.search.list({
-        part: 'snippet',
-        q: query,
-        type: 'video',
-        maxResults: maxResults || 5,
-      });
-      const result = response.data.items.map((item) => ({
-        title: item.snippet.title,
-        description: item.snippet.description,
-        url: `https://www.youtube.com/watch?v=${item.id.videoId}`,
-      }));
-      return JSON.stringify(result, null, 2);
-    },
-    {
-      name: 'youtube_search',
-      description: `Search for YouTube videos by keyword or phrase.
-- Required: query (search terms to find videos)
-- Optional: maxResults (number of videos to return, 1-50, default: 5)
-- Returns: List of videos with titles, descriptions, and URLs
-- Use for: Finding specific videos, exploring content, research
-Example: query="cooking pasta tutorials" maxResults=3`,
-      schema: z.object({
-        query: z.string().describe('Search query terms'),
-        maxResults: z.number().int().min(1).max(50).optional().describe('Number of results (1-50)'),
-      }),
-    },
-  );
+  const searchTool = tool(async ({ query, maxResults = 5 }) => {
+    const response = await youtubeClient.search.list({
+      part: 'snippet',
+      q: query,
+      type: 'video',
+      maxResults: maxResults || 5,
+    });
+    const result = response.data.items.map((item) => ({
+      title: item.snippet.title,
+      description: item.snippet.description,
+      url: `https://www.youtube.com/watch?v=${item.id.videoId}`,
+    }));
+    return JSON.stringify(result, null, 2);
+  }, ytToolkit.youtube_search);
 
-  const infoTool = tool(
-    async ({ url }) => {
-      const videoId = extractVideoId(url);
-      if (!videoId) {
-        throw new Error('Invalid YouTube URL or video ID');
-      }
+  const infoTool = tool(async ({ url }) => {
+    const videoId = extractVideoId(url);
+    if (!videoId) {
+      throw new Error('Invalid YouTube URL or video ID');
+    }
 
-      const response = await youtubeClient.videos.list({
-        part: 'snippet,statistics',
-        id: videoId,
-      });
+    const response = await youtubeClient.videos.list({
+      part: 'snippet,statistics',
+      id: videoId,
+    });
 
-      if (!response.data.items?.length) {
-        throw new Error('Video not found');
-      }
-      const video = response.data.items[0];
+    if (!response.data.items?.length) {
+      throw new Error('Video not found');
+    }
+    const video = response.data.items[0];
 
-      const result = {
-        title: video.snippet.title,
-        description: video.snippet.description,
-        views: video.statistics.viewCount,
-        likes: video.statistics.likeCount,
-        comments: video.statistics.commentCount,
-      };
-      return JSON.stringify(result, null, 2);
-    },
-    {
-      name: 'youtube_info',
-      description: `Get detailed metadata and statistics for a specific YouTube video.
-- Required: url (full YouTube URL or video ID)
-- Returns: Video title, description, view count, like count, comment count
-- Use for: Getting video metrics and basic metadata
-- DO NOT USE FOR VIDEO SUMMARIES, USE TRANSCRIPTS FOR COMPREHENSIVE ANALYSIS
-- Accepts both full URLs and video IDs
-Example: url="https://youtube.com/watch?v=abc123" or url="abc123"`,
-      schema: z.object({
-        url: z.string().describe('YouTube video URL or ID'),
-      }),
-    },
-  );
+    const result = {
+      title: video.snippet.title,
+      description: video.snippet.description,
+      views: video.statistics.viewCount,
+      likes: video.statistics.likeCount,
+      comments: video.statistics.commentCount,
+    };
+    return JSON.stringify(result, null, 2);
+  }, ytToolkit.youtube_info);
 
-  const commentsTool = tool(
-    async ({ url, maxResults = 10 }) => {
-      const videoId = extractVideoId(url);
-      if (!videoId) {
-        throw new Error('Invalid YouTube URL or video ID');
-      }
+  const commentsTool = tool(async ({ url, maxResults = 10 }) => {
+    const videoId = extractVideoId(url);
+    if (!videoId) {
+      throw new Error('Invalid YouTube URL or video ID');
+    }
 
-      const response = await youtubeClient.commentThreads.list({
-        part: 'snippet',
-        videoId,
-        maxResults: maxResults || 10,
-      });
+    const response = await youtubeClient.commentThreads.list({
+      part: 'snippet',
+      videoId,
+      maxResults: maxResults || 10,
+    });
 
-      const result = response.data.items.map((item) => ({
-        author: item.snippet.topLevelComment.snippet.authorDisplayName,
-        text: item.snippet.topLevelComment.snippet.textDisplay,
-        likes: item.snippet.topLevelComment.snippet.likeCount,
-      }));
-      return JSON.stringify(result, null, 2);
-    },
-    {
-      name: 'youtube_comments',
-      description: `Retrieve top-level comments from a YouTube video.
-- Required: url (full YouTube URL or video ID)
-- Optional: maxResults (number of comments, 1-50, default: 10)
-- Returns: Comment text, author names, like counts
-- Use for: Sentiment analysis, audience feedback, engagement review
-Example: url="abc123" maxResults=20`,
-      schema: z.object({
-        url: z.string().describe('YouTube video URL or ID'),
-        maxResults: z
-          .number()
-          .int()
-          .min(1)
-          .max(50)
-          .optional()
-          .describe('Number of comments to retrieve'),
-      }),
-    },
-  );
+    const result = response.data.items.map((item) => ({
+      author: item.snippet.topLevelComment.snippet.authorDisplayName,
+      text: item.snippet.topLevelComment.snippet.textDisplay,
+      likes: item.snippet.topLevelComment.snippet.likeCount,
+    }));
+    return JSON.stringify(result, null, 2);
+  }, ytToolkit.youtube_comments);
 
-  const transcriptTool = tool(
-    async ({ url }) => {
-      const videoId = extractVideoId(url);
-      if (!videoId) {
-        throw new Error('Invalid YouTube URL or video ID');
+  const transcriptTool = tool(async ({ url }) => {
+    const videoId = extractVideoId(url);
+    if (!videoId) {
+      throw new Error('Invalid YouTube URL or video ID');
+    }
+
+    try {
+      try {
+        const transcript = await YoutubeTranscript.fetchTranscript(videoId, { lang: 'en' });
+        return parseTranscript(transcript);
+      } catch (e) {
+        logger.error(e);
       }
 
       try {
-        try {
-          const transcript = await YoutubeTranscript.fetchTranscript(videoId, { lang: 'en' });
-          return parseTranscript(transcript);
-        } catch (e) {
-          logger.error(e);
-        }
-
-        try {
-          const transcript = await YoutubeTranscript.fetchTranscript(videoId, { lang: 'de' });
-          return parseTranscript(transcript);
-        } catch (e) {
-          logger.error(e);
-        }
-
-        const transcript = await YoutubeTranscript.fetchTranscript(videoId);
+        const transcript = await YoutubeTranscript.fetchTranscript(videoId, { lang: 'de' });
         return parseTranscript(transcript);
-      } catch (error) {
-        throw new Error(`Failed to fetch transcript: ${error.message}`);
+      } catch (e) {
+        logger.error(e);
       }
-    },
-    {
-      name: 'youtube_transcript',
-      description: `Fetch and parse the transcript/captions of a YouTube video.
-- Required: url (full YouTube URL or video ID)
-- Returns: Full video transcript as plain text
-- Use for: Content analysis, summarization, translation reference
-- This is the "Go-to" tool for analyzing actual video content
-- Attempts to fetch English first, then German, then any available language
-Example: url="https://youtube.com/watch?v=abc123"`,
-      schema: z.object({
-        url: z.string().describe('YouTube video URL or ID'),
-      }),
-    },
-  );
+
+      const transcript = await YoutubeTranscript.fetchTranscript(videoId);
+      return parseTranscript(transcript);
+    } catch (error) {
+      throw new Error(`Failed to fetch transcript: ${error.message}`);
+    }
+  }, ytToolkit.youtube_transcript);
 
   return [searchTool, infoTool, commentsTool, transcriptTool];
 }

api/app/clients/tools/structured/specs/DALLE3-proxy.spec.js

@@ -0,0 +1,60 @@
+const DALLE3 = require('../DALLE3');
+const { ProxyAgent } = require('undici');
+
+jest.mock('tiktoken');
+const processFileURL = jest.fn();
+
+describe('DALLE3 Proxy Configuration', () => {
+  let originalEnv;
+
+  beforeAll(() => {
+    originalEnv = { ...process.env };
+  });
+
+  beforeEach(() => {
+    jest.resetModules();
+    process.env = { ...originalEnv };
+  });
+
+  afterEach(() => {
+    process.env = originalEnv;
+  });
+
+  it('should configure ProxyAgent in fetchOptions.dispatcher when PROXY env is set', () => {
+    // Set proxy environment variable
+    process.env.PROXY = 'http://proxy.example.com:8080';
+    process.env.DALLE_API_KEY = 'test-api-key';
+
+    // Create instance
+    const dalleWithProxy = new DALLE3({ processFileURL });
+
+    // Check that the openai client exists
+    expect(dalleWithProxy.openai).toBeDefined();
+
+    // Check that _options exists and has fetchOptions with a dispatcher
+    expect(dalleWithProxy.openai._options).toBeDefined();
+    expect(dalleWithProxy.openai._options.fetchOptions).toBeDefined();
+    expect(dalleWithProxy.openai._options.fetchOptions.dispatcher).toBeDefined();
+    expect(dalleWithProxy.openai._options.fetchOptions.dispatcher).toBeInstanceOf(ProxyAgent);
+  });
+
+  it('should not configure ProxyAgent when PROXY env is not set', () => {
+    // Ensure PROXY is not set
+    delete process.env.PROXY;
+    process.env.DALLE_API_KEY = 'test-api-key';
+
+    // Create instance
+    const dalleWithoutProxy = new DALLE3({ processFileURL });
+
+    // Check that the openai client exists
+    expect(dalleWithoutProxy.openai).toBeDefined();
+
+    // Check that _options exists but fetchOptions either doesn't exist or doesn't have a dispatcher
+    expect(dalleWithoutProxy.openai._options).toBeDefined();
+
+    // fetchOptions should either not exist or not have a dispatcher
+    if (dalleWithoutProxy.openai._options.fetchOptions) {
+      expect(dalleWithoutProxy.openai._options.fetchOptions.dispatcher).toBeUndefined();
+    }
+  });
+});

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

@@ -1,31 +1,30 @@
 const OpenAI = require('openai');
+const { logger } = require('@librechat/data-schemas');
 const DALLE3 = require('../DALLE3');
 
-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', () => ({
-  getImageBasename: jest.fn().mockImplementation((url) => {
-    // Split the URL by '/'
-    const parts = url.split('/');
-
-    // Get the last part of the URL
-    const lastPart = parts.pop();
-
-    // Check if the last part of the URL matches the image extension regex
-    const imageExtensionRegex = /\.(jpg|jpeg|png|gif|bmp|tiff|svg)$/i;
-    if (imageExtensionRegex.test(lastPart)) {
-      return lastPart;
-    }
-
-    // If the regex test fails, return an empty string
-    return '';
-  }),
-}));
-
 const generate = jest.fn();
 OpenAI.mockImplementation(() => ({
   images: {
@@ -37,6 +36,11 @@ 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/addOpenAPISpecs.js

@@ -1,30 +0,0 @@
-const { loadSpecs } = require('./loadSpecs');
-
-function transformSpec(input) {
-  return {
-    name: input.name_for_human,
-    pluginKey: input.name_for_model,
-    description: input.description_for_human,
-    icon: input?.logo_url ?? 'https://placehold.co/70x70.png',
-    // TODO: add support for authentication
-    isAuthRequired: 'false',
-    authConfig: [],
-  };
-}
-
-async function addOpenAPISpecs(availableTools) {
-  try {
-    const specs = (await loadSpecs({})).map(transformSpec);
-    if (specs.length > 0) {
-      return [...specs, ...availableTools];
-    }
-    return availableTools;
-  } catch (error) {
-    return availableTools;
-  }
-}
-
-module.exports = {
-  transformSpec,
-  addOpenAPISpecs,
-};

api/app/clients/tools/util/addOpenAPISpecs.spec.js

@@ -1,76 +0,0 @@
-const { addOpenAPISpecs, transformSpec } = require('./addOpenAPISpecs');
-const { loadSpecs } = require('./loadSpecs');
-const { createOpenAPIPlugin } = require('../dynamic/OpenAPIPlugin');
-
-jest.mock('./loadSpecs');
-jest.mock('../dynamic/OpenAPIPlugin');
-
-describe('transformSpec', () => {
-  it('should transform input spec to a desired format', () => {
-    const input = {
-      name_for_human: 'Human Name',
-      name_for_model: 'Model Name',
-      description_for_human: 'Human Description',
-      logo_url: 'https://example.com/logo.png',
-    };
-
-    const expectedOutput = {
-      name: 'Human Name',
-      pluginKey: 'Model Name',
-      description: 'Human Description',
-      icon: 'https://example.com/logo.png',
-      isAuthRequired: 'false',
-      authConfig: [],
-    };
-
-    expect(transformSpec(input)).toEqual(expectedOutput);
-  });
-
-  it('should use default icon if logo_url is not provided', () => {
-    const input = {
-      name_for_human: 'Human Name',
-      name_for_model: 'Model Name',
-      description_for_human: 'Human Description',
-    };
-
-    const expectedOutput = {
-      name: 'Human Name',
-      pluginKey: 'Model Name',
-      description: 'Human Description',
-      icon: 'https://placehold.co/70x70.png',
-      isAuthRequired: 'false',
-      authConfig: [],
-    };
-
-    expect(transformSpec(input)).toEqual(expectedOutput);
-  });
-});
-
-describe('addOpenAPISpecs', () => {
-  it('should add specs to available tools', async () => {
-    const availableTools = ['Tool1', 'Tool2'];
-    const specs = [
-      {
-        name_for_human: 'Human Name',
-        name_for_model: 'Model Name',
-        description_for_human: 'Human Description',
-        logo_url: 'https://example.com/logo.png',
-      },
-    ];
-
-    loadSpecs.mockResolvedValue(specs);
-    createOpenAPIPlugin.mockReturnValue('Plugin');
-
-    const result = await addOpenAPISpecs(availableTools);
-    expect(result).toEqual([...specs.map(transformSpec), ...availableTools]);
-  });
-
-  it('should return available tools if specs loading fails', async () => {
-    const availableTools = ['Tool1', 'Tool2'];
-
-    loadSpecs.mockRejectedValue(new Error('Failed to load specs'));
-
-    const result = await addOpenAPISpecs(availableTools);
-    expect(result).toEqual(availableTools);
-  });
-});

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

@@ -1,26 +1,46 @@
 const { z } = require('zod');
 const axios = require('axios');
 const { tool } = require('@langchain/core/tools');
+const { logger } = require('@librechat/data-schemas');
+const { generateShortLivedToken } = require('@librechat/api');
 const { Tools, EToolResources } = require('librechat-data-provider');
+const { filterFilesByAgentAccess } = require('~/server/services/Files/permissions');
 const { getFiles } = require('~/models/File');
-const { logger } = require('~/config');
 
 /**
  *
  * @param {Object} options
  * @param {ServerRequest} options.req
  * @param {Agent['tool_resources']} options.tool_resources
+ * @param {string} [options.agentId] - The agent ID for file access control
  * @returns {Promise<{
  *   files: Array<{ file_id: string; filename: string }>,
  *   toolContext: string
  * }>}
  */
 const primeFiles = async (options) => {
-  const { tool_resources } = options;
+  const { tool_resources, req, agentId } = options;
   const file_ids = tool_resources?.[EToolResources.file_search]?.file_ids ?? [];
   const agentResourceIds = new Set(file_ids);
   const resourceFiles = tool_resources?.[EToolResources.file_search]?.files ?? [];
-  const dbFiles = ((await getFiles({ file_id: { $in: file_ids } })) ?? []).concat(resourceFiles);
+
+  // Get all files first
+  const allFiles = (await getFiles({ file_id: { $in: file_ids } }, null, { text: 0 })) ?? [];
+
+  // Filter by access if user and agent are provided
+  let dbFiles;
+  if (req?.user?.id && agentId) {
+    dbFiles = await filterFilesByAgentAccess({
+      files: allFiles,
+      userId: req.user.id,
+      role: req.user.role,
+      agentId,
+    });
+  } else {
+    dbFiles = allFiles;
+  }
+
+  dbFiles = dbFiles.concat(resourceFiles);
 
   let toolContext = `- Note: Semantic search is available through the ${Tools.file_search} tool but no files are currently loaded. Request the user to upload documents to search through.`;
 
@@ -48,18 +68,19 @@ const primeFiles = async (options) => {
 /**
  *
  * @param {Object} options
- * @param {ServerRequest} options.req
+ * @param {string} options.userId
  * @param {Array<{ file_id: string; filename: string }>} options.files
  * @param {string} [options.entity_id]
+ * @param {boolean} [options.fileCitations=false] - Whether to include citation instructions
  * @returns
  */
-const createFileSearchTool = async ({ req, files, entity_id }) => {
+const createFileSearchTool = async ({ userId, files, entity_id, fileCitations = false }) => {
   return tool(
     async ({ query }) => {
       if (files.length === 0) {
         return 'No files to search. Instruct the user to add files for the search.';
       }
-      const jwtToken = req.headers.authorization.split(' ')[1];
+      const jwtToken = generateShortLivedToken(userId);
       if (!jwtToken) {
         return 'There was an error authenticating the file search request.';
       }
@@ -105,11 +126,13 @@ const createFileSearchTool = async ({ req, files, entity_id }) => {
       }
 
       const formattedResults = validResults
-        .flatMap((result) =>
+        .flatMap((result, fileIndex) =>
           result.data.map(([docInfo, distance]) => ({
             filename: docInfo.metadata.source.split('/').pop(),
             content: docInfo.page_content,
             distance,
+            file_id: files[fileIndex]?.file_id,
+            page: docInfo.metadata.page || null,
           })),
         )
         // TODO: results should be sorted by relevance, not distance
@@ -119,23 +142,46 @@ const createFileSearchTool = async ({ req, files, entity_id }) => {
 
       const formattedString = formattedResults
         .map(
-          (result) =>
-            `File: ${result.filename}\nRelevance: ${1.0 - result.distance.toFixed(4)}\nContent: ${
-              result.content
-            }\n`,
+          (result, index) =>
+            `File: ${result.filename}${
+              fileCitations ? `\nAnchor: \\ue202turn0file${index} (${result.filename})` : ''
+            }\nRelevance: ${(1.0 - result.distance).toFixed(4)}\nContent: ${result.content}\n`,
         )
         .join('\n---\n');
 
-      return formattedString;
+      const sources = formattedResults.map((result) => ({
+        type: 'file',
+        fileId: result.file_id,
+        content: result.content,
+        fileName: result.filename,
+        relevance: 1.0 - result.distance,
+        pages: result.page ? [result.page] : [],
+        pageRelevance: result.page ? { [result.page]: 1.0 - result.distance } : {},
+      }));
+
+      return [formattedString, { [Tools.file_search]: { sources, fileCitations } }];
     },
     {
       name: Tools.file_search,
-      description: `Performs semantic search across attached "${Tools.file_search}" documents using natural language queries. This tool analyzes the content of uploaded files to find relevant information, quotes, and passages that best match your query. Use this to extract specific information or find relevant sections within the available documents.`,
+      responseFormat: 'content_and_artifact',
+      description: `Performs semantic search across attached "${Tools.file_search}" documents using natural language queries. This tool analyzes the content of uploaded files to find relevant information, quotes, and passages that best match your query. Use this to extract specific information or find relevant sections within the available documents.${
+        fileCitations
+          ? `
+
+**CITE FILE SEARCH RESULTS:**
+Use anchor markers immediately after statements derived from file content. Reference the filename in your text:
+- File citation: "The document.pdf states that... \\ue202turn0file0"  
+- Page reference: "According to report.docx... \\ue202turn0file1"
+- Multi-file: "Multiple sources confirm... \\ue200\\ue202turn0file0\\ue202turn0file1\\ue201"
+
+**ALWAYS mention the filename in your text before the citation marker. NEVER use markdown links or footnotes.**`
+          : ''
+      }`,
       schema: z.object({
         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/handleOpenAIErrors.js

@@ -1,5 +1,5 @@
 const OpenAI = require('openai');
-const { logger } = require('~/config');
+const { logger } = require('@librechat/data-schemas');
 
 /**
  * Handles errors that may occur when making requests to OpenAI's API.

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

@@ -1,8 +1,21 @@
+const { logger } = require('@librechat/data-schemas');
 const { SerpAPI } = require('@langchain/community/tools/serpapi');
 const { Calculator } = require('@langchain/community/tools/calculator');
-const { createCodeExecutionTool, EnvVar } = require('@librechat/agents');
-const { Tools, Constants, EToolResources } = require('librechat-data-provider');
-const { getUserPluginAuthValue } = require('~/server/services/PluginService');
+const { EnvVar, createCodeExecutionTool, createSearchTool } = require('@librechat/agents');
+const {
+  checkAccess,
+  createSafeUser,
+  mcpToolPattern,
+  loadWebSearchAuth,
+} = require('@librechat/api');
+const {
+  Tools,
+  Constants,
+  Permissions,
+  EToolResources,
+  PermissionTypes,
+  replaceSpecialVars,
+} = require('librechat-data-provider');
 const {
   availableTools,
   manifestToolMap,
@@ -22,12 +35,11 @@ const {
 } = require('../');
 const { primeFiles: primeCodeFiles } = require('~/server/services/Files/Code/process');
 const { createFileSearchTool, primeFiles: primeSearchFiles } = require('./fileSearch');
+const { getUserPluginAuthValue } = require('~/server/services/PluginService');
+const { createMCPTool, createMCPTools } = require('~/server/services/MCP');
 const { loadAuthValues } = require('~/server/services/Tools/credentials');
-const { createMCPTool } = require('~/server/services/MCP');
-const { loadSpecs } = require('./loadSpecs');
-const { logger } = require('~/config');
-
-const mcpToolPattern = new RegExp(`^.+${Constants.mcp_delimiter}.+$`);
+const { getMCPServerTools } = require('~/server/services/Config');
+const { getRoleByName } = require('~/models/Role');
 
 /**
  * Validates the availability and authentication of tools for a user based on environment variables or user-specific plugin authentication values.
@@ -88,7 +100,7 @@ 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('There was a problem validating tools');
+    throw new Error(err);
   }
 };
 
@@ -122,28 +134,37 @@ const getAuthFields = (toolKey) => {
 
 /**
  *
- * @param {object} object
- * @param {string} object.user
- * @param {Pick<Agent, 'id' | 'provider' | 'model'>} [object.agent]
- * @param {string} [object.model]
- * @param {EModelEndpoint} [object.endpoint]
- * @param {LoadToolOptions} [object.options]
- * @param {boolean} [object.useSpecs]
- * @param {Array<string>} object.tools
- * @param {boolean} [object.functions]
- * @param {boolean} [object.returnMap]
+ * @param {object} params
+ * @param {string} params.user
+ * @param {Record<string, Record<string, string>>} [object.userMCPAuthMap]
+ * @param {AbortSignal} [object.signal]
+ * @param {Pick<Agent, 'id' | 'provider' | 'model'>} [params.agent]
+ * @param {string} [params.model]
+ * @param {EModelEndpoint} [params.endpoint]
+ * @param {LoadToolOptions} [params.options]
+ * @param {boolean} [params.useSpecs]
+ * @param {Array<string>} params.tools
+ * @param {boolean} [params.functions]
+ * @param {boolean} [params.returnMap]
+ * @param {AppConfig['webSearch']} [params.webSearch]
+ * @param {AppConfig['fileStrategy']} [params.fileStrategy]
+ * @param {AppConfig['imageOutputType']} [params.imageOutputType]
  * @returns {Promise<{ loadedTools: Tool[], toolContextMap: Object<string, any> } | Record<string,Tool>>}
  */
 const loadTools = async ({
   user,
   agent,
   model,
+  signal,
   endpoint,
-  useSpecs,
+  userMCPAuthMap,
   tools = [],
   options = {},
   functions = true,
   returnMap = false,
+  webSearch,
+  fileStrategy,
+  imageOutputType,
 }) => {
   const toolConstructors = {
     flux: FluxAPI,
@@ -202,6 +223,8 @@ const loadTools = async ({
         ...authValues,
         isAgent: !!agent,
         req: options.req,
+        imageOutputType,
+        fileStrategy,
         imageFiles,
       });
     },
@@ -217,7 +240,7 @@ const loadTools = async ({
   const imageGenOptions = {
     isAgent: !!agent,
     req: options.req,
-    fileStrategy: options.fileStrategy,
+    fileStrategy,
     processFileURL: options.processFileURL,
     returnMetadata: options.returnMetadata,
     uploadImageBuffer: options.uploadImageBuffer,
@@ -232,8 +255,7 @@ const loadTools = async ({
 
   /** @type {Record<string, string>} */
   const toolContextMap = {};
-  const remainingTools = [];
-  const appTools = options.req?.app?.locals?.availableTools ?? {};
+  const requestedMCPTools = {};
 
   for (const tool of tools) {
     if (tool === Tools.execute_code) {
@@ -243,7 +265,13 @@ const loadTools = async ({
           authFields: [EnvVar.CODE_API_KEY],
         });
         const codeApiKey = authValues[EnvVar.CODE_API_KEY];
-        const { files, toolContext } = await primeCodeFiles(options, codeApiKey);
+        const { files, toolContext } = await primeCodeFiles(
+          {
+            ...options,
+            agentId: agent?.id,
+          },
+          codeApiKey,
+        );
         if (toolContext) {
           toolContextMap[tool] = toolContext;
         }
@@ -258,21 +286,92 @@ const loadTools = async ({
       continue;
     } else if (tool === Tools.file_search) {
       requestedTools[tool] = async () => {
-        const { files, toolContext } = await primeSearchFiles(options);
+        const { files, toolContext } = await primeSearchFiles({
+          ...options,
+          agentId: agent?.id,
+        });
         if (toolContext) {
           toolContextMap[tool] = toolContext;
         }
-        return createFileSearchTool({ req: options.req, files, entity_id: agent?.id });
+
+        /** @type {boolean | undefined} Check if user has FILE_CITATIONS permission */
+        let fileCitations;
+        if (fileCitations == null && options.req?.user != null) {
+          try {
+            fileCitations = await checkAccess({
+              user: options.req.user,
+              permissionType: PermissionTypes.FILE_CITATIONS,
+              permissions: [Permissions.USE],
+              getRoleByName,
+            });
+          } catch (error) {
+            logger.error('[handleTools] FILE_CITATIONS permission check failed:', error);
+            fileCitations = false;
+          }
+        }
+
+        return createFileSearchTool({
+          userId: user,
+          files,
+          entity_id: agent?.id,
+          fileCitations,
+        });
       };
       continue;
-    } else if (tool && appTools[tool] && mcpToolPattern.test(tool)) {
-      requestedTools[tool] = async () =>
-        createMCPTool({
-          req: options.req,
-          toolKey: tool,
-          model: agent?.model ?? model,
-          provider: agent?.provider ?? endpoint,
+    } else if (tool === Tools.web_search) {
+      const result = await loadWebSearchAuth({
+        userId: user,
+        loadAuthValues,
+        webSearchConfig: webSearch,
+      });
+      const { onSearchResults, onGetHighlights } = options?.[Tools.web_search] ?? {};
+      requestedTools[tool] = async () => {
+        toolContextMap[tool] = `# \`${tool}\`:
+Current Date & Time: ${replaceSpecialVars({ text: '{{iso_datetime}}' })}
+1. **Execute immediately without preface** when using \`${tool}\`.
+2. **After the search, begin with a brief summary** that directly addresses the query without headers or explaining your process.
+3. **Structure your response clearly** using Markdown formatting (Level 2 headers for sections, lists for multiple points, tables for comparisons).
+4. **Cite sources properly** according to the citation anchor format, utilizing group anchors when appropriate.
+5. **Tailor your approach to the query type** (academic, news, coding, etc.) while maintaining an expert, journalistic, unbiased tone.
+6. **Provide comprehensive information** with specific details, examples, and as much relevant context as possible from search results.
+7. **Avoid moralizing language.**
+`.trim();
+        return createSearchTool({
+          ...result.authResult,
+          onSearchResults,
+          onGetHighlights,
+          logger,
         });
+      };
+      continue;
+    } else if (tool && mcpToolPattern.test(tool)) {
+      const [toolName, serverName] = tool.split(Constants.mcp_delimiter);
+      if (toolName === Constants.mcp_server) {
+        /** Placeholder used for UI purposes */
+        continue;
+      }
+      if (serverName && options.req?.config?.mcpConfig?.[serverName] == null) {
+        logger.warn(
+          `MCP server "${serverName}" for "${toolName}" tool is not configured${agent?.id != null && agent.id ? ` but attached to "${agent.id}"` : ''}`,
+        );
+        continue;
+      }
+      if (toolName === Constants.mcp_all) {
+        requestedMCPTools[serverName] = [
+          {
+            type: 'all',
+            serverName,
+          },
+        ];
+        continue;
+      }
+
+      requestedMCPTools[serverName] = requestedMCPTools[serverName] || [];
+      requestedMCPTools[serverName].push({
+        type: 'single',
+        toolKey: tool,
+        serverName,
+      });
       continue;
     }
 
@@ -292,30 +391,6 @@ const loadTools = async ({
       requestedTools[tool] = toolInstance;
       continue;
     }
-
-    if (functions === true) {
-      remainingTools.push(tool);
-    }
-  }
-
-  let specs = null;
-  if (useSpecs === true && functions === true && remainingTools.length > 0) {
-    specs = await loadSpecs({
-      llm: model,
-      user,
-      message: options.message,
-      memory: options.memory,
-      signal: options.signal,
-      tools: remainingTools,
-      map: true,
-      verbose: false,
-    });
-  }
-
-  for (const tool of remainingTools) {
-    if (specs && specs[tool]) {
-      requestedTools[tool] = specs[tool];
-    }
   }
 
   if (returnMap) {
@@ -336,6 +411,75 @@ const loadTools = async ({
   }
 
   const loadedTools = (await Promise.all(toolPromises)).flatMap((plugin) => plugin || []);
+  const mcpToolPromises = [];
+  /** MCP server tools are initialized sequentially by server */
+  let index = -1;
+  const failedMCPServers = new Set();
+  const safeUser = createSafeUser(options.req?.user);
+  for (const [serverName, toolConfigs] of Object.entries(requestedMCPTools)) {
+    index++;
+    /** @type {LCAvailableTools} */
+    let availableTools;
+    for (const config of toolConfigs) {
+      try {
+        if (failedMCPServers.has(serverName)) {
+          continue;
+        }
+        const mcpParams = {
+          index,
+          signal,
+          user: safeUser,
+          userMCPAuthMap,
+          res: options.res,
+          model: agent?.model ?? model,
+          serverName: config.serverName,
+          provider: agent?.provider ?? endpoint,
+        };
+
+        if (config.type === 'all' && toolConfigs.length === 1) {
+          /** Handle async loading for single 'all' tool config */
+          mcpToolPromises.push(
+            createMCPTools(mcpParams).catch((error) => {
+              logger.error(`Error loading ${serverName} tools:`, error);
+              return null;
+            }),
+          );
+          continue;
+        }
+        if (!availableTools) {
+          try {
+            availableTools = await getMCPServerTools(serverName);
+          } catch (error) {
+            logger.error(`Error fetching available tools for MCP server ${serverName}:`, error);
+          }
+        }
+
+        /** Handle synchronous loading */
+        const mcpTool =
+          config.type === 'all'
+            ? await createMCPTools(mcpParams)
+            : await createMCPTool({
+                ...mcpParams,
+                availableTools,
+                toolKey: config.toolKey,
+              });
+
+        if (Array.isArray(mcpTool)) {
+          loadedTools.push(...mcpTool);
+        } else if (mcpTool) {
+          loadedTools.push(mcpTool);
+        } else {
+          failedMCPServers.add(serverName);
+          logger.warn(
+            `MCP tool creation failed for "${config.toolKey}", server may be unavailable or unauthenticated.`,
+          );
+        }
+      } catch (error) {
+        logger.error(`Error loading MCP tool for server ${serverName}:`, error);
+      }
+    }
+  }
+  loadedTools.push(...(await Promise.all(mcpToolPromises)).flatMap((plugin) => plugin || []));
   return { loadedTools, toolContextMap };
 };
 

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

@@ -1,8 +1,5 @@
-const mockUser = {
-  _id: 'fakeId',
-  save: jest.fn(),
-  findByIdAndDelete: jest.fn(),
-};
+const mongoose = require('mongoose');
+const { MongoMemoryServer } = require('mongodb-memory-server');
 
 const mockPluginService = {
   updateUserPluginAuth: jest.fn(),
@@ -10,23 +7,39 @@ const mockPluginService = {
   getUserPluginAuthValue: jest.fn(),
 };
 
-jest.mock('~/models/User', () => {
-  return function () {
-    return mockUser;
-  };
-});
-
 jest.mock('~/server/services/PluginService', () => mockPluginService);
 
+jest.mock('~/server/services/Config', () => ({
+  getAppConfig: jest.fn().mockResolvedValue({
+    // Default app config for tool tests
+    paths: { uploads: '/tmp' },
+    fileStrategy: 'local',
+    filteredTools: [],
+    includedTools: [],
+  }),
+  getCachedTools: jest.fn().mockResolvedValue({
+    // Default cached tools for tests
+    dalle: {
+      type: 'function',
+      function: {
+        name: 'dalle',
+        description: 'DALL-E image generation',
+        parameters: {},
+      },
+    },
+  }),
+}));
+
 const { BaseLLM } = require('@langchain/openai');
 const { Calculator } = require('@langchain/community/tools/calculator');
 
-const User = require('~/models/User');
+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';
@@ -37,7 +50,9 @@ describe('Tool Handlers', () => {
   const authConfigs = mainPlugin.authConfig;
 
   beforeAll(async () => {
-    mockUser.save.mockResolvedValue(undefined);
+    mongoServer = await MongoMemoryServer.create();
+    const mongoUri = mongoServer.getUri();
+    await mongoose.connect(mongoUri);
 
     const userAuthValues = {};
     mockPluginService.getUserPluginAuthValue.mockImplementation((userId, authField) => {
@@ -78,9 +93,36 @@ describe('Tool Handlers', () => {
   });
 
   afterAll(async () => {
-    await mockUser.findByIdAndDelete(fakeUser._id);
+    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
     for (const authConfig of authConfigs) {
-      await PluginService.deleteUserPluginAuth(fakeUser._id, authConfig.authField);
+      await PluginService.updateUserPluginAuth(
+        fakeUser._id,
+        authConfig.authField,
+        pluginKey,
+        mockCredential,
+      );
     }
   });
 
@@ -218,7 +260,6 @@ describe('Tool Handlers', () => {
       try {
         await loadTool2();
       } catch (error) {
-        // eslint-disable-next-line jest/no-conditional-expect
         expect(error).toBeDefined();
       }
     });

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

@@ -1,117 +0,0 @@
-const fs = require('fs');
-const path = require('path');
-const { z } = require('zod');
-const { logger } = require('~/config');
-const { createOpenAPIPlugin } = require('~/app/clients/tools/dynamic/OpenAPIPlugin');
-
-// The minimum Manifest definition
-const ManifestDefinition = z.object({
-  schema_version: z.string().optional(),
-  name_for_human: z.string(),
-  name_for_model: z.string(),
-  description_for_human: z.string(),
-  description_for_model: z.string(),
-  auth: z.object({}).optional(),
-  api: z.object({
-    // Spec URL or can be the filename of the OpenAPI spec yaml file,
-    // located in api\app\clients\tools\.well-known\openapi
-    url: z.string(),
-    type: z.string().optional(),
-    is_user_authenticated: z.boolean().nullable().optional(),
-    has_user_authentication: z.boolean().nullable().optional(),
-  }),
-  // use to override any params that the LLM will consistently get wrong
-  params: z.object({}).optional(),
-  logo_url: z.string().optional(),
-  contact_email: z.string().optional(),
-  legal_info_url: z.string().optional(),
-});
-
-function validateJson(json) {
-  try {
-    return ManifestDefinition.parse(json);
-  } catch (error) {
-    logger.debug('[validateJson] manifest parsing error', error);
-    return false;
-  }
-}
-
-// omit the LLM to return the well known jsons as objects
-async function loadSpecs({ llm, user, message, tools = [], map = false, memory, signal }) {
-  const directoryPath = path.join(__dirname, '..', '.well-known');
-  let files = [];
-
-  for (let i = 0; i < tools.length; i++) {
-    const filePath = path.join(directoryPath, tools[i] + '.json');
-
-    try {
-      // If the access Promise is resolved, it means that the file exists
-      // Then we can add it to the files array
-      await fs.promises.access(filePath, fs.constants.F_OK);
-      files.push(tools[i] + '.json');
-    } catch (err) {
-      logger.error(`[loadSpecs] File ${tools[i] + '.json'} does not exist`, err);
-    }
-  }
-
-  if (files.length === 0) {
-    files = (await fs.promises.readdir(directoryPath)).filter(
-      (file) => path.extname(file) === '.json',
-    );
-  }
-
-  const validJsons = [];
-  const constructorMap = {};
-
-  logger.debug('[validateJson] files', files);
-
-  for (const file of files) {
-    if (path.extname(file) === '.json') {
-      const filePath = path.join(directoryPath, file);
-      const fileContent = await fs.promises.readFile(filePath, 'utf8');
-      const json = JSON.parse(fileContent);
-
-      if (!validateJson(json)) {
-        logger.debug('[validateJson] Invalid json', json);
-        continue;
-      }
-
-      if (llm && map) {
-        constructorMap[json.name_for_model] = async () =>
-          await createOpenAPIPlugin({
-            data: json,
-            llm,
-            message,
-            memory,
-            signal,
-            user,
-          });
-        continue;
-      }
-
-      if (llm) {
-        validJsons.push(createOpenAPIPlugin({ data: json, llm }));
-        continue;
-      }
-
-      validJsons.push(json);
-    }
-  }
-
-  if (map) {
-    return constructorMap;
-  }
-
-  const plugins = (await Promise.all(validJsons)).filter((plugin) => plugin);
-
-  //   logger.debug('[validateJson] plugins', plugins);
-  //   logger.debug(plugins[0].name);
-
-  return plugins;
-}
-
-module.exports = {
-  loadSpecs,
-  validateJson,
-  ManifestDefinition,
-};

api/app/clients/tools/util/loadSpecs.spec.js

@@ -1,101 +0,0 @@
-const fs = require('fs');
-const { validateJson, loadSpecs, ManifestDefinition } = require('./loadSpecs');
-const { createOpenAPIPlugin } = require('../dynamic/OpenAPIPlugin');
-
-jest.mock('../dynamic/OpenAPIPlugin');
-
-describe('ManifestDefinition', () => {
-  it('should validate correct json', () => {
-    const json = {
-      name_for_human: 'Test',
-      name_for_model: 'Test',
-      description_for_human: 'Test',
-      description_for_model: 'Test',
-      api: {
-        url: 'http://test.com',
-      },
-    };
-
-    expect(() => ManifestDefinition.parse(json)).not.toThrow();
-  });
-
-  it('should not validate incorrect json', () => {
-    const json = {
-      name_for_human: 'Test',
-      name_for_model: 'Test',
-      description_for_human: 'Test',
-      description_for_model: 'Test',
-      api: {
-        url: 123, // incorrect type
-      },
-    };
-
-    expect(() => ManifestDefinition.parse(json)).toThrow();
-  });
-});
-
-describe('validateJson', () => {
-  it('should return parsed json if valid', () => {
-    const json = {
-      name_for_human: 'Test',
-      name_for_model: 'Test',
-      description_for_human: 'Test',
-      description_for_model: 'Test',
-      api: {
-        url: 'http://test.com',
-      },
-    };
-
-    expect(validateJson(json)).toEqual(json);
-  });
-
-  it('should return false if json is not valid', () => {
-    const json = {
-      name_for_human: 'Test',
-      name_for_model: 'Test',
-      description_for_human: 'Test',
-      description_for_model: 'Test',
-      api: {
-        url: 123, // incorrect type
-      },
-    };
-
-    expect(validateJson(json)).toEqual(false);
-  });
-});
-
-describe('loadSpecs', () => {
-  beforeEach(() => {
-    jest.spyOn(fs.promises, 'readdir').mockResolvedValue(['test.json']);
-    jest.spyOn(fs.promises, 'readFile').mockResolvedValue(
-      JSON.stringify({
-        name_for_human: 'Test',
-        name_for_model: 'Test',
-        description_for_human: 'Test',
-        description_for_model: 'Test',
-        api: {
-          url: 'http://test.com',
-        },
-      }),
-    );
-    createOpenAPIPlugin.mockResolvedValue({});
-  });
-
-  afterEach(() => {
-    jest.restoreAllMocks();
-  });
-
-  it('should return plugins', async () => {
-    const plugins = await loadSpecs({ llm: true, verbose: false });
-
-    expect(plugins).toHaveLength(1);
-    expect(createOpenAPIPlugin).toHaveBeenCalledTimes(1);
-  });
-
-  it('should return constructorMap if map is true', async () => {
-    const plugins = await loadSpecs({ llm: {}, map: true, verbose: false });
-
-    expect(plugins).toHaveProperty('Test');
-    expect(createOpenAPIPlugin).not.toHaveBeenCalled();
-  });
-});

api/cache/banViolation.js

@@ -1,8 +1,9 @@
+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 { logger } = require('~/config');
 
 const { BAN_VIOLATIONS, BAN_INTERVAL } = process.env ?? {};
 const interval = math(BAN_INTERVAL, 20);
@@ -32,7 +33,6 @@ const banViolation = async (req, res, errorMessage) => {
   if (!isEnabled(BAN_VIOLATIONS)) {
     return;
   }
-
   if (!errorMessage) {
     return;
   }
@@ -51,7 +51,6 @@ const banViolation = async (req, res, errorMessage) => {
 
   const banLogs = getLogStores(ViolationTypes.BAN);
   const duration = errorMessage.duration || banLogs.opts.ttl;
-
   if (duration <= 0) {
     return;
   }

api/cache/banViolation.spec.js

@@ -1,48 +1,28 @@
+const mongoose = require('mongoose');
+const { MongoMemoryServer } = require('mongodb-memory-server');
 const banViolation = require('./banViolation');
 
-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),
-    });
-  });
-});
+// Mock deleteAllUserSessions since we're testing ban logic, not session deletion
+jest.mock('~/models', () => ({
+  ...jest.requireActual('~/models'),
+  deleteAllUserSessions: jest.fn().mockResolvedValue(true),
+}));
 
 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',
@@ -55,7 +35,7 @@ describe('banViolation', () => {
     };
     errorMessage = {
       type: 'someViolation',
-      user_id: '12345',
+      user_id: new mongoose.Types.ObjectId().toString(), // Use valid ObjectId
       prev_count: 0,
       violation_count: 0,
     };

api/cache/cacheConfig.js

@@ -0,0 +1,89 @@
+const fs = require('fs');
+const { logger } = require('@librechat/data-schemas');
+const { math, isEnabled } = require('@librechat/api');
+const { CacheKeys } = require('librechat-data-provider');
+
+// To ensure that different deployments do not interfere with each other's cache, we use a prefix for the Redis keys.
+// This prefix is usually the deployment ID, which is often passed to the container or pod as an env var.
+// Set REDIS_KEY_PREFIX_VAR to the env var that contains the deployment ID.
+const REDIS_KEY_PREFIX_VAR = process.env.REDIS_KEY_PREFIX_VAR;
+const REDIS_KEY_PREFIX = process.env.REDIS_KEY_PREFIX;
+if (REDIS_KEY_PREFIX_VAR && REDIS_KEY_PREFIX) {
+  throw new Error('Only either REDIS_KEY_PREFIX_VAR or REDIS_KEY_PREFIX can be set.');
+}
+
+const USE_REDIS = isEnabled(process.env.USE_REDIS);
+if (USE_REDIS && !process.env.REDIS_URI) {
+  throw new Error('USE_REDIS is enabled but REDIS_URI is not set.');
+}
+
+// Comma-separated list of cache namespaces that should be forced to use in-memory storage
+// even when Redis is enabled. This allows selective performance optimization for specific caches.
+const FORCED_IN_MEMORY_CACHE_NAMESPACES = process.env.FORCED_IN_MEMORY_CACHE_NAMESPACES
+  ? process.env.FORCED_IN_MEMORY_CACHE_NAMESPACES.split(',').map((key) => key.trim())
+  : [];
+
+// Validate against CacheKeys enum
+if (FORCED_IN_MEMORY_CACHE_NAMESPACES.length > 0) {
+  const validKeys = Object.values(CacheKeys);
+  const invalidKeys = FORCED_IN_MEMORY_CACHE_NAMESPACES.filter((key) => !validKeys.includes(key));
+
+  if (invalidKeys.length > 0) {
+    throw new Error(
+      `Invalid cache keys in FORCED_IN_MEMORY_CACHE_NAMESPACES: ${invalidKeys.join(', ')}. Valid keys: ${validKeys.join(', ')}`,
+    );
+  }
+}
+
+/** Helper function to safely read Redis CA certificate from file
+ * @returns {string|null} The contents of the CA certificate file, or null if not set or on error
+ */
+const getRedisCA = () => {
+  const caPath = process.env.REDIS_CA;
+  if (!caPath) {
+    return null;
+  }
+
+  try {
+    if (fs.existsSync(caPath)) {
+      return fs.readFileSync(caPath, 'utf8');
+    } else {
+      logger.warn(`Redis CA certificate file not found: ${caPath}`);
+      return null;
+    }
+  } catch (error) {
+    logger.error(`Failed to read Redis CA certificate file '${caPath}':`, error);
+    return null;
+  }
+};
+
+const cacheConfig = {
+  FORCED_IN_MEMORY_CACHE_NAMESPACES,
+  USE_REDIS,
+  REDIS_URI: process.env.REDIS_URI,
+  REDIS_USERNAME: process.env.REDIS_USERNAME,
+  REDIS_PASSWORD: process.env.REDIS_PASSWORD,
+  REDIS_CA: getRedisCA(),
+  REDIS_KEY_PREFIX: process.env[REDIS_KEY_PREFIX_VAR] || REDIS_KEY_PREFIX || '',
+  REDIS_MAX_LISTENERS: math(process.env.REDIS_MAX_LISTENERS, 40),
+  REDIS_PING_INTERVAL: math(process.env.REDIS_PING_INTERVAL, 0),
+  /** Max delay between reconnection attempts in ms */
+  REDIS_RETRY_MAX_DELAY: math(process.env.REDIS_RETRY_MAX_DELAY, 3000),
+  /** Max number of reconnection attempts (0 = infinite) */
+  REDIS_RETRY_MAX_ATTEMPTS: math(process.env.REDIS_RETRY_MAX_ATTEMPTS, 10),
+  /** Connection timeout in ms */
+  REDIS_CONNECT_TIMEOUT: math(process.env.REDIS_CONNECT_TIMEOUT, 10000),
+  /** Queue commands when disconnected */
+  REDIS_ENABLE_OFFLINE_QUEUE: isEnabled(process.env.REDIS_ENABLE_OFFLINE_QUEUE ?? 'true'),
+  /** flag to modify redis connection by adding dnsLookup this is required when connecting to elasticache for ioredis
+   * see "Special Note: Aws Elasticache Clusters with TLS" on this webpage:  https://www.npmjs.com/package/ioredis **/
+  REDIS_USE_ALTERNATIVE_DNS_LOOKUP: isEnabled(process.env.REDIS_USE_ALTERNATIVE_DNS_LOOKUP),
+  /** Enable redis cluster without the need of multiple URIs */
+  USE_REDIS_CLUSTER: isEnabled(process.env.USE_REDIS_CLUSTER ?? 'false'),
+  CI: isEnabled(process.env.CI),
+  DEBUG_MEMORY_CACHE: isEnabled(process.env.DEBUG_MEMORY_CACHE),
+
+  BAN_DURATION: math(process.env.BAN_DURATION, 7200000), // 2 hours
+};
+
+module.exports = { cacheConfig };

api/cache/cacheConfig.spec.js

@@ -0,0 +1,189 @@
+const fs = require('fs');
+
+describe('cacheConfig', () => {
+  let originalEnv;
+  let originalReadFileSync;
+
+  beforeEach(() => {
+    originalEnv = { ...process.env };
+    originalReadFileSync = fs.readFileSync;
+
+    // Clear all related env vars first
+    delete process.env.REDIS_URI;
+    delete process.env.REDIS_CA;
+    delete process.env.REDIS_KEY_PREFIX_VAR;
+    delete process.env.REDIS_KEY_PREFIX;
+    delete process.env.USE_REDIS;
+    delete process.env.USE_REDIS_CLUSTER;
+    delete process.env.REDIS_PING_INTERVAL;
+    delete process.env.FORCED_IN_MEMORY_CACHE_NAMESPACES;
+
+    // Clear require cache
+    jest.resetModules();
+  });
+
+  afterEach(() => {
+    process.env = originalEnv;
+    fs.readFileSync = originalReadFileSync;
+    jest.resetModules();
+  });
+
+  describe('REDIS_KEY_PREFIX validation and resolution', () => {
+    test('should throw error when both REDIS_KEY_PREFIX_VAR and REDIS_KEY_PREFIX are set', () => {
+      process.env.REDIS_KEY_PREFIX_VAR = 'DEPLOYMENT_ID';
+      process.env.REDIS_KEY_PREFIX = 'manual-prefix';
+
+      expect(() => {
+        require('./cacheConfig');
+      }).toThrow('Only either REDIS_KEY_PREFIX_VAR or REDIS_KEY_PREFIX can be set.');
+    });
+
+    test('should resolve REDIS_KEY_PREFIX from variable reference', () => {
+      process.env.REDIS_KEY_PREFIX_VAR = 'DEPLOYMENT_ID';
+      process.env.DEPLOYMENT_ID = 'test-deployment-123';
+
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.REDIS_KEY_PREFIX).toBe('test-deployment-123');
+    });
+
+    test('should use direct REDIS_KEY_PREFIX value', () => {
+      process.env.REDIS_KEY_PREFIX = 'direct-prefix';
+
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.REDIS_KEY_PREFIX).toBe('direct-prefix');
+    });
+
+    test('should default to empty string when no prefix is configured', () => {
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.REDIS_KEY_PREFIX).toBe('');
+    });
+
+    test('should handle empty variable reference', () => {
+      process.env.REDIS_KEY_PREFIX_VAR = 'EMPTY_VAR';
+      process.env.EMPTY_VAR = '';
+
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.REDIS_KEY_PREFIX).toBe('');
+    });
+
+    test('should handle undefined variable reference', () => {
+      process.env.REDIS_KEY_PREFIX_VAR = 'UNDEFINED_VAR';
+
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.REDIS_KEY_PREFIX).toBe('');
+    });
+  });
+
+  describe('USE_REDIS and REDIS_URI validation', () => {
+    test('should throw error when USE_REDIS is enabled but REDIS_URI is not set', () => {
+      process.env.USE_REDIS = 'true';
+
+      expect(() => {
+        require('./cacheConfig');
+      }).toThrow('USE_REDIS is enabled but REDIS_URI is not set.');
+    });
+
+    test('should not throw error when USE_REDIS is enabled and REDIS_URI is set', () => {
+      process.env.USE_REDIS = 'true';
+      process.env.REDIS_URI = 'redis://localhost:6379';
+
+      expect(() => {
+        require('./cacheConfig');
+      }).not.toThrow();
+    });
+
+    test('should handle empty REDIS_URI when USE_REDIS is enabled', () => {
+      process.env.USE_REDIS = 'true';
+      process.env.REDIS_URI = '';
+
+      expect(() => {
+        require('./cacheConfig');
+      }).toThrow('USE_REDIS is enabled but REDIS_URI is not set.');
+    });
+  });
+
+  describe('USE_REDIS_CLUSTER configuration', () => {
+    test('should default to false when USE_REDIS_CLUSTER is not set', () => {
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.USE_REDIS_CLUSTER).toBe(false);
+    });
+
+    test('should be false when USE_REDIS_CLUSTER is set to false', () => {
+      process.env.USE_REDIS_CLUSTER = 'false';
+
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.USE_REDIS_CLUSTER).toBe(false);
+    });
+
+    test('should be true when USE_REDIS_CLUSTER is set to true', () => {
+      process.env.USE_REDIS_CLUSTER = 'true';
+
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.USE_REDIS_CLUSTER).toBe(true);
+    });
+
+    test('should work with USE_REDIS enabled and REDIS_URI set', () => {
+      process.env.USE_REDIS_CLUSTER = 'true';
+      process.env.USE_REDIS = 'true';
+      process.env.REDIS_URI = 'redis://localhost:6379';
+
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.USE_REDIS_CLUSTER).toBe(true);
+      expect(cacheConfig.USE_REDIS).toBe(true);
+      expect(cacheConfig.REDIS_URI).toBe('redis://localhost:6379');
+    });
+  });
+
+  describe('REDIS_CA file reading', () => {
+    test('should be null when REDIS_CA is not set', () => {
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.REDIS_CA).toBeNull();
+    });
+  });
+
+  describe('REDIS_PING_INTERVAL configuration', () => {
+    test('should default to 0 when REDIS_PING_INTERVAL is not set', () => {
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.REDIS_PING_INTERVAL).toBe(0);
+    });
+
+    test('should use provided REDIS_PING_INTERVAL value', () => {
+      process.env.REDIS_PING_INTERVAL = '300';
+
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.REDIS_PING_INTERVAL).toBe(300);
+    });
+  });
+
+  describe('FORCED_IN_MEMORY_CACHE_NAMESPACES validation', () => {
+    test('should parse comma-separated cache keys correctly', () => {
+      process.env.FORCED_IN_MEMORY_CACHE_NAMESPACES = ' ROLES, MESSAGES ';
+
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES).toEqual([
+        'ROLES',
+        'MESSAGES',
+      ]);
+    });
+
+    test('should throw error for invalid cache keys', () => {
+      process.env.FORCED_IN_MEMORY_CACHE_NAMESPACES = 'INVALID_KEY,ROLES';
+
+      expect(() => {
+        require('./cacheConfig');
+      }).toThrow('Invalid cache keys in FORCED_IN_MEMORY_CACHE_NAMESPACES: INVALID_KEY');
+    });
+
+    test('should handle empty string gracefully', () => {
+      process.env.FORCED_IN_MEMORY_CACHE_NAMESPACES = '';
+
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES).toEqual([]);
+    });
+
+    test('should handle undefined env var gracefully', () => {
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES).toEqual([]);
+    });
+  });
+});

api/cache/cacheFactory.js

@@ -0,0 +1,108 @@
+const KeyvRedis = require('@keyv/redis').default;
+const { Keyv } = require('keyv');
+const { RedisStore } = require('rate-limit-redis');
+const { Time } = require('librechat-data-provider');
+const { logger } = require('@librechat/data-schemas');
+const { RedisStore: ConnectRedis } = require('connect-redis');
+const MemoryStore = require('memorystore')(require('express-session'));
+const { keyvRedisClient, ioredisClient, GLOBAL_PREFIX_SEPARATOR } = require('./redisClients');
+const { cacheConfig } = require('./cacheConfig');
+const { violationFile } = require('./keyvFiles');
+
+/**
+ * Creates a cache instance using Redis or a fallback store. Suitable for general caching needs.
+ * @param {string} namespace - The cache namespace.
+ * @param {number} [ttl] - Time to live for cache entries.
+ * @param {object} [fallbackStore] - Optional fallback store if Redis is not used.
+ * @returns {Keyv} Cache instance.
+ */
+const standardCache = (namespace, ttl = undefined, fallbackStore = undefined) => {
+  if (
+    cacheConfig.USE_REDIS &&
+    !cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES?.includes(namespace)
+  ) {
+    try {
+      const keyvRedis = new KeyvRedis(keyvRedisClient);
+      const cache = new Keyv(keyvRedis, { namespace, ttl });
+      keyvRedis.namespace = cacheConfig.REDIS_KEY_PREFIX;
+      keyvRedis.keyPrefixSeparator = GLOBAL_PREFIX_SEPARATOR;
+
+      cache.on('error', (err) => {
+        logger.error(`Cache error in namespace ${namespace}:`, err);
+      });
+
+      return cache;
+    } catch (err) {
+      logger.error(`Failed to create Redis cache for namespace ${namespace}:`, err);
+      throw err;
+    }
+  }
+  if (fallbackStore) return new Keyv({ store: fallbackStore, namespace, ttl });
+  return new Keyv({ namespace, ttl });
+};
+
+/**
+ * Creates a cache instance for storing violation data.
+ * Uses a file-based fallback store if Redis is not enabled.
+ * @param {string} namespace - The cache namespace for violations.
+ * @param {number} [ttl] - Time to live for cache entries.
+ * @returns {Keyv} Cache instance for violations.
+ */
+const violationCache = (namespace, ttl = undefined) => {
+  return standardCache(`violations:${namespace}`, ttl, violationFile);
+};
+
+/**
+ * Creates a session cache instance using Redis or in-memory store.
+ * @param {string} namespace - The session namespace.
+ * @param {number} [ttl] - Time to live for session entries.
+ * @returns {MemoryStore | ConnectRedis} Session store instance.
+ */
+const sessionCache = (namespace, ttl = undefined) => {
+  namespace = namespace.endsWith(':') ? namespace : `${namespace}:`;
+  if (!cacheConfig.USE_REDIS) return new MemoryStore({ ttl, checkPeriod: Time.ONE_DAY });
+  const store = new ConnectRedis({ client: ioredisClient, ttl, prefix: namespace });
+  if (ioredisClient) {
+    ioredisClient.on('error', (err) => {
+      logger.error(`Session store Redis error for namespace ${namespace}:`, err);
+    });
+  }
+  return store;
+};
+
+/**
+ * Creates a rate limiter cache using Redis.
+ * @param {string} prefix - The key prefix for rate limiting.
+ * @returns {RedisStore|undefined} RedisStore instance or undefined if Redis is not used.
+ */
+const limiterCache = (prefix) => {
+  if (!prefix) throw new Error('prefix is required');
+  if (!cacheConfig.USE_REDIS) return undefined;
+  prefix = prefix.endsWith(':') ? prefix : `${prefix}:`;
+
+  try {
+    if (!ioredisClient) {
+      logger.warn(`Redis client not available for rate limiter with prefix ${prefix}`);
+      return undefined;
+    }
+
+    return new RedisStore({ sendCommand, prefix });
+  } catch (err) {
+    logger.error(`Failed to create Redis rate limiter for prefix ${prefix}:`, err);
+    return undefined;
+  }
+};
+
+const sendCommand = (...args) => {
+  if (!ioredisClient) {
+    logger.warn('Redis client not available for command execution');
+    return Promise.reject(new Error('Redis client not available'));
+  }
+
+  return ioredisClient.call(...args).catch((err) => {
+    logger.error('Redis command execution failed:', err);
+    throw err;
+  });
+};
+
+module.exports = { standardCache, sessionCache, violationCache, limiterCache };

api/cache/cacheFactory.spec.js

@@ -0,0 +1,432 @@
+const { Time } = require('librechat-data-provider');
+
+// Mock dependencies first
+const mockKeyvRedis = {
+  namespace: '',
+  keyPrefixSeparator: '',
+};
+
+const mockKeyv = jest.fn().mockReturnValue({
+  mock: 'keyv',
+  on: jest.fn(),
+});
+const mockConnectRedis = jest.fn().mockReturnValue({ mock: 'connectRedis' });
+const mockMemoryStore = jest.fn().mockReturnValue({ mock: 'memoryStore' });
+const mockRedisStore = jest.fn().mockReturnValue({ mock: 'redisStore' });
+
+const mockIoredisClient = {
+  call: jest.fn(),
+  on: jest.fn(),
+};
+
+const mockKeyvRedisClient = {};
+const mockViolationFile = {};
+
+// Mock modules before requiring the main module
+jest.mock('@keyv/redis', () => ({
+  default: jest.fn().mockImplementation(() => mockKeyvRedis),
+}));
+
+jest.mock('keyv', () => ({
+  Keyv: mockKeyv,
+}));
+
+jest.mock('./cacheConfig', () => ({
+  cacheConfig: {
+    USE_REDIS: false,
+    REDIS_KEY_PREFIX: 'test',
+    FORCED_IN_MEMORY_CACHE_NAMESPACES: [],
+  },
+}));
+
+jest.mock('./redisClients', () => ({
+  keyvRedisClient: mockKeyvRedisClient,
+  ioredisClient: mockIoredisClient,
+  GLOBAL_PREFIX_SEPARATOR: '::',
+}));
+
+jest.mock('./keyvFiles', () => ({
+  violationFile: mockViolationFile,
+}));
+
+jest.mock('connect-redis', () => ({ RedisStore: mockConnectRedis }));
+
+jest.mock('memorystore', () => jest.fn(() => mockMemoryStore));
+
+jest.mock('rate-limit-redis', () => ({
+  RedisStore: mockRedisStore,
+}));
+
+jest.mock('@librechat/data-schemas', () => ({
+  logger: {
+    error: jest.fn(),
+    warn: jest.fn(),
+    info: jest.fn(),
+  },
+}));
+
+// Import after mocking
+const { standardCache, sessionCache, violationCache, limiterCache } = require('./cacheFactory');
+const { cacheConfig } = require('./cacheConfig');
+
+describe('cacheFactory', () => {
+  beforeEach(() => {
+    jest.clearAllMocks();
+
+    // Reset cache config mock
+    cacheConfig.USE_REDIS = false;
+    cacheConfig.REDIS_KEY_PREFIX = 'test';
+    cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES = [];
+  });
+
+  describe('redisCache', () => {
+    it('should create Redis cache when USE_REDIS is true', () => {
+      cacheConfig.USE_REDIS = true;
+      const namespace = 'test-namespace';
+      const ttl = 3600;
+
+      standardCache(namespace, ttl);
+
+      expect(require('@keyv/redis').default).toHaveBeenCalledWith(mockKeyvRedisClient);
+      expect(mockKeyv).toHaveBeenCalledWith(mockKeyvRedis, { namespace, ttl });
+      expect(mockKeyvRedis.namespace).toBe(cacheConfig.REDIS_KEY_PREFIX);
+      expect(mockKeyvRedis.keyPrefixSeparator).toBe('::');
+    });
+
+    it('should create Redis cache with undefined ttl when not provided', () => {
+      cacheConfig.USE_REDIS = true;
+      const namespace = 'test-namespace';
+
+      standardCache(namespace);
+
+      expect(mockKeyv).toHaveBeenCalledWith(mockKeyvRedis, { namespace, ttl: undefined });
+    });
+
+    it('should use fallback store when USE_REDIS is false and fallbackStore is provided', () => {
+      cacheConfig.USE_REDIS = false;
+      const namespace = 'test-namespace';
+      const ttl = 3600;
+      const fallbackStore = { some: 'store' };
+
+      standardCache(namespace, ttl, fallbackStore);
+
+      expect(mockKeyv).toHaveBeenCalledWith({ store: fallbackStore, namespace, ttl });
+    });
+
+    it('should create default Keyv instance when USE_REDIS is false and no fallbackStore', () => {
+      cacheConfig.USE_REDIS = false;
+      const namespace = 'test-namespace';
+      const ttl = 3600;
+
+      standardCache(namespace, ttl);
+
+      expect(mockKeyv).toHaveBeenCalledWith({ namespace, ttl });
+    });
+
+    it('should handle namespace and ttl as undefined', () => {
+      cacheConfig.USE_REDIS = false;
+
+      standardCache();
+
+      expect(mockKeyv).toHaveBeenCalledWith({ namespace: undefined, ttl: undefined });
+    });
+
+    it('should use fallback when namespace is in FORCED_IN_MEMORY_CACHE_NAMESPACES', () => {
+      cacheConfig.USE_REDIS = true;
+      cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES = ['forced-memory'];
+      const namespace = 'forced-memory';
+      const ttl = 3600;
+
+      standardCache(namespace, ttl);
+
+      expect(require('@keyv/redis').default).not.toHaveBeenCalled();
+      expect(mockKeyv).toHaveBeenCalledWith({ namespace, ttl });
+    });
+
+    it('should use Redis when namespace is not in FORCED_IN_MEMORY_CACHE_NAMESPACES', () => {
+      cacheConfig.USE_REDIS = true;
+      cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES = ['other-namespace'];
+      const namespace = 'test-namespace';
+      const ttl = 3600;
+
+      standardCache(namespace, ttl);
+
+      expect(require('@keyv/redis').default).toHaveBeenCalledWith(mockKeyvRedisClient);
+      expect(mockKeyv).toHaveBeenCalledWith(mockKeyvRedis, { namespace, ttl });
+    });
+
+    it('should throw error when Redis cache creation fails', () => {
+      cacheConfig.USE_REDIS = true;
+      const namespace = 'test-namespace';
+      const ttl = 3600;
+      const testError = new Error('Redis connection failed');
+
+      const KeyvRedis = require('@keyv/redis').default;
+      KeyvRedis.mockImplementationOnce(() => {
+        throw testError;
+      });
+
+      expect(() => standardCache(namespace, ttl)).toThrow('Redis connection failed');
+
+      const { logger } = require('@librechat/data-schemas');
+      expect(logger.error).toHaveBeenCalledWith(
+        `Failed to create Redis cache for namespace ${namespace}:`,
+        testError,
+      );
+
+      expect(mockKeyv).not.toHaveBeenCalled();
+    });
+  });
+
+  describe('violationCache', () => {
+    it('should create violation cache with prefixed namespace', () => {
+      const namespace = 'test-violations';
+      const ttl = 7200;
+
+      // We can't easily mock the internal redisCache call since it's in the same module
+      // But we can test that the function executes without throwing
+      expect(() => violationCache(namespace, ttl)).not.toThrow();
+    });
+
+    it('should create violation cache with undefined ttl', () => {
+      const namespace = 'test-violations';
+
+      violationCache(namespace);
+
+      // The function should call redisCache with violations: prefixed namespace
+      // Since we can't easily mock the internal redisCache call, we test the behavior
+      expect(() => violationCache(namespace)).not.toThrow();
+    });
+
+    it('should handle undefined namespace', () => {
+      expect(() => violationCache(undefined)).not.toThrow();
+    });
+  });
+
+  describe('sessionCache', () => {
+    it('should return MemoryStore when USE_REDIS is false', () => {
+      cacheConfig.USE_REDIS = false;
+      const namespace = 'sessions';
+      const ttl = 86400;
+
+      const result = sessionCache(namespace, ttl);
+
+      expect(mockMemoryStore).toHaveBeenCalledWith({ ttl, checkPeriod: Time.ONE_DAY });
+      expect(result).toBe(mockMemoryStore());
+    });
+
+    it('should return ConnectRedis when USE_REDIS is true', () => {
+      cacheConfig.USE_REDIS = true;
+      const namespace = 'sessions';
+      const ttl = 86400;
+
+      const result = sessionCache(namespace, ttl);
+
+      expect(mockConnectRedis).toHaveBeenCalledWith({
+        client: mockIoredisClient,
+        ttl,
+        prefix: `${namespace}:`,
+      });
+      expect(result).toBe(mockConnectRedis());
+    });
+
+    it('should add colon to namespace if not present', () => {
+      cacheConfig.USE_REDIS = true;
+      const namespace = 'sessions';
+
+      sessionCache(namespace);
+
+      expect(mockConnectRedis).toHaveBeenCalledWith({
+        client: mockIoredisClient,
+        ttl: undefined,
+        prefix: 'sessions:',
+      });
+    });
+
+    it('should not add colon to namespace if already present', () => {
+      cacheConfig.USE_REDIS = true;
+      const namespace = 'sessions:';
+
+      sessionCache(namespace);
+
+      expect(mockConnectRedis).toHaveBeenCalledWith({
+        client: mockIoredisClient,
+        ttl: undefined,
+        prefix: 'sessions:',
+      });
+    });
+
+    it('should handle undefined ttl', () => {
+      cacheConfig.USE_REDIS = false;
+      const namespace = 'sessions';
+
+      sessionCache(namespace);
+
+      expect(mockMemoryStore).toHaveBeenCalledWith({
+        ttl: undefined,
+        checkPeriod: Time.ONE_DAY,
+      });
+    });
+
+    it('should throw error when ConnectRedis constructor fails', () => {
+      cacheConfig.USE_REDIS = true;
+      const namespace = 'sessions';
+      const ttl = 86400;
+
+      // Mock ConnectRedis to throw an error during construction
+      const redisError = new Error('Redis connection failed');
+      mockConnectRedis.mockImplementationOnce(() => {
+        throw redisError;
+      });
+
+      // The error should propagate up, not be caught
+      expect(() => sessionCache(namespace, ttl)).toThrow('Redis connection failed');
+
+      // Verify that MemoryStore was NOT used as fallback
+      expect(mockMemoryStore).not.toHaveBeenCalled();
+    });
+
+    it('should register error handler but let errors propagate to Express', () => {
+      cacheConfig.USE_REDIS = true;
+      const namespace = 'sessions';
+
+      // Create a mock session store with middleware methods
+      const mockSessionStore = {
+        get: jest.fn(),
+        set: jest.fn(),
+        destroy: jest.fn(),
+      };
+      mockConnectRedis.mockReturnValue(mockSessionStore);
+
+      const store = sessionCache(namespace);
+
+      // Verify error handler was registered
+      expect(mockIoredisClient.on).toHaveBeenCalledWith('error', expect.any(Function));
+
+      // Get the error handler
+      const errorHandler = mockIoredisClient.on.mock.calls.find((call) => call[0] === 'error')[1];
+
+      // Simulate an error from Redis during a session operation
+      const redisError = new Error('Socket closed unexpectedly');
+
+      // The error handler should log but not swallow the error
+      const { logger } = require('@librechat/data-schemas');
+      errorHandler(redisError);
+
+      expect(logger.error).toHaveBeenCalledWith(
+        `Session store Redis error for namespace ${namespace}::`,
+        redisError,
+      );
+
+      // Now simulate what happens when session middleware tries to use the store
+      const callback = jest.fn();
+      mockSessionStore.get.mockImplementation((sid, cb) => {
+        cb(new Error('Redis connection lost'));
+      });
+
+      // Call the store's get method (as Express session would)
+      store.get('test-session-id', callback);
+
+      // The error should be passed to the callback, not swallowed
+      expect(callback).toHaveBeenCalledWith(new Error('Redis connection lost'));
+    });
+
+    it('should handle null ioredisClient gracefully', () => {
+      cacheConfig.USE_REDIS = true;
+      const namespace = 'sessions';
+
+      // Temporarily set ioredisClient to null (simulating connection not established)
+      const originalClient = require('./redisClients').ioredisClient;
+      require('./redisClients').ioredisClient = null;
+
+      // ConnectRedis might accept null client but would fail on first use
+      // The important thing is it doesn't throw uncaught exceptions during construction
+      const store = sessionCache(namespace);
+      expect(store).toBeDefined();
+
+      // Restore original client
+      require('./redisClients').ioredisClient = originalClient;
+    });
+  });
+
+  describe('limiterCache', () => {
+    it('should return undefined when USE_REDIS is false', () => {
+      cacheConfig.USE_REDIS = false;
+      const result = limiterCache('prefix');
+
+      expect(result).toBeUndefined();
+    });
+
+    it('should return RedisStore when USE_REDIS is true', () => {
+      cacheConfig.USE_REDIS = true;
+      const result = limiterCache('rate-limit');
+
+      expect(mockRedisStore).toHaveBeenCalledWith({
+        sendCommand: expect.any(Function),
+        prefix: `rate-limit:`,
+      });
+      expect(result).toBe(mockRedisStore());
+    });
+
+    it('should add colon to prefix if not present', () => {
+      cacheConfig.USE_REDIS = true;
+      limiterCache('rate-limit');
+
+      expect(mockRedisStore).toHaveBeenCalledWith({
+        sendCommand: expect.any(Function),
+        prefix: 'rate-limit:',
+      });
+    });
+
+    it('should not add colon to prefix if already present', () => {
+      cacheConfig.USE_REDIS = true;
+      limiterCache('rate-limit:');
+
+      expect(mockRedisStore).toHaveBeenCalledWith({
+        sendCommand: expect.any(Function),
+        prefix: 'rate-limit:',
+      });
+    });
+
+    it('should pass sendCommand function that calls ioredisClient.call', async () => {
+      cacheConfig.USE_REDIS = true;
+      mockIoredisClient.call.mockResolvedValue('test-value');
+
+      limiterCache('rate-limit');
+
+      const sendCommandCall = mockRedisStore.mock.calls[0][0];
+      const sendCommand = sendCommandCall.sendCommand;
+
+      // Test that sendCommand properly delegates to ioredisClient.call
+      const args = ['GET', 'test-key'];
+      const result = await sendCommand(...args);
+
+      expect(mockIoredisClient.call).toHaveBeenCalledWith(...args);
+      expect(result).toBe('test-value');
+    });
+
+    it('should handle sendCommand errors properly', async () => {
+      cacheConfig.USE_REDIS = true;
+
+      // Mock the call method to reject with an error
+      const testError = new Error('Redis error');
+      mockIoredisClient.call.mockRejectedValue(testError);
+
+      limiterCache('rate-limit');
+
+      const sendCommandCall = mockRedisStore.mock.calls[0][0];
+      const sendCommand = sendCommandCall.sendCommand;
+
+      // Test that sendCommand properly handles errors
+      const args = ['GET', 'test-key'];
+
+      await expect(sendCommand(...args)).rejects.toThrow('Redis error');
+      expect(mockIoredisClient.call).toHaveBeenCalledWith(...args);
+    });
+
+    it('should handle undefined prefix', () => {
+      cacheConfig.USE_REDIS = true;
+      expect(() => limiterCache()).toThrow('prefix is required');
+    });
+  });
+});

api/cache/clearPendingReq.js

@@ -1,5 +1,5 @@
+const { isEnabled } = require('@librechat/api');
 const { Time, CacheKeys } = require('librechat-data-provider');
-const { isEnabled } = require('~/server/utils');
 const getLogStores = require('./getLogStores');
 
 const { USE_REDIS, LIMIT_CONCURRENT_MESSAGES } = process.env ?? {};

api/cache/getLogStores.js

@@ -1,103 +1,52 @@
+const { cacheConfig } = require('./cacheConfig');
 const { Keyv } = require('keyv');
 const { CacheKeys, ViolationTypes, Time } = require('librechat-data-provider');
-const { logFile, violationFile } = require('./keyvFiles');
-const { math, isEnabled } = require('~/server/utils');
-const keyvRedis = require('./keyvRedis');
+const { logFile } = require('./keyvFiles');
 const keyvMongo = require('./keyvMongo');
-
-const { BAN_DURATION, USE_REDIS, DEBUG_MEMORY_CACHE, CI } = process.env ?? {};
-
-const duration = math(BAN_DURATION, 7200000);
-const isRedisEnabled = isEnabled(USE_REDIS);
-const debugMemoryCache = isEnabled(DEBUG_MEMORY_CACHE);
-
-const createViolationInstance = (namespace) => {
-  const config = isRedisEnabled ? { store: keyvRedis } : { store: violationFile, namespace };
-  return new Keyv(config);
-};
-
-// Serve cache from memory so no need to clear it on startup/exit
-const pending_req = isRedisEnabled
-  ? new Keyv({ store: keyvRedis })
-  : new Keyv({ namespace: CacheKeys.PENDING_REQ });
-
-const config = isRedisEnabled
-  ? new Keyv({ store: keyvRedis })
-  : new Keyv({ namespace: CacheKeys.CONFIG_STORE });
-
-const roles = isRedisEnabled
-  ? new Keyv({ store: keyvRedis })
-  : new Keyv({ namespace: CacheKeys.ROLES });
-
-const audioRuns = isRedisEnabled
-  ? new Keyv({ store: keyvRedis, ttl: Time.TEN_MINUTES })
-  : new Keyv({ namespace: CacheKeys.AUDIO_RUNS, ttl: Time.TEN_MINUTES });
-
-const messages = isRedisEnabled
-  ? new Keyv({ store: keyvRedis, ttl: Time.ONE_MINUTE })
-  : new Keyv({ namespace: CacheKeys.MESSAGES, ttl: Time.ONE_MINUTE });
-
-const flows = isRedisEnabled
-  ? new Keyv({ store: keyvRedis, ttl: Time.TWO_MINUTES })
-  : new Keyv({ namespace: CacheKeys.FLOWS, ttl: Time.ONE_MINUTE * 3 });
-
-const tokenConfig = isRedisEnabled
-  ? new Keyv({ store: keyvRedis, ttl: Time.THIRTY_MINUTES })
-  : new Keyv({ namespace: CacheKeys.TOKEN_CONFIG, ttl: Time.THIRTY_MINUTES });
-
-const genTitle = isRedisEnabled
-  ? new Keyv({ store: keyvRedis, ttl: Time.TWO_MINUTES })
-  : new Keyv({ namespace: CacheKeys.GEN_TITLE, ttl: Time.TWO_MINUTES });
-
-const s3ExpiryInterval = isRedisEnabled
-  ? new Keyv({ store: keyvRedis, ttl: Time.THIRTY_MINUTES })
-  : new Keyv({ namespace: CacheKeys.S3_EXPIRY_INTERVAL, ttl: Time.THIRTY_MINUTES });
-
-const modelQueries = isEnabled(process.env.USE_REDIS)
-  ? new Keyv({ store: keyvRedis })
-  : new Keyv({ namespace: CacheKeys.MODEL_QUERIES });
-
-const abortKeys = isRedisEnabled
-  ? new Keyv({ store: keyvRedis })
-  : new Keyv({ namespace: CacheKeys.ABORT_KEYS, ttl: Time.TEN_MINUTES });
+const { standardCache, sessionCache, violationCache } = require('./cacheFactory');
 
 const namespaces = {
-  [CacheKeys.ROLES]: roles,
-  [CacheKeys.CONFIG_STORE]: config,
-  [CacheKeys.PENDING_REQ]: pending_req,
-  [ViolationTypes.BAN]: new Keyv({ store: keyvMongo, namespace: CacheKeys.BANS, ttl: duration }),
-  [CacheKeys.ENCODED_DOMAINS]: new Keyv({
+  [ViolationTypes.GENERAL]: new Keyv({ store: logFile, namespace: 'violations' }),
+  [ViolationTypes.LOGINS]: violationCache(ViolationTypes.LOGINS),
+  [ViolationTypes.CONCURRENT]: violationCache(ViolationTypes.CONCURRENT),
+  [ViolationTypes.NON_BROWSER]: violationCache(ViolationTypes.NON_BROWSER),
+  [ViolationTypes.MESSAGE_LIMIT]: violationCache(ViolationTypes.MESSAGE_LIMIT),
+  [ViolationTypes.REGISTRATIONS]: violationCache(ViolationTypes.REGISTRATIONS),
+  [ViolationTypes.TOKEN_BALANCE]: violationCache(ViolationTypes.TOKEN_BALANCE),
+  [ViolationTypes.TTS_LIMIT]: violationCache(ViolationTypes.TTS_LIMIT),
+  [ViolationTypes.STT_LIMIT]: violationCache(ViolationTypes.STT_LIMIT),
+  [ViolationTypes.CONVO_ACCESS]: violationCache(ViolationTypes.CONVO_ACCESS),
+  [ViolationTypes.TOOL_CALL_LIMIT]: violationCache(ViolationTypes.TOOL_CALL_LIMIT),
+  [ViolationTypes.FILE_UPLOAD_LIMIT]: violationCache(ViolationTypes.FILE_UPLOAD_LIMIT),
+  [ViolationTypes.VERIFY_EMAIL_LIMIT]: violationCache(ViolationTypes.VERIFY_EMAIL_LIMIT),
+  [ViolationTypes.RESET_PASSWORD_LIMIT]: violationCache(ViolationTypes.RESET_PASSWORD_LIMIT),
+  [ViolationTypes.ILLEGAL_MODEL_REQUEST]: violationCache(ViolationTypes.ILLEGAL_MODEL_REQUEST),
+  [ViolationTypes.BAN]: new Keyv({
     store: keyvMongo,
-    namespace: CacheKeys.ENCODED_DOMAINS,
-    ttl: 0,
+    namespace: CacheKeys.BANS,
+    ttl: cacheConfig.BAN_DURATION,
   }),
-  general: new Keyv({ store: logFile, namespace: 'violations' }),
-  concurrent: createViolationInstance('concurrent'),
-  non_browser: createViolationInstance('non_browser'),
-  message_limit: createViolationInstance('message_limit'),
-  token_balance: createViolationInstance(ViolationTypes.TOKEN_BALANCE),
-  registrations: createViolationInstance('registrations'),
-  [ViolationTypes.TTS_LIMIT]: createViolationInstance(ViolationTypes.TTS_LIMIT),
-  [ViolationTypes.STT_LIMIT]: createViolationInstance(ViolationTypes.STT_LIMIT),
-  [ViolationTypes.CONVO_ACCESS]: createViolationInstance(ViolationTypes.CONVO_ACCESS),
-  [ViolationTypes.TOOL_CALL_LIMIT]: createViolationInstance(ViolationTypes.TOOL_CALL_LIMIT),
-  [ViolationTypes.FILE_UPLOAD_LIMIT]: createViolationInstance(ViolationTypes.FILE_UPLOAD_LIMIT),
-  [ViolationTypes.VERIFY_EMAIL_LIMIT]: createViolationInstance(ViolationTypes.VERIFY_EMAIL_LIMIT),
-  [ViolationTypes.RESET_PASSWORD_LIMIT]: createViolationInstance(
-    ViolationTypes.RESET_PASSWORD_LIMIT,
+
+  [CacheKeys.OPENID_SESSION]: sessionCache(CacheKeys.OPENID_SESSION),
+  [CacheKeys.SAML_SESSION]: sessionCache(CacheKeys.SAML_SESSION),
+
+  [CacheKeys.ROLES]: standardCache(CacheKeys.ROLES),
+  [CacheKeys.APP_CONFIG]: standardCache(CacheKeys.APP_CONFIG),
+  [CacheKeys.CONFIG_STORE]: standardCache(CacheKeys.CONFIG_STORE),
+  [CacheKeys.PENDING_REQ]: standardCache(CacheKeys.PENDING_REQ),
+  [CacheKeys.ENCODED_DOMAINS]: new Keyv({ store: keyvMongo, namespace: CacheKeys.ENCODED_DOMAINS }),
+  [CacheKeys.ABORT_KEYS]: standardCache(CacheKeys.ABORT_KEYS, Time.TEN_MINUTES),
+  [CacheKeys.TOKEN_CONFIG]: standardCache(CacheKeys.TOKEN_CONFIG, Time.THIRTY_MINUTES),
+  [CacheKeys.GEN_TITLE]: standardCache(CacheKeys.GEN_TITLE, Time.TWO_MINUTES),
+  [CacheKeys.S3_EXPIRY_INTERVAL]: standardCache(CacheKeys.S3_EXPIRY_INTERVAL, Time.THIRTY_MINUTES),
+  [CacheKeys.MODEL_QUERIES]: standardCache(CacheKeys.MODEL_QUERIES),
+  [CacheKeys.AUDIO_RUNS]: standardCache(CacheKeys.AUDIO_RUNS, Time.TEN_MINUTES),
+  [CacheKeys.MESSAGES]: standardCache(CacheKeys.MESSAGES, Time.ONE_MINUTE),
+  [CacheKeys.FLOWS]: standardCache(CacheKeys.FLOWS, Time.ONE_MINUTE * 3),
+  [CacheKeys.OPENID_EXCHANGED_TOKENS]: standardCache(
+    CacheKeys.OPENID_EXCHANGED_TOKENS,
+    Time.TEN_MINUTES,
   ),
-  [ViolationTypes.ILLEGAL_MODEL_REQUEST]: createViolationInstance(
-    ViolationTypes.ILLEGAL_MODEL_REQUEST,
-  ),
-  logins: createViolationInstance('logins'),
-  [CacheKeys.ABORT_KEYS]: abortKeys,
-  [CacheKeys.TOKEN_CONFIG]: tokenConfig,
-  [CacheKeys.GEN_TITLE]: genTitle,
-  [CacheKeys.S3_EXPIRY_INTERVAL]: s3ExpiryInterval,
-  [CacheKeys.MODEL_QUERIES]: modelQueries,
-  [CacheKeys.AUDIO_RUNS]: audioRuns,
-  [CacheKeys.MESSAGES]: messages,
-  [CacheKeys.FLOWS]: flows,
 };
 
 /**
@@ -106,7 +55,10 @@ const namespaces = {
  */
 function getTTLStores() {
   return Object.values(namespaces).filter(
-    (store) => store instanceof Keyv && typeof store.opts?.ttl === 'number' && store.opts.ttl > 0,
+    (store) =>
+      store instanceof Keyv &&
+      parseInt(store.opts?.ttl ?? '0') > 0 &&
+      !store.opts?.store?.constructor?.name?.includes('Redis'), // Only include non-Redis stores
   );
 }
 
@@ -142,18 +94,18 @@ async function clearExpiredFromCache(cache) {
       if (data?.expires && data.expires <= expiryTime) {
         const deleted = await cache.opts.store.delete(key);
         if (!deleted) {
-          debugMemoryCache &&
+          cacheConfig.DEBUG_MEMORY_CACHE &&
             console.warn(`[Cache] Error deleting entry: ${key} from ${cache.opts.namespace}`);
           continue;
         }
         cleared++;
       }
     } catch (error) {
-      debugMemoryCache &&
+      cacheConfig.DEBUG_MEMORY_CACHE &&
         console.log(`[Cache] Error processing entry from ${cache.opts.namespace}:`, error);
       const deleted = await cache.opts.store.delete(key);
       if (!deleted) {
-        debugMemoryCache &&
+        cacheConfig.DEBUG_MEMORY_CACHE &&
           console.warn(`[Cache] Error deleting entry: ${key} from ${cache.opts.namespace}`);
         continue;
       }
@@ -162,7 +114,7 @@ async function clearExpiredFromCache(cache) {
   }
 
   if (cleared > 0) {
-    debugMemoryCache &&
+    cacheConfig.DEBUG_MEMORY_CACHE &&
       console.log(
         `[Cache] Cleared ${cleared} entries older than ${ttl}ms from ${cache.opts.namespace}`,
       );
@@ -203,7 +155,7 @@ async function clearAllExpiredFromCache() {
   }
 }
 
-if (!isRedisEnabled && !isEnabled(CI)) {
+if (!cacheConfig.USE_REDIS && !cacheConfig.CI) {
   /** @type {Set<NodeJS.Timeout>} */
   const cleanupIntervals = new Set();
 
@@ -214,7 +166,7 @@ if (!isRedisEnabled && !isEnabled(CI)) {
 
   cleanupIntervals.add(cleanup);
 
-  if (debugMemoryCache) {
+  if (cacheConfig.DEBUG_MEMORY_CACHE) {
     const monitor = setInterval(() => {
       const ttlStores = getTTLStores();
       const memory = process.memoryUsage();
@@ -235,13 +187,13 @@ if (!isRedisEnabled && !isEnabled(CI)) {
   }
 
   const dispose = () => {
-    debugMemoryCache && console.log('[Cache] Cleaning up and shutting down...');
+    cacheConfig.DEBUG_MEMORY_CACHE && console.log('[Cache] Cleaning up and shutting down...');
     cleanupIntervals.forEach((interval) => clearInterval(interval));
     cleanupIntervals.clear();
 
     // One final cleanup before exit
     clearAllExpiredFromCache().then(() => {
-      debugMemoryCache && console.log('[Cache] Final cleanup completed');
+      cacheConfig.DEBUG_MEMORY_CACHE && console.log('[Cache] Final cleanup completed');
       process.exit(0);
     });
   };

api/cache/ioredisClient.js

@@ -1,92 +0,0 @@
-const fs = require('fs');
-const Redis = require('ioredis');
-const { isEnabled } = require('~/server/utils');
-const logger = require('~/config/winston');
-
-const { REDIS_URI, USE_REDIS, USE_REDIS_CLUSTER, REDIS_CA, REDIS_MAX_LISTENERS } = process.env;
-
-/** @type {import('ioredis').Redis | import('ioredis').Cluster} */
-let ioredisClient;
-const redis_max_listeners = Number(REDIS_MAX_LISTENERS) || 40;
-
-function mapURI(uri) {
-  const regex =
-    /^(?:(?<scheme>\w+):\/\/)?(?:(?<user>[^:@]+)(?::(?<password>[^@]+))?@)?(?<host>[\w.-]+)(?::(?<port>\d{1,5}))?$/;
-  const match = uri.match(regex);
-
-  if (match) {
-    const { scheme, user, password, host, port } = match.groups;
-
-    return {
-      scheme: scheme || 'none',
-      user: user || null,
-      password: password || null,
-      host: host || null,
-      port: port || null,
-    };
-  } else {
-    const parts = uri.split(':');
-    if (parts.length === 2) {
-      return {
-        scheme: 'none',
-        user: null,
-        password: null,
-        host: parts[0],
-        port: parts[1],
-      };
-    }
-
-    return {
-      scheme: 'none',
-      user: null,
-      password: null,
-      host: uri,
-      port: null,
-    };
-  }
-}
-
-if (REDIS_URI && isEnabled(USE_REDIS)) {
-  let redisOptions = null;
-
-  if (REDIS_CA) {
-    const ca = fs.readFileSync(REDIS_CA);
-    redisOptions = { tls: { ca } };
-  }
-
-  if (isEnabled(USE_REDIS_CLUSTER)) {
-    const hosts = REDIS_URI.split(',').map((item) => {
-      var value = mapURI(item);
-
-      return {
-        host: value.host,
-        port: value.port,
-      };
-    });
-    ioredisClient = new Redis.Cluster(hosts, { redisOptions });
-  } else {
-    ioredisClient = new Redis(REDIS_URI, redisOptions);
-  }
-
-  ioredisClient.on('ready', () => {
-    logger.info('IoRedis connection ready');
-  });
-  ioredisClient.on('reconnecting', () => {
-    logger.info('IoRedis connection reconnecting');
-  });
-  ioredisClient.on('end', () => {
-    logger.info('IoRedis connection ended');
-  });
-  ioredisClient.on('close', () => {
-    logger.info('IoRedis connection closed');
-  });
-  ioredisClient.on('error', (err) => logger.error('IoRedis connection error:', err));
-  ioredisClient.setMaxListeners(redis_max_listeners);
-  logger.info(
-    '[Optional] IoRedis initialized for rate limiters. If you have issues, disable Redis or restart the server.',
-  );
-} else {
-  logger.info('[Optional] IoRedis not initialized for rate limiters.');
-}
-
-module.exports = ioredisClient;

api/cache/keyvMongo.js

@@ -2,7 +2,7 @@
 const mongoose = require('mongoose');
 const EventEmitter = require('events');
 const { GridFSBucket } = require('mongodb');
-const { logger } = require('~/config');
+const { logger } = require('@librechat/data-schemas');
 
 const storeMap = new Map();
 

api/cache/keyvRedis.js

@@ -1,106 +0,0 @@
-const fs = require('fs');
-const ioredis = require('ioredis');
-const KeyvRedis = require('@keyv/redis').default;
-const { isEnabled } = require('~/server/utils');
-const logger = require('~/config/winston');
-
-const { REDIS_URI, USE_REDIS, USE_REDIS_CLUSTER, REDIS_CA, REDIS_KEY_PREFIX, REDIS_MAX_LISTENERS } =
-  process.env;
-
-let keyvRedis;
-const redis_prefix = REDIS_KEY_PREFIX || '';
-const redis_max_listeners = Number(REDIS_MAX_LISTENERS) || 40;
-
-function mapURI(uri) {
-  const regex =
-    /^(?:(?<scheme>\w+):\/\/)?(?:(?<user>[^:@]+)(?::(?<password>[^@]+))?@)?(?<host>[\w.-]+)(?::(?<port>\d{1,5}))?$/;
-  const match = uri.match(regex);
-
-  if (match) {
-    const { scheme, user, password, host, port } = match.groups;
-
-    return {
-      scheme: scheme || 'none',
-      user: user || null,
-      password: password || null,
-      host: host || null,
-      port: port || null,
-    };
-  } else {
-    const parts = uri.split(':');
-    if (parts.length === 2) {
-      return {
-        scheme: 'none',
-        user: null,
-        password: null,
-        host: parts[0],
-        port: parts[1],
-      };
-    }
-
-    return {
-      scheme: 'none',
-      user: null,
-      password: null,
-      host: uri,
-      port: null,
-    };
-  }
-}
-
-if (REDIS_URI && isEnabled(USE_REDIS)) {
-  let redisOptions = null;
-  /** @type {import('@keyv/redis').KeyvRedisOptions} */
-  let keyvOpts = {
-    useRedisSets: false,
-    keyPrefix: redis_prefix,
-  };
-
-  if (REDIS_CA) {
-    const ca = fs.readFileSync(REDIS_CA);
-    redisOptions = { tls: { ca } };
-  }
-
-  if (isEnabled(USE_REDIS_CLUSTER)) {
-    const hosts = REDIS_URI.split(',').map((item) => {
-      var value = mapURI(item);
-
-      return {
-        host: value.host,
-        port: value.port,
-      };
-    });
-    const cluster = new ioredis.Cluster(hosts, { redisOptions });
-    keyvRedis = new KeyvRedis(cluster, keyvOpts);
-  } else {
-    keyvRedis = new KeyvRedis(REDIS_URI, keyvOpts);
-  }
-
-  const pingInterval = setInterval(() => {
-    logger.debug('KeyvRedis ping');
-    keyvRedis.client.ping().catch(err => logger.error('Redis keep-alive ping failed:', err));
-  }, 5 * 60 * 1000);
-
-  keyvRedis.on('ready', () => {
-    logger.info('KeyvRedis connection ready');
-  });
-  keyvRedis.on('reconnecting', () => {
-    logger.info('KeyvRedis connection reconnecting');
-  });
-  keyvRedis.on('end', () => {
-    logger.info('KeyvRedis connection ended');
-  });
-  keyvRedis.on('close', () => {
-    clearInterval(pingInterval);
-    logger.info('KeyvRedis connection closed');
-  });
-  keyvRedis.on('error', (err) => logger.error('KeyvRedis connection error:', err));
-  keyvRedis.setMaxListeners(redis_max_listeners);
-  logger.info(
-    '[Optional] Redis initialized. If you have issues, or seeing older values, disable it or flush cache to refresh values.',
-  );
-} else {
-  logger.info('[Optional] Redis not initialized.');
-}
-
-module.exports = keyvRedis;

api/cache/logViolation.js

@@ -1,4 +1,5 @@
-const { isEnabled } = require('~/server/utils');
+const { isEnabled } = require('@librechat/api');
+const { ViolationTypes } = require('librechat-data-provider');
 const getLogStores = require('./getLogStores');
 const banViolation = require('./banViolation');
 
@@ -9,14 +10,14 @@ const banViolation = require('./banViolation');
  * @param {Object} res - Express response object.
  * @param {string} type - The type of violation.
  * @param {Object} errorMessage - The error message to log.
- * @param {number} [score=1] - The severity of the violation. Defaults to 1
+ * @param {number | string} [score=1] - The severity of the violation. Defaults to 1
  */
 const logViolation = async (req, res, type, errorMessage, score = 1) => {
   const userId = req.user?.id ?? req.user?._id;
   if (!userId) {
     return;
   }
-  const logs = getLogStores('general');
+  const logs = getLogStores(ViolationTypes.GENERAL);
   const violationLogs = getLogStores(type);
   const key = isEnabled(process.env.USE_REDIS) ? `${type}:${userId}` : userId;
 

api/cache/redisClients.js

@@ -0,0 +1,210 @@
+const IoRedis = require('ioredis');
+const { logger } = require('@librechat/data-schemas');
+const { createClient, createCluster } = require('@keyv/redis');
+const { cacheConfig } = require('./cacheConfig');
+
+const GLOBAL_PREFIX_SEPARATOR = '::';
+
+const urls = cacheConfig.REDIS_URI?.split(',').map((uri) => new URL(uri));
+const username = urls?.[0].username || cacheConfig.REDIS_USERNAME;
+const password = urls?.[0].password || cacheConfig.REDIS_PASSWORD;
+const ca = cacheConfig.REDIS_CA;
+
+/** @type {import('ioredis').Redis | import('ioredis').Cluster | null} */
+let ioredisClient = null;
+if (cacheConfig.USE_REDIS) {
+  /** @type {import('ioredis').RedisOptions | import('ioredis').ClusterOptions} */
+  const redisOptions = {
+    username: username,
+    password: password,
+    tls: ca ? { ca } : undefined,
+    keyPrefix: `${cacheConfig.REDIS_KEY_PREFIX}${GLOBAL_PREFIX_SEPARATOR}`,
+    maxListeners: cacheConfig.REDIS_MAX_LISTENERS,
+    retryStrategy: (times) => {
+      if (
+        cacheConfig.REDIS_RETRY_MAX_ATTEMPTS > 0 &&
+        times > cacheConfig.REDIS_RETRY_MAX_ATTEMPTS
+      ) {
+        logger.error(
+          `ioredis giving up after ${cacheConfig.REDIS_RETRY_MAX_ATTEMPTS} reconnection attempts`,
+        );
+        return null;
+      }
+      const delay = Math.min(times * 50, cacheConfig.REDIS_RETRY_MAX_DELAY);
+      logger.info(`ioredis reconnecting... attempt ${times}, delay ${delay}ms`);
+      return delay;
+    },
+    reconnectOnError: (err) => {
+      const targetError = 'READONLY';
+      if (err.message.includes(targetError)) {
+        logger.warn('ioredis reconnecting due to READONLY error');
+        return 2; // Return retry delay instead of boolean
+      }
+      return false;
+    },
+    enableOfflineQueue: cacheConfig.REDIS_ENABLE_OFFLINE_QUEUE,
+    connectTimeout: cacheConfig.REDIS_CONNECT_TIMEOUT,
+    maxRetriesPerRequest: 3,
+  };
+
+  ioredisClient =
+    urls.length === 1 && !cacheConfig.USE_REDIS_CLUSTER
+      ? new IoRedis(cacheConfig.REDIS_URI, redisOptions)
+      : new IoRedis.Cluster(
+          urls.map((url) => ({ host: url.hostname, port: parseInt(url.port, 10) || 6379 })),
+          {
+            ...(cacheConfig.REDIS_USE_ALTERNATIVE_DNS_LOOKUP
+              ? { dnsLookup: (address, callback) => callback(null, address) }
+              : {}),
+            redisOptions,
+            clusterRetryStrategy: (times) => {
+              if (
+                cacheConfig.REDIS_RETRY_MAX_ATTEMPTS > 0 &&
+                times > cacheConfig.REDIS_RETRY_MAX_ATTEMPTS
+              ) {
+                logger.error(
+                  `ioredis cluster giving up after ${cacheConfig.REDIS_RETRY_MAX_ATTEMPTS} reconnection attempts`,
+                );
+                return null;
+              }
+              const delay = Math.min(times * 100, cacheConfig.REDIS_RETRY_MAX_DELAY);
+              logger.info(`ioredis cluster reconnecting... attempt ${times}, delay ${delay}ms`);
+              return delay;
+            },
+            enableOfflineQueue: cacheConfig.REDIS_ENABLE_OFFLINE_QUEUE,
+          },
+        );
+
+  ioredisClient.on('error', (err) => {
+    logger.error('ioredis client error:', err);
+  });
+
+  ioredisClient.on('connect', () => {
+    logger.info('ioredis client connected');
+  });
+
+  ioredisClient.on('ready', () => {
+    logger.info('ioredis client ready');
+  });
+
+  ioredisClient.on('reconnecting', (delay) => {
+    logger.info(`ioredis client reconnecting in ${delay}ms`);
+  });
+
+  ioredisClient.on('close', () => {
+    logger.warn('ioredis client connection closed');
+  });
+
+  /** Ping Interval to keep the Redis server connection alive (if enabled) */
+  let pingInterval = null;
+  const clearPingInterval = () => {
+    if (pingInterval) {
+      clearInterval(pingInterval);
+      pingInterval = null;
+    }
+  };
+
+  if (cacheConfig.REDIS_PING_INTERVAL > 0) {
+    pingInterval = setInterval(() => {
+      if (ioredisClient && ioredisClient.status === 'ready') {
+        ioredisClient.ping().catch((err) => {
+          logger.error('ioredis ping failed:', err);
+        });
+      }
+    }, cacheConfig.REDIS_PING_INTERVAL * 1000);
+    ioredisClient.on('close', clearPingInterval);
+    ioredisClient.on('end', clearPingInterval);
+  }
+}
+
+/** @type {import('@keyv/redis').RedisClient | import('@keyv/redis').RedisCluster | null} */
+let keyvRedisClient = null;
+if (cacheConfig.USE_REDIS) {
+  /**
+   * ** WARNING ** Keyv Redis client does not support Prefix like ioredis above.
+   * The prefix feature will be handled by the Keyv-Redis store in cacheFactory.js
+   * @type {import('@keyv/redis').RedisClientOptions | import('@keyv/redis').RedisClusterOptions}
+   */
+  const redisOptions = {
+    username,
+    password,
+    socket: {
+      tls: ca != null,
+      ca,
+      connectTimeout: cacheConfig.REDIS_CONNECT_TIMEOUT,
+      reconnectStrategy: (retries) => {
+        if (
+          cacheConfig.REDIS_RETRY_MAX_ATTEMPTS > 0 &&
+          retries > cacheConfig.REDIS_RETRY_MAX_ATTEMPTS
+        ) {
+          logger.error(
+            `@keyv/redis client giving up after ${cacheConfig.REDIS_RETRY_MAX_ATTEMPTS} reconnection attempts`,
+          );
+          return new Error('Max reconnection attempts reached');
+        }
+        const delay = Math.min(retries * 100, cacheConfig.REDIS_RETRY_MAX_DELAY);
+        logger.info(`@keyv/redis reconnecting... attempt ${retries}, delay ${delay}ms`);
+        return delay;
+      },
+    },
+    disableOfflineQueue: !cacheConfig.REDIS_ENABLE_OFFLINE_QUEUE,
+  };
+
+  keyvRedisClient =
+    urls.length === 1 && !cacheConfig.USE_REDIS_CLUSTER
+      ? createClient({ url: cacheConfig.REDIS_URI, ...redisOptions })
+      : createCluster({
+          rootNodes: urls.map((url) => ({ url: url.href })),
+          defaults: redisOptions,
+        });
+
+  keyvRedisClient.setMaxListeners(cacheConfig.REDIS_MAX_LISTENERS);
+
+  keyvRedisClient.on('error', (err) => {
+    logger.error('@keyv/redis client error:', err);
+  });
+
+  keyvRedisClient.on('connect', () => {
+    logger.info('@keyv/redis client connected');
+  });
+
+  keyvRedisClient.on('ready', () => {
+    logger.info('@keyv/redis client ready');
+  });
+
+  keyvRedisClient.on('reconnecting', () => {
+    logger.info('@keyv/redis client reconnecting...');
+  });
+
+  keyvRedisClient.on('disconnect', () => {
+    logger.warn('@keyv/redis client disconnected');
+  });
+
+  keyvRedisClient.connect().catch((err) => {
+    logger.error('@keyv/redis initial connection failed:', err);
+    throw err;
+  });
+
+  /** Ping Interval to keep the Redis server connection alive (if enabled) */
+  let pingInterval = null;
+  const clearPingInterval = () => {
+    if (pingInterval) {
+      clearInterval(pingInterval);
+      pingInterval = null;
+    }
+  };
+
+  if (cacheConfig.REDIS_PING_INTERVAL > 0) {
+    pingInterval = setInterval(() => {
+      if (keyvRedisClient && keyvRedisClient.isReady) {
+        keyvRedisClient.ping().catch((err) => {
+          logger.error('@keyv/redis ping failed:', err);
+        });
+      }
+    }, cacheConfig.REDIS_PING_INTERVAL * 1000);
+    keyvRedisClient.on('disconnect', clearPingInterval);
+    keyvRedisClient.on('end', clearPingInterval);
+  }
+}
+
+module.exports = { ioredisClient, keyvRedisClient, GLOBAL_PREFIX_SEPARATOR };

api/config/index.js

@@ -1,28 +1,13 @@
-const axios = require('axios');
 const { EventSource } = require('eventsource');
-const { Time, CacheKeys } = require('librechat-data-provider');
-const { MCPManager, FlowStateManager } = require('librechat-mcp');
+const { Time } = require('librechat-data-provider');
+const { MCPManager, FlowStateManager, OAuthReconnectionManager } = require('@librechat/api');
 const logger = require('./winston');
 
 global.EventSource = EventSource;
 
 /** @type {MCPManager} */
-let mcpManager = null;
 let flowManager = null;
 
-/**
- * @param {string} [userId] - Optional user ID, to avoid disconnecting the current user.
- * @returns {MCPManager}
- */
-function getMCPManager(userId) {
-  if (!mcpManager) {
-    mcpManager = MCPManager.getInstance(logger);
-  } else {
-    mcpManager.checkIdleConnections(userId);
-  }
-  return mcpManager;
-}
-
 /**
  * @param {Keyv} flowsCache
  * @returns {FlowStateManager}
@@ -31,66 +16,16 @@ 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,
+  createMCPManager: MCPManager.createInstance,
+  getMCPManager: MCPManager.getInstance,
   getFlowStateManager,
+  createOAuthReconnectionManager: OAuthReconnectionManager.createInstance,
+  getOAuthReconnectionManager: OAuthReconnectionManager.getInstance,
 };

api/db/connect.js

@@ -0,0 +1,79 @@
+require('dotenv').config();
+const { isEnabled } = require('@librechat/api');
+const { logger } = require('@librechat/data-schemas');
+
+const mongoose = require('mongoose');
+const MONGO_URI = process.env.MONGO_URI;
+
+if (!MONGO_URI) {
+  throw new Error('Please define the MONGO_URI environment variable');
+}
+/** The maximum number of connections in the connection pool. */
+const maxPoolSize = parseInt(process.env.MONGO_MAX_POOL_SIZE) || undefined;
+/** The minimum number of connections in the connection pool. */
+const minPoolSize = parseInt(process.env.MONGO_MIN_POOL_SIZE) || undefined;
+/** The maximum number of connections that may be in the process of being established concurrently by the connection pool. */
+const maxConnecting = parseInt(process.env.MONGO_MAX_CONNECTING) || undefined;
+/** The maximum number of milliseconds that a connection can remain idle in the pool before being removed and closed. */
+const maxIdleTimeMS = parseInt(process.env.MONGO_MAX_IDLE_TIME_MS) || undefined;
+/** The maximum time in milliseconds that a thread can wait for a connection to become available. */
+const waitQueueTimeoutMS = parseInt(process.env.MONGO_WAIT_QUEUE_TIMEOUT_MS) || undefined;
+/** Set to false to disable automatic index creation for all models associated with this connection. */
+const autoIndex =
+  process.env.MONGO_AUTO_INDEX != undefined
+    ? isEnabled(process.env.MONGO_AUTO_INDEX) || false
+    : undefined;
+
+/** Set to `false` to disable Mongoose automatically calling `createCollection()` on every model created on this connection. */
+const autoCreate =
+  process.env.MONGO_AUTO_CREATE != undefined
+    ? isEnabled(process.env.MONGO_AUTO_CREATE) || false
+    : undefined;
+/**
+ * Global is used here to maintain a cached connection across hot reloads
+ * in development. This prevents connections growing exponentially
+ * during API Route usage.
+ */
+let cached = global.mongoose;
+
+if (!cached) {
+  cached = global.mongoose = { conn: null, promise: null };
+}
+
+async function connectDb() {
+  if (cached.conn && cached.conn?._readyState === 1) {
+    return cached.conn;
+  }
+
+  const disconnected = cached.conn && cached.conn?._readyState !== 1;
+  if (!cached.promise || disconnected) {
+    const opts = {
+      bufferCommands: false,
+      ...(maxPoolSize ? { maxPoolSize } : {}),
+      ...(minPoolSize ? { minPoolSize } : {}),
+      ...(maxConnecting ? { maxConnecting } : {}),
+      ...(maxIdleTimeMS ? { maxIdleTimeMS } : {}),
+      ...(waitQueueTimeoutMS ? { waitQueueTimeoutMS } : {}),
+      ...(autoIndex != undefined ? { autoIndex } : {}),
+      ...(autoCreate != undefined ? { autoCreate } : {}),
+      // useNewUrlParser: true,
+      // useUnifiedTopology: true,
+      // bufferMaxEntries: 0,
+      // useFindAndModify: true,
+      // useCreateIndex: true
+    };
+    logger.info('Mongo Connection options');
+    logger.info(JSON.stringify(opts, null, 2));
+    mongoose.set('strictQuery', true);
+    cached.promise = mongoose.connect(MONGO_URI, opts).then((mongoose) => {
+      return mongoose;
+    });
+  }
+  cached.conn = await cached.promise;
+
+  return cached.conn;
+}
+
+module.exports = {
+  connectDb,
+};

api/db/index.js

@@ -0,0 +1,8 @@
+const mongoose = require('mongoose');
+const { createModels } = require('@librechat/data-schemas');
+const { connectDb } = require('./connect');
+const indexSync = require('./indexSync');
+
+createModels(mongoose);
+
+module.exports = { connectDb, indexSync };

api/db/indexSync.js

@@ -0,0 +1,262 @@
+const mongoose = require('mongoose');
+const { MeiliSearch } = require('meilisearch');
+const { logger } = require('@librechat/data-schemas');
+const { CacheKeys } = require('librechat-data-provider');
+const { isEnabled, FlowStateManager } = require('@librechat/api');
+const { getLogStores } = require('~/cache');
+
+const Conversation = mongoose.models.Conversation;
+const Message = mongoose.models.Message;
+
+const searchEnabled = isEnabled(process.env.SEARCH);
+const indexingDisabled = isEnabled(process.env.MEILI_NO_SYNC);
+let currentTimeout = null;
+
+class MeiliSearchClient {
+  static instance = null;
+
+  static getInstance() {
+    if (!MeiliSearchClient.instance) {
+      if (!process.env.MEILI_HOST || !process.env.MEILI_MASTER_KEY) {
+        throw new Error('Meilisearch configuration is missing.');
+      }
+      MeiliSearchClient.instance = new MeiliSearch({
+        host: process.env.MEILI_HOST,
+        apiKey: process.env.MEILI_MASTER_KEY,
+      });
+    }
+    return MeiliSearchClient.instance;
+  }
+}
+
+/**
+ * Ensures indexes have proper filterable attributes configured and checks if documents have user field
+ * @param {MeiliSearch} client - MeiliSearch client instance
+ * @returns {Promise<boolean>} - true if configuration was updated or re-sync is needed
+ */
+async function ensureFilterableAttributes(client) {
+  try {
+    // Check and update messages index
+    try {
+      const messagesIndex = client.index('messages');
+      const settings = await messagesIndex.getSettings();
+
+      if (!settings.filterableAttributes || !settings.filterableAttributes.includes('user')) {
+        logger.info('[indexSync] Configuring messages index to filter by user...');
+        await messagesIndex.updateSettings({
+          filterableAttributes: ['user'],
+        });
+        logger.info('[indexSync] Messages index configured for user filtering');
+        logger.info('[indexSync] Index configuration updated. Full re-sync will be triggered.');
+        return true;
+      }
+
+      // Check if existing documents have user field indexed
+      try {
+        const searchResult = await messagesIndex.search('', { limit: 1 });
+        if (searchResult.hits.length > 0 && !searchResult.hits[0].user) {
+          logger.info('[indexSync] Existing messages missing user field, re-sync needed');
+          return true;
+        }
+      } catch (searchError) {
+        logger.debug('[indexSync] Could not check message documents:', searchError.message);
+      }
+    } catch (error) {
+      if (error.code !== 'index_not_found') {
+        logger.warn('[indexSync] Could not check/update messages index settings:', error.message);
+      }
+    }
+
+    // Check and update conversations index
+    try {
+      const convosIndex = client.index('convos');
+      const settings = await convosIndex.getSettings();
+
+      if (!settings.filterableAttributes || !settings.filterableAttributes.includes('user')) {
+        logger.info('[indexSync] Configuring convos index to filter by user...');
+        await convosIndex.updateSettings({
+          filterableAttributes: ['user'],
+        });
+        logger.info('[indexSync] Convos index configured for user filtering');
+        logger.info('[indexSync] Index configuration updated. Full re-sync will be triggered.');
+        return true;
+      }
+
+      // Check if existing documents have user field indexed
+      try {
+        const searchResult = await convosIndex.search('', { limit: 1 });
+        if (searchResult.hits.length > 0 && !searchResult.hits[0].user) {
+          logger.info('[indexSync] Existing conversations missing user field, re-sync needed');
+          return true;
+        }
+      } catch (searchError) {
+        logger.debug('[indexSync] Could not check conversation documents:', searchError.message);
+      }
+    } catch (error) {
+      if (error.code !== 'index_not_found') {
+        logger.warn('[indexSync] Could not check/update convos index settings:', error.message);
+      }
+    }
+  } catch (error) {
+    logger.error('[indexSync] Error ensuring filterable attributes:', error);
+  }
+
+  return false;
+}
+
+/**
+ * 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 };
+  }
+
+  /** Ensures indexes have proper filterable attributes configured */
+  const configUpdated = await ensureFilterableAttributes(client);
+
+  let messagesSync = false;
+  let convosSync = false;
+
+  // If configuration was just updated or documents are missing user field, force a full re-sync
+  if (configUpdated) {
+    logger.info('[indexSync] Forcing full re-sync to ensure user field is properly indexed...');
+
+    // Reset sync flags to force full re-sync
+    await Message.collection.updateMany({ _meiliIndex: true }, { $set: { _meiliIndex: false } });
+    await Conversation.collection.updateMany(
+      { _meiliIndex: true },
+      { $set: { _meiliIndex: false } },
+    );
+  }
+
+  // Check if we need to sync messages
+  const messageProgress = await Message.getSyncProgress();
+  if (!messageProgress.isComplete || configUpdated) {
+    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 || configUpdated) {
+    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 flowManager = new FlowStateManager(flowsCache, {
+      ttl: 60000 * 10, // 10 minutes TTL for sync operations
+    });
+
+    // Use a unique flow ID for the sync operation
+    const flowId = 'meili-index-sync';
+    const flowType = 'MEILI_SYNC';
+
+    // This will only execute the handler if no other instance is running the sync
+    const result = await flowManager.createFlowWithHandler(flowId, flowType, performSync);
+
+    if (result.messagesSync || result.convosSync) {
+      logger.info('[indexSync] Sync completed successfully');
+    } else {
+      logger.debug('[indexSync] No sync was needed');
+    }
+
+    return result;
+  } catch (err) {
+    if (err.message.includes('flow already exists')) {
+      logger.info('[indexSync] Sync already running on another instance');
+      return;
+    }
+
+    if (err.message.includes('not found')) {
+      logger.debug('[indexSync] Creating indices...');
+      currentTimeout = setTimeout(async () => {
+        try {
+          await Message.syncWithMeili();
+          await Conversation.syncWithMeili();
+        } catch (err) {
+          logger.error('[indexSync] Trouble creating indices, try restarting the server.', err);
+        }
+      }, 750);
+    } else if (err.message.includes('Meilisearch not configured')) {
+      logger.info('[indexSync] Meilisearch not configured, search will be disabled.');
+    } else {
+      logger.error('[indexSync] error', err);
+    }
+  }
+}
+
+process.on('exit', () => {
+  logger.debug('[indexSync] Clearing sync timeouts before exiting...');
+  clearTimeout(currentTimeout);
+});
+
+module.exports = indexSync;

api/db/models.js

@@ -0,0 +1,5 @@
+const mongoose = require('mongoose');
+const { createModels } = require('@librechat/data-schemas');
+const models = createModels(mongoose);
+
+module.exports = { ...models };

api/jest.config.js

@@ -3,6 +3,7 @@ module.exports = {
   clearMocks: true,
   roots: ['<rootDir>'],
   coverageDirectory: 'coverage',
+  testTimeout: 30000, // 30 seconds timeout for all tests
   setupFiles: [
     './test/jestSetup.js',
     './test/__mocks__/logger.js',
@@ -11,5 +12,8 @@ module.exports = {
   moduleNameMapper: {
     '~/(.*)': '<rootDir>/$1',
     '~/data/auth.json': '<rootDir>/__mocks__/auth.mock.json',
+    '^openid-client/passport$': '<rootDir>/test/__mocks__/openid-client-passport.js', // Mock for the passport strategy part
+    '^openid-client$': '<rootDir>/test/__mocks__/openid-client.js',
   },
+  transformIgnorePatterns: ['/node_modules/(?!(openid-client|oauth4webapi|jose)/).*/'],
 };

api/lib/db/connectDb.js

@@ -1,45 +0,0 @@
-require('dotenv').config();
-const mongoose = require('mongoose');
-const MONGO_URI = process.env.MONGO_URI;
-
-if (!MONGO_URI) {
-  throw new Error('Please define the MONGO_URI environment variable');
-}
-
-/**
- * Global is used here to maintain a cached connection across hot reloads
- * in development. This prevents connections growing exponentially
- * during API Route usage.
- */
-let cached = global.mongoose;
-
-if (!cached) {
-  cached = global.mongoose = { conn: null, promise: null };
-}
-
-async function connectDb() {
-  if (cached.conn && cached.conn?._readyState === 1) {
-    return cached.conn;
-  }
-
-  const disconnected = cached.conn && cached.conn?._readyState !== 1;
-  if (!cached.promise || disconnected) {
-    const opts = {
-      bufferCommands: false,
-      // useNewUrlParser: true,
-      // useUnifiedTopology: true,
-      // bufferMaxEntries: 0,
-      // useFindAndModify: true,
-      // useCreateIndex: true
-    };
-
-    mongoose.set('strictQuery', true);
-    cached.promise = mongoose.connect(MONGO_URI, opts).then((mongoose) => {
-      return mongoose;
-    });
-  }
-  cached.conn = await cached.promise;
-  return cached.conn;
-}
-
-module.exports = connectDb;

api/lib/db/index.js

@@ -1,4 +0,0 @@
-const connectDb = require('./connectDb');
-const indexSync = require('./indexSync');
-
-module.exports = { connectDb, indexSync };

api/lib/db/indexSync.js

@@ -1,89 +0,0 @@
-const { MeiliSearch } = require('meilisearch');
-const { Conversation } = require('~/models/Conversation');
-const { Message } = require('~/models/Message');
-const { isEnabled } = require('~/server/utils');
-const { logger } = require('~/config');
-
-const searchEnabled = isEnabled(process.env.SEARCH);
-const indexingDisabled = isEnabled(process.env.MEILI_NO_SYNC);
-let currentTimeout = null;
-
-class MeiliSearchClient {
-  static instance = null;
-
-  static getInstance() {
-    if (!MeiliSearchClient.instance) {
-      if (!process.env.MEILI_HOST || !process.env.MEILI_MASTER_KEY) {
-        throw new Error('Meilisearch configuration is missing.');
-      }
-      MeiliSearchClient.instance = new MeiliSearch({
-        host: process.env.MEILI_HOST,
-        apiKey: process.env.MEILI_MASTER_KEY,
-      });
-    }
-    return MeiliSearchClient.instance;
-  }
-}
-
-async function indexSync() {
-  if (!searchEnabled) {
-    return;
-  }
-
-  try {
-    const client = MeiliSearchClient.getInstance();
-
-    const { status } = await client.health();
-    if (status !== 'available') {
-      throw new Error('Meilisearch not available');
-    }
-
-    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 () => {
-        try {
-          await Message.syncWithMeili();
-          await Conversation.syncWithMeili();
-        } catch (err) {
-          logger.error('[indexSync] Trouble creating indices, try restarting the server.', err);
-        }
-      }, 750);
-    } else if (err.message.includes('Meilisearch not configured')) {
-      logger.info('[indexSync] Meilisearch not configured, search will be disabled.');
-    } else {
-      logger.error('[indexSync] error', err);
-    }
-  }
-}
-
-process.on('exit', () => {
-  logger.debug('[indexSync] Clearing sync timeouts before exiting...');
-  clearTimeout(currentTimeout);
-});
-
-module.exports = indexSync;

api/models/Action.js

@@ -1,7 +1,4 @@
-const mongoose = require('mongoose');
-const { actionSchema } = require('@librechat/data-schemas');
-
-const Action = mongoose.model('action', actionSchema);
+const { Action } = require('~/db/models');
 
 /**
  * Update an action with new data without overwriting existing properties,

api/models/Agent.js

@@ -1,18 +1,19 @@
 const mongoose = require('mongoose');
-const { agentSchema } = require('@librechat/data-schemas');
-const { SystemRoles, Tools } = require('librechat-data-provider');
-const { GLOBAL_PROJECT_NAME, EPHEMERAL_AGENT_ID, mcp_delimiter } =
+const crypto = require('node:crypto');
+const { logger } = require('@librechat/data-schemas');
+const { ResourceType, SystemRoles, Tools, actionDelimiter } = require('librechat-data-provider');
+const { GLOBAL_PROJECT_NAME, EPHEMERAL_AGENT_ID, mcp_all, mcp_delimiter } =
   require('librechat-data-provider').Constants;
-const { CONFIG_STORE, STARTUP_CONFIG } = require('librechat-data-provider').CacheKeys;
 const {
-  getProjectByName,
-  addAgentIdsToProject,
-  removeAgentIdsFromProject,
   removeAgentFromAllProjects,
+  removeAgentIdsFromProject,
+  addAgentIdsToProject,
+  getProjectByName,
 } = require('./Project');
-const getLogStores = require('~/cache/getLogStores');
-
-const Agent = mongoose.model('agent', agentSchema);
+const { removeAllPermissions } = require('~/server/services/PermissionService');
+const { getMCPServerTools } = require('~/server/services/Config');
+const { getActions } = require('./Action');
+const { Agent } = require('~/db/models');
 
 /**
  * Create an agent with the provided data.
@@ -21,7 +22,21 @@ const Agent = mongoose.model('agent', agentSchema);
  * @throws {Error} If the agent creation fails.
  */
 const createAgent = async (agentData) => {
-  return (await Agent.create(agentData)).toObject();
+  const { author: _author, ...versionData } = agentData;
+  const timestamp = new Date();
+  const initialAgentData = {
+    ...agentData,
+    versions: [
+      {
+        ...versionData,
+        createdAt: timestamp,
+        updatedAt: timestamp,
+      },
+    ],
+    category: agentData.category || 'general',
+  };
+
+  return (await Agent.create(initialAgentData)).toObject();
 };
 
 /**
@@ -34,6 +49,14 @@ const createAgent = async (agentData) => {
  */
 const getAgent = async (searchParameter) => await Agent.findOne(searchParameter).lean();
 
+/**
+ * Get multiple agent documents based on the provided search parameters.
+ *
+ * @param {Object} searchParameter - The search parameters to find agents.
+ * @returns {Promise<Agent[]>} Array of agent documents as plain objects.
+ */
+const getAgents = async (searchParameter) => await Agent.find(searchParameter).lean();
+
 /**
  * Load an agent based on the provided ID
  *
@@ -42,33 +65,44 @@ 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 {Agent|null} The agent document as a plain object, or null if not found.
+ * @returns {Promise<Agent|null>} The agent document as a plain object, or null if not found.
  */
-const loadEphemeralAgent = ({ req, agent_id, endpoint, model_parameters: _m }) => {
+const loadEphemeralAgent = async ({ req, agent_id, endpoint, model_parameters: _m }) => {
   const { model, ...model_parameters } = _m;
-  /** @type {Record<string, FunctionTool>} */
-  const availableTools = req.app.locals.availableTools;
-  const mcpServers = new Set(req.body.ephemeralAgent?.mcp);
+  /** @type {TEphemeralAgent | null} */
+  const ephemeralAgent = req.body.ephemeralAgent;
+  const mcpServers = new Set(ephemeralAgent?.mcp);
   /** @type {string[]} */
   const tools = [];
-  if (req.body.ephemeralAgent?.execute_code === true) {
+  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);
+  }
 
+  const addedServers = new Set();
   if (mcpServers.size > 0) {
-    for (const toolName of Object.keys(availableTools)) {
-      if (!toolName.includes(mcp_delimiter)) {
+    for (const mcpServer of mcpServers) {
+      if (addedServers.has(mcpServer)) {
         continue;
       }
-      const mcpServer = toolName.split(mcp_delimiter)?.[1];
-      if (mcpServer && mcpServers.has(mcpServer)) {
-        tools.push(toolName);
+      const serverTools = await getMCPServerTools(mcpServer);
+      if (!serverTools) {
+        tools.push(`${mcp_all}${mcp_delimiter}${mcpServer}`);
+        addedServers.add(mcpServer);
+        continue;
       }
+      tools.push(...Object.keys(serverTools));
+      addedServers.add(mcpServer);
     }
   }
 
   const instructions = req.body.promptPrefix;
-  return {
+  const result = {
     id: agent_id,
     instructions,
     provider: endpoint,
@@ -76,6 +110,11 @@ const loadEphemeralAgent = ({ req, agent_id, endpoint, model_parameters: _m }) =
     model,
     tools,
   };
+
+  if (ephemeralAgent?.artifacts != null && ephemeralAgent.artifacts) {
+    result.artifacts = ephemeralAgent.artifacts;
+  }
+  return result;
 };
 
 /**
@@ -93,7 +132,7 @@ const loadAgent = async ({ req, agent_id, endpoint, model_parameters }) => {
     return null;
   }
   if (agent_id === EPHEMERAL_AGENT_ID) {
-    return loadEphemeralAgent({ req, agent_id, endpoint, model_parameters });
+    return await loadEphemeralAgent({ req, agent_id, endpoint, model_parameters });
   }
   const agent = await getAgent({
     id: agent_id,
@@ -103,43 +142,271 @@ const loadAgent = async ({ req, agent_id, endpoint, model_parameters }) => {
     return null;
   }
 
-  if (agent.author.toString() === req.user.id) {
-    return agent;
-  }
+  agent.version = agent.versions ? agent.versions.length : 0;
+  return agent;
+};
 
-  if (!agent.projectIds) {
+/**
+ * Check if a version already exists in the versions array, excluding timestamp and author fields
+ * @param {Object} updateData - The update data to compare
+ * @param {Object} currentData - The current agent data
+ * @param {Array} versions - The existing versions array
+ * @param {string} [actionsHash] - Hash of current action metadata
+ * @returns {Object|null} - The matching version if found, null otherwise
+ */
+const isDuplicateVersion = (updateData, currentData, versions, actionsHash = null) => {
+  if (!versions || versions.length === 0) {
     return null;
   }
 
-  const cache = getLogStores(CONFIG_STORE);
-  /** @type {TStartupConfig} */
-  const cachedStartupConfig = await cache.get(STARTUP_CONFIG);
-  let { instanceProjectId } = cachedStartupConfig ?? {};
-  if (!instanceProjectId) {
-    instanceProjectId = (await getProjectByName(GLOBAL_PROJECT_NAME, '_id'))._id.toString();
+  const excludeFields = [
+    '_id',
+    'id',
+    'createdAt',
+    'updatedAt',
+    'author',
+    'updatedBy',
+    'created_at',
+    'updated_at',
+    '__v',
+    'versions',
+    'actionsHash', // Exclude actionsHash from direct comparison
+  ];
+
+  const { $push: _$push, $pull: _$pull, $addToSet: _$addToSet, ...directUpdates } = updateData;
+
+  if (Object.keys(directUpdates).length === 0 && !actionsHash) {
+    return null;
   }
 
-  for (const projectObjectId of agent.projectIds) {
-    const projectId = projectObjectId.toString();
-    if (projectId === instanceProjectId) {
-      return agent;
+  const wouldBeVersion = { ...currentData, ...directUpdates };
+  const lastVersion = versions[versions.length - 1];
+
+  if (actionsHash && lastVersion.actionsHash !== actionsHash) {
+    return null;
+  }
+
+  const allFields = new Set([...Object.keys(wouldBeVersion), ...Object.keys(lastVersion)]);
+
+  const importantFields = Array.from(allFields).filter((field) => !excludeFields.includes(field));
+
+  let isMatch = true;
+  for (const field of importantFields) {
+    const wouldBeValue = wouldBeVersion[field];
+    const lastVersionValue = lastVersion[field];
+
+    // Skip if both are undefined/null
+    if (!wouldBeValue && !lastVersionValue) {
+      continue;
+    }
+
+    // Handle arrays
+    if (Array.isArray(wouldBeValue) || Array.isArray(lastVersionValue)) {
+      // Normalize: treat undefined/null as empty array for comparison
+      let wouldBeArr;
+      if (Array.isArray(wouldBeValue)) {
+        wouldBeArr = wouldBeValue;
+      } else if (wouldBeValue == null) {
+        wouldBeArr = [];
+      } else {
+        wouldBeArr = [wouldBeValue];
+      }
+
+      let lastVersionArr;
+      if (Array.isArray(lastVersionValue)) {
+        lastVersionArr = lastVersionValue;
+      } else if (lastVersionValue == null) {
+        lastVersionArr = [];
+      } else {
+        lastVersionArr = [lastVersionValue];
+      }
+
+      if (wouldBeArr.length !== lastVersionArr.length) {
+        isMatch = false;
+        break;
+      }
+
+      // Special handling for projectIds (MongoDB ObjectIds)
+      if (field === 'projectIds') {
+        const wouldBeIds = wouldBeArr.map((id) => id.toString()).sort();
+        const versionIds = lastVersionArr.map((id) => id.toString()).sort();
+
+        if (!wouldBeIds.every((id, i) => id === versionIds[i])) {
+          isMatch = false;
+          break;
+        }
+      }
+      // Handle arrays of objects
+      else if (
+        wouldBeArr.length > 0 &&
+        typeof wouldBeArr[0] === 'object' &&
+        wouldBeArr[0] !== null
+      ) {
+        const sortedWouldBe = [...wouldBeArr].map((item) => JSON.stringify(item)).sort();
+        const sortedVersion = [...lastVersionArr].map((item) => JSON.stringify(item)).sort();
+
+        if (!sortedWouldBe.every((item, i) => item === sortedVersion[i])) {
+          isMatch = false;
+          break;
+        }
+      } else {
+        const sortedWouldBe = [...wouldBeArr].sort();
+        const sortedVersion = [...lastVersionArr].sort();
+
+        if (!sortedWouldBe.every((item, i) => item === sortedVersion[i])) {
+          isMatch = false;
+          break;
+        }
+      }
+    }
+    // Handle objects
+    else if (typeof wouldBeValue === 'object' && wouldBeValue !== null) {
+      const lastVersionObj =
+        typeof lastVersionValue === 'object' && lastVersionValue !== null ? lastVersionValue : {};
+
+      // For empty objects, normalize the comparison
+      const wouldBeKeys = Object.keys(wouldBeValue);
+      const lastVersionKeys = Object.keys(lastVersionObj);
+
+      // If both are empty objects, they're equal
+      if (wouldBeKeys.length === 0 && lastVersionKeys.length === 0) {
+        continue;
+      }
+
+      // Otherwise do a deep comparison
+      if (JSON.stringify(wouldBeValue) !== JSON.stringify(lastVersionObj)) {
+        isMatch = false;
+        break;
+      }
+    }
+    // Handle primitive values
+    else {
+      // For primitives, handle the case where one is undefined and the other is a default value
+      if (wouldBeValue !== lastVersionValue) {
+        // Special handling for boolean false vs undefined
+        if (
+          typeof wouldBeValue === 'boolean' &&
+          wouldBeValue === false &&
+          lastVersionValue === undefined
+        ) {
+          continue;
+        }
+        // Special handling for empty string vs undefined
+        if (
+          typeof wouldBeValue === 'string' &&
+          wouldBeValue === '' &&
+          lastVersionValue === undefined
+        ) {
+          continue;
+        }
+        isMatch = false;
+        break;
+      }
     }
   }
+
+  return isMatch ? lastVersion : null;
 };
 
 /**
  * Update an agent with new data without overwriting existing
  *  properties, or create a new agent if it doesn't exist.
+ * When an agent is updated, a copy of the current state will be saved to the versions array.
  *
  * @param {Object} searchParameter - The search parameters to find the agent to update.
  * @param {string} searchParameter.id - The ID of the agent to update.
  * @param {string} [searchParameter.author] - The user ID of the agent's author.
  * @param {Object} updateData - An object containing the properties to update.
+ * @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) => {
-  const options = { new: true, upsert: false };
-  return Agent.findOneAndUpdate(searchParameter, updateData, options).lean();
+const updateAgent = async (searchParameter, updateData, options = {}) => {
+  const { updatingUserId = null, forceVersion = false, skipVersioning = false } = options;
+  const mongoOptions = { new: true, upsert: false };
+
+  const currentAgent = await Agent.findOne(searchParameter);
+  if (currentAgent) {
+    const {
+      __v,
+      _id,
+      id: __id,
+      versions,
+      author: _author,
+      ...versionData
+    } = currentAgent.toObject();
+    const { $push, $pull, $addToSet, ...directUpdates } = updateData;
+
+    let actionsHash = null;
+
+    // Generate actions hash if agent has actions
+    if (currentAgent.actions && currentAgent.actions.length > 0) {
+      // Extract action IDs from the format "domain_action_id"
+      const actionIds = currentAgent.actions
+        .map((action) => {
+          const parts = action.split(actionDelimiter);
+          return parts[1]; // Get just the action ID part
+        })
+        .filter(Boolean);
+
+      if (actionIds.length > 0) {
+        try {
+          const actions = await getActions(
+            {
+              action_id: { $in: actionIds },
+            },
+            true,
+          ); // Include sensitive data for hash
+
+          actionsHash = await generateActionMetadataHash(currentAgent.actions, actions);
+        } catch (error) {
+          logger.error('Error fetching actions for hash generation:', error);
+        }
+      }
+    }
+
+    const shouldCreateVersion =
+      !skipVersioning &&
+      (forceVersion || Object.keys(directUpdates).length > 0 || $push || $pull || $addToSet);
+
+    if (shouldCreateVersion) {
+      const duplicateVersion = isDuplicateVersion(updateData, versionData, versions, actionsHash);
+      if (duplicateVersion && !forceVersion) {
+        // No changes detected, return the current agent without creating a new version
+        const agentObj = currentAgent.toObject();
+        agentObj.version = versions.length;
+        return agentObj;
+      }
+    }
+
+    const versionEntry = {
+      ...versionData,
+      ...directUpdates,
+      updatedAt: new Date(),
+    };
+
+    // Include actions hash in version if available
+    if (actionsHash) {
+      versionEntry.actionsHash = actionsHash;
+    }
+
+    // Always store updatedBy field to track who made the change
+    if (updatingUserId) {
+      versionEntry.updatedBy = new mongoose.Types.ObjectId(updatingUserId);
+    }
+
+    if (shouldCreateVersion) {
+      updateData.$push = {
+        ...($push || {}),
+        versions: versionEntry,
+      };
+    }
+  }
+
+  return Agent.findOneAndUpdate(searchParameter, updateData, mongoOptions).lean();
 };
 
 /**
@@ -151,7 +418,7 @@ const updateAgent = async (searchParameter, updateData) => {
  * @param {string} params.file_id
  * @returns {Promise<Agent>} The updated agent.
  */
-const addAgentResourceFile = async ({ agent_id, tool_resource, file_id }) => {
+const addAgentResourceFile = async ({ req, agent_id, tool_resource, file_id }) => {
   const searchParameter = { id: agent_id };
   let agent = await getAgent(searchParameter);
   if (!agent) {
@@ -177,7 +444,9 @@ const addAgentResourceFile = async ({ agent_id, tool_resource, file_id }) => {
     },
   };
 
-  const updatedAgent = await updateAgent(searchParameter, updateData);
+  const updatedAgent = await updateAgent(searchParameter, updateData, {
+    updatingUserId: req?.user?.id,
+  });
   if (updatedAgent) {
     return updatedAgent;
   } else {
@@ -248,12 +517,117 @@ const deleteAgent = async (searchParameter) => {
   const agent = await Agent.findOneAndDelete(searchParameter);
   if (agent) {
     await removeAgentFromAllProjects(agent.id);
+    await removeAllPermissions({
+      resourceType: ResourceType.AGENT,
+      resourceId: agent._id,
+    });
   }
   return agent;
 };
 
+/**
+ * Get agents by accessible IDs with optional cursor-based pagination.
+ * @param {Object} params - The parameters for getting accessible agents.
+ * @param {Array} [params.accessibleIds] - Array of agent ObjectIds the user has ACL access to.
+ * @param {Object} [params.otherParams] - Additional query parameters (including author filter).
+ * @param {number} [params.limit] - Number of agents to return (max 100). If not provided, returns all agents.
+ * @param {string} [params.after] - Cursor for pagination - get agents after this cursor. // base64 encoded JSON string with updatedAt and _id.
+ * @returns {Promise<Object>} A promise that resolves to an object containing the agents data and pagination info.
+ */
+const getListAgentsByAccess = async ({
+  accessibleIds = [],
+  otherParams = {},
+  limit = null,
+  after = null,
+}) => {
+  const isPaginated = limit !== null && limit !== undefined;
+  const normalizedLimit = isPaginated ? Math.min(Math.max(1, parseInt(limit) || 20), 100) : null;
+
+  // Build base query combining ACL accessible agents with other filters
+  const baseQuery = { ...otherParams, _id: { $in: accessibleIds } };
+
+  // Add cursor condition
+  if (after) {
+    try {
+      const cursor = JSON.parse(Buffer.from(after, 'base64').toString('utf8'));
+      const { updatedAt, _id } = cursor;
+
+      const cursorCondition = {
+        $or: [
+          { updatedAt: { $lt: new Date(updatedAt) } },
+          { updatedAt: new Date(updatedAt), _id: { $gt: new mongoose.Types.ObjectId(_id) } },
+        ],
+      };
+
+      // Merge cursor condition with base query
+      if (Object.keys(baseQuery).length > 0) {
+        baseQuery.$and = [{ ...baseQuery }, cursorCondition];
+        // Remove the original conditions from baseQuery to avoid duplication
+        Object.keys(baseQuery).forEach((key) => {
+          if (key !== '$and') delete baseQuery[key];
+        });
+      } else {
+        Object.assign(baseQuery, cursorCondition);
+      }
+    } catch (error) {
+      logger.warn('Invalid cursor:', error.message);
+    }
+  }
+
+  let query = Agent.find(baseQuery, {
+    id: 1,
+    _id: 1,
+    name: 1,
+    avatar: 1,
+    author: 1,
+    projectIds: 1,
+    description: 1,
+    updatedAt: 1,
+    category: 1,
+    support_contact: 1,
+    is_promoted: 1,
+  }).sort({ updatedAt: -1, _id: 1 });
+
+  // Only apply limit if pagination is requested
+  if (isPaginated) {
+    query = query.limit(normalizedLimit + 1);
+  }
+
+  const agents = await query.lean();
+
+  const hasMore = isPaginated ? agents.length > normalizedLimit : false;
+  const data = (isPaginated ? agents.slice(0, normalizedLimit) : agents).map((agent) => {
+    if (agent.author) {
+      agent.author = agent.author.toString();
+    }
+    return agent;
+  });
+
+  // Generate next cursor only if paginated
+  let nextCursor = null;
+  if (isPaginated && hasMore && data.length > 0) {
+    const lastAgent = agents[normalizedLimit - 1];
+    nextCursor = Buffer.from(
+      JSON.stringify({
+        updatedAt: lastAgent.updatedAt.toISOString(),
+        _id: lastAgent._id.toString(),
+      }),
+    ).toString('base64');
+  }
+
+  return {
+    object: 'list',
+    data,
+    first_id: data.length > 0 ? data[0].id : null,
+    last_id: data.length > 0 ? data[data.length - 1].id : null,
+    has_more: hasMore,
+    after: nextCursor,
+  };
+};
+
 /**
  * Get all agents.
+ * @deprecated Use getListAgentsByAccess for ACL-aware agent listing
  * @param {Object} searchParameter - The search parameters to find matching agents.
  * @param {string} searchParameter.author - The user ID of the agent's author.
  * @returns {Promise<Object>} A promise that resolves to an object containing the agents data and pagination info.
@@ -269,17 +643,18 @@ const getListAgents = async (searchParameter) => {
     delete globalQuery.author;
     query = { $or: [globalQuery, query] };
   }
-
   const agents = (
     await Agent.find(query, {
       id: 1,
-      _id: 0,
+      _id: 1,
       name: 1,
       avatar: 1,
       author: 1,
       projectIds: 1,
       description: 1,
+      // @deprecated - isCollaborative replaced by ACL permissions
       isCollaborative: 1,
+      category: 1,
     }).lean()
   ).map((agent) => {
     if (agent.author?.toString() !== author) {
@@ -308,7 +683,7 @@ const getListAgents = async (searchParameter) => {
  * This function also updates the corresponding projects to include or exclude the agent ID.
  *
  * @param {Object} params - Parameters for updating the agent's projects.
- * @param {MongoUser} params.user - Parameters for updating the agent's projects.
+ * @param {IUser} params.user - Parameters for updating the agent's projects.
  * @param {string} params.agentId - The ID of the agent to update.
  * @param {string[]} [params.projectIds] - Array of project IDs to add to the agent.
  * @param {string[]} [params.removeProjectIds] - Array of project IDs to remove from the agent.
@@ -341,7 +716,10 @@ const updateAgentProjects = async ({ user, agentId, projectIds, removeProjectIds
     delete updateQuery.author;
   }
 
-  const updatedAgent = await updateAgent(updateQuery, updateOps);
+  const updatedAgent = await updateAgent(updateQuery, updateOps, {
+    updatingUserId: user.id,
+    skipVersioning: true,
+  });
   if (updatedAgent) {
     return updatedAgent;
   }
@@ -358,15 +736,118 @@ const updateAgentProjects = async ({ user, agentId, projectIds, removeProjectIds
   return await getAgent({ id: agentId });
 };
 
+/**
+ * Reverts an agent to a specific version in its version history.
+ * @param {Object} searchParameter - The search parameters to find the agent to revert.
+ * @param {string} searchParameter.id - The ID of the agent to revert.
+ * @param {string} [searchParameter.author] - The user ID of the agent's author.
+ * @param {number} versionIndex - The index of the version to revert to in the versions array.
+ * @returns {Promise<MongoAgent>} The updated agent document after reverting.
+ * @throws {Error} If the agent is not found or the specified version does not exist.
+ */
+const revertAgentVersion = async (searchParameter, versionIndex) => {
+  const agent = await Agent.findOne(searchParameter);
+  if (!agent) {
+    throw new Error('Agent not found');
+  }
+
+  if (!agent.versions || !agent.versions[versionIndex]) {
+    throw new Error(`Version ${versionIndex} not found`);
+  }
+
+  const revertToVersion = agent.versions[versionIndex];
+
+  const updateData = {
+    ...revertToVersion,
+  };
+
+  delete updateData._id;
+  delete updateData.id;
+  delete updateData.versions;
+  delete updateData.author;
+  delete updateData.updatedBy;
+
+  return Agent.findOneAndUpdate(searchParameter, updateData, { new: true }).lean();
+};
+
+/**
+ * Generates a hash of action metadata for version comparison
+ * @param {string[]} actionIds - Array of action IDs in format "domain_action_id"
+ * @param {Action[]} actions - Array of action documents
+ * @returns {Promise<string>} - SHA256 hash of the action metadata
+ */
+const generateActionMetadataHash = async (actionIds, actions) => {
+  if (!actionIds || actionIds.length === 0) {
+    return '';
+  }
+
+  // Create a map of action_id to metadata for quick lookup
+  const actionMap = new Map();
+  actions.forEach((action) => {
+    actionMap.set(action.action_id, action.metadata);
+  });
+
+  // Sort action IDs for consistent hashing
+  const sortedActionIds = [...actionIds].sort();
+
+  // Build a deterministic string representation of all action metadata
+  const metadataString = sortedActionIds
+    .map((actionFullId) => {
+      // Extract just the action_id part (after the delimiter)
+      const parts = actionFullId.split(actionDelimiter);
+      const actionId = parts[1];
+
+      const metadata = actionMap.get(actionId);
+      if (!metadata) {
+        return `${actionId}:null`;
+      }
+
+      // Sort metadata keys for deterministic output
+      const sortedKeys = Object.keys(metadata).sort();
+      const metadataStr = sortedKeys
+        .map((key) => `${key}:${JSON.stringify(metadata[key])}`)
+        .join(',');
+      return `${actionId}:{${metadataStr}}`;
+    })
+    .join(';');
+
+  // Use Web Crypto API to generate hash
+  const encoder = new TextEncoder();
+  const data = encoder.encode(metadataString);
+  const hashBuffer = await crypto.webcrypto.subtle.digest('SHA-256', data);
+  const hashArray = Array.from(new Uint8Array(hashBuffer));
+  const hashHex = hashArray.map((b) => b.toString(16).padStart(2, '0')).join('');
+
+  return hashHex;
+};
+/**
+ * Counts the number of promoted agents.
+ * @returns  {Promise<number>} - The count of promoted agents
+ */
+const countPromotedAgents = async () => {
+  const count = await Agent.countDocuments({ is_promoted: true });
+  return count;
+};
+
+/**
+ * Load a default agent based on the endpoint
+ * @param {string} endpoint
+ * @returns {Agent | null}
+ */
+
 module.exports = {
-  Agent,
   getAgent,
+  getAgents,
   loadAgent,
   createAgent,
   updateAgent,
   deleteAgent,
   getListAgents,
+  revertAgentVersion,
   updateAgentProjects,
   addAgentResourceFile,
+  getListAgentsByAccess,
   removeAgentResourceFiles,
+  generateActionMetadataHash,
+  countPromotedAgents,
 };

api/models/Assistant.js

@@ -1,7 +1,4 @@
-const mongoose = require('mongoose');
-const { assistantSchema } = require('@librechat/data-schemas');
-
-const Assistant = mongoose.model('assistant', assistantSchema);
+const { Assistant } = require('~/db/models');
 
 /**
  * Update an assistant with new data without overwriting existing properties,

api/models/Balance.js

@@ -1,4 +0,0 @@
-const mongoose = require('mongoose');
-const { balanceSchema } = require('@librechat/data-schemas');
-
-module.exports = mongoose.model('Balance', balanceSchema);

api/models/Banner.js

@@ -1,8 +1,5 @@
-const mongoose = require('mongoose');
-const logger = require('~/config/winston');
-const { bannerSchema } = require('@librechat/data-schemas');
-
-const Banner = mongoose.model('Banner', bannerSchema);
+const { logger } = require('@librechat/data-schemas');
+const { Banner } = require('~/db/models');
 
 /**
  * Retrieves the current active banner.
@@ -28,4 +25,4 @@ const getBanner = async (user) => {
   }
 };
 
-module.exports = { Banner, getBanner };
+module.exports = { getBanner };

api/models/Categories.js

@@ -1,4 +1,4 @@
-const { logger } = require('~/config');
+const { logger } = require('@librechat/data-schemas');
 
 const options = [
   {