chore: bring back file search and ocr options

- Add comprehensive debug logging throughout PDF processing pipeline
- Refactor attachment processing to separate image and document handling
- Create distinct addImageURLs(), addDocuments(), and processAttachments() methods
- Fix critical bugs in stream handling and parameter passing
- Add streamToBuffer utility for proper stream-to-buffer conversion
- Remove api/agents submodule from repository

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

.env.example

@@ -58,7 +58,7 @@ DEBUG_CONSOLE=false
 #                     Endpoints                     #
 #===================================================#
 
-# ENDPOINTS=openAI,assistants,azureOpenAI,google,gptPlugins,anthropic
+# ENDPOINTS=openAI,assistants,azureOpenAI,google,anthropic
 
 PROXY=
 
@@ -142,10 +142,10 @@ GOOGLE_KEY=user_provided
 # GOOGLE_AUTH_HEADER=true
 
 # Gemini API (AI Studio)
-# GOOGLE_MODELS=gemini-2.5-pro-preview-05-06,gemini-2.5-flash-preview-04-17,gemini-2.0-flash-001,gemini-2.0-flash-exp,gemini-2.0-flash-lite-001,gemini-1.5-pro-002,gemini-1.5-flash-002
+# GOOGLE_MODELS=gemini-2.5-pro,gemini-2.5-flash,gemini-2.5-flash-lite-preview-06-17,gemini-2.0-flash,gemini-2.0-flash-lite
 
 # Vertex AI
-# GOOGLE_MODELS=gemini-2.5-pro-preview-05-06,gemini-2.5-flash-preview-04-17,gemini-2.0-flash-001,gemini-2.0-flash-exp,gemini-2.0-flash-lite-001,gemini-1.5-pro-002,gemini-1.5-flash-002
+# GOOGLE_MODELS=gemini-2.5-pro,gemini-2.5-flash,gemini-2.5-flash-lite-preview-06-17,gemini-2.0-flash-001,gemini-2.0-flash-lite-001
 
 # GOOGLE_TITLE_MODEL=gemini-2.0-flash-lite-001
 
@@ -349,6 +349,11 @@ 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
@@ -437,6 +442,8 @@ 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=
@@ -453,8 +460,8 @@ OPENID_REUSE_TOKENS=
 OPENID_JWKS_URL_CACHE_ENABLED=
 OPENID_JWKS_URL_CACHE_TIME= # 600000 ms eq to 10 minutes leave empty to disable caching
 #Set to true to trigger token exchange flow to acquire access token for the userinfo endpoint.
-OPENID_ON_BEHALF_FLOW_FOR_USERINFRO_REQUIRED=
-OPENID_ON_BEHALF_FLOW_USERINFRO_SCOPE = "user.read" # example for Scope Needed for Microsoft Graph API
+OPENID_ON_BEHALF_FLOW_FOR_USERINFO_REQUIRED=
+OPENID_ON_BEHALF_FLOW_USERINFO_SCOPE="user.read" # example for Scope Needed for Microsoft Graph API
 # Set to true to use the OpenID Connect end session endpoint for logout
 OPENID_USE_END_SESSION_ENDPOINT=
 
@@ -515,6 +522,18 @@ 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           #
 #========================#
@@ -563,6 +582,10 @@ 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                         #
 #===================================================#
@@ -580,11 +603,40 @@ 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
+
+# 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., STATIC_CONFIG,ROLES,MESSAGES)
+# FORCED_IN_MEMORY_CACHE_NAMESPACES=STATIC_CONFIG,ROLES
 
 #==================================================#
 #                      Others                      #
@@ -645,4 +697,4 @@ OPENWEATHER_API_KEY=
 # Reranker (Required)
 # JINA_API_KEY=your_jina_api_key
 # or
-# COHERE_API_KEY=your_cohere_api_key
+# COHERE_API_KEY=your_cohere_api_key

.github/CONTRIBUTING.md

@@ -30,8 +30,8 @@ Project maintainers have the right and responsibility to remove, edit, or reject
 2. Install typescript globally: `npm i -g typescript`.
 3. Run `npm ci` to install dependencies.
 4. Build the data provider: `npm run build:data-provider`.
-5. Build 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`.

.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-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

@@ -1,95 +0,0 @@
-name: Generate Release Changelog PR
-
-on:
-  push:
-    tags:
-      - 'v*.*.*'
-  workflow_dispatch:
-
-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: "📜 docs: 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: "📜 docs: 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
@@ -26,15 +27,49 @@ jobs:
         uses: azure/setup-helm@v4
         env:
           GITHUB_TOKEN: "${{ secrets.GITHUB_TOKEN }}"
+      
       - name: Build Subchart Deps
         run: |
-          cd helm/librechat-rag-api
-          helm dependency build 
+          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
+      - 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:
-          charts_dir: helm
-          skip_existing: true
-        env:
-          CR_TOKEN: "${{ secrets.GITHUB_TOKEN }}"
+          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

@@ -5,12 +5,14 @@ on:
     paths:
       - "client/src/**"
       - "api/**"
+      - "packages/data-provider/src/**"
+      - "packages/client/**"
 
 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 +24,7 @@ jobs:
 
           # Define paths
           I18N_FILE="client/src/locales/en/translation.json"
-          SOURCE_DIRS=("client/src" "api" "packages/data-provider/src")
+          SOURCE_DIRS=("client/src" "api" "packages/data-provider/src" "packages/client")
 
           # Check if translation file exists
           if [[ ! -f "$I18N_FILE" ]]; then

.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
 
@@ -55,6 +58,7 @@ bower_components/
 # AI
 .clineignore
 .cursor
+.aider*
 
 # Floobits
 .floo
@@ -124,3 +128,12 @@ helm/**/.values.yaml
 
 # SAML Idp cert
 *.cert
+
+# AI Assistants
+/.claude/
+/.cursor/
+/.copilot/
+/.aider/
+/.openai/
+/.tabnine/
+/.codeium

Dockerfile

@@ -1,4 +1,4 @@
-# v0.7.8
+# v0.8.0-rc2
 
 # Base node image
 FROM node:20-alpine AS node

Dockerfile.multi

@@ -1,5 +1,5 @@
 # Dockerfile.multi
-# v0.7.8
+# v0.8.0-rc2
 
 # 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,
@@ -66,10 +66,9 @@
 - 🔦 **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
+    - 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
@@ -150,8 +149,8 @@ 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

@@ -10,6 +10,7 @@ const {
   validateVisionModel,
 } = require('librechat-data-provider');
 const { SplitStreamHandler: _Handler } = require('@librechat/agents');
+const { Tokenizer, createFetch, createStreamEventHandlers } = require('@librechat/api');
 const {
   truncateText,
   formatMessage,
@@ -26,8 +27,7 @@ const {
 const { getModelMaxTokens, getModelMaxOutputTokens, matchModelName } = require('~/utils');
 const { spendTokens, spendStructuredTokens } = require('~/models/spendTokens');
 const { encodeAndFormat } = require('~/server/services/Files/images/encode');
-const { createFetch, createStreamEventHandlers } = require('./generators');
-const Tokenizer = require('~/server/services/Tokenizer');
+const { encodeAndFormatDocuments } = require('~/server/services/Files/documents');
 const { sleep } = require('~/server/utils');
 const BaseClient = require('./BaseClient');
 const { logger } = require('~/config');
@@ -191,10 +191,11 @@ 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) {
@@ -312,6 +313,33 @@ class AnthropicClient extends BaseClient {
     return files;
   }
 
+  async addDocuments(message, attachments) {
+    // Only process documents
+    const documentResult = await encodeAndFormatDocuments(
+      this.options.req,
+      attachments,
+      EModelEndpoint.anthropic,
+    );
+
+    message.documents =
+      documentResult.documents && documentResult.documents.length
+        ? documentResult.documents
+        : undefined;
+
+    return documentResult.files;
+  }
+
+  async processAttachments(message, attachments) {
+    // Process both images and documents
+    const [imageFiles, documentFiles] = await Promise.all([
+      this.addImageURLs(message, attachments),
+      this.addDocuments(message, attachments),
+    ]);
+
+    // Combine files from both processors
+    return [...imageFiles, ...documentFiles];
+  }
+
   /**
    * @param {object} params
    * @param {number} params.promptTokens
@@ -382,7 +410,7 @@ class AnthropicClient extends BaseClient {
         };
       }
 
-      const files = await this.addImageURLs(latestMessage, attachments);
+      const files = await this.processAttachments(latestMessage, attachments);
 
       this.options.attachments = files;
     }
@@ -941,7 +969,7 @@ class AnthropicClient extends BaseClient {
       const content = `<conversation_context>
   ${convo}
   </conversation_context>
-  
+
   Please generate a title for this conversation.`;
 
       const titleMessage = { role: 'user', content };

api/app/clients/BaseClient.js

@@ -13,7 +13,6 @@ const {
 const { getMessages, saveMessage, updateMessage, saveConvo, getConvo } = require('~/models');
 const { checkBalance } = require('~/models/balanceMethods');
 const { truncateToolCallOutputs } = require('./prompts');
-const { addSpaceIfNeeded } = require('~/server/utils');
 const { getFiles } = require('~/models/File');
 const TextStream = require('./TextStream');
 const { logger } = require('~/config');
@@ -109,12 +108,15 @@ 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 {number} promptTokens
    * @param {number} completionTokens
    * @returns {Promise<void>}
    */
-  async recordTokenUsage({ promptTokens, completionTokens }) {
+  async recordTokenUsage({ model, promptTokens, completionTokens }) {
     logger.debug('[BaseClient] `recordTokenUsage` not implemented.', {
+      model,
       promptTokens,
       completionTokens,
     });
@@ -198,6 +200,10 @@ class BaseClient {
       this.currentMessages[this.currentMessages.length - 1].messageId = head;
     }
 
+    if (opts.isRegenerate && responseMessageId.endsWith('_')) {
+      responseMessageId = crypto.randomUUID();
+    }
+
     this.responseMessageId = responseMessageId;
 
     return {
@@ -572,7 +578,7 @@ class BaseClient {
       });
     }
 
-    const { generation = '' } = opts;
+    const { editedContent } = opts;
 
     // It's not necessary to push to currentMessages
     // depending on subclass implementation of handling messages
@@ -587,11 +593,21 @@ class BaseClient {
           isCreatedByUser: false,
           model: this.modelOptions?.model ?? this.model,
           sender: this.sender,
-          text: generation,
         };
         this.currentMessages.push(userMessage, latestMessage);
-      } else {
-        latestMessage.text = generation;
+      } else if (editedContent != null) {
+        // Handle editedContent for content parts
+        if (editedContent && latestMessage.content && Array.isArray(latestMessage.content)) {
+          const { index, text, type } = editedContent;
+          if (index >= 0 && index < latestMessage.content.length) {
+            const contentPart = latestMessage.content[index];
+            if (type === ContentTypes.THINK && contentPart.type === ContentTypes.THINK) {
+              contentPart[ContentTypes.THINK] = text;
+            } else if (type === ContentTypes.TEXT && contentPart.type === ContentTypes.TEXT) {
+              contentPart[ContentTypes.TEXT] = text;
+            }
+          }
+        }
       }
       this.continued = true;
     } else {
@@ -672,16 +688,32 @@ 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 +744,13 @@ class BaseClient {
       } else {
         responseMessage.tokenCount = this.getTokenCountForResponse(responseMessage);
         completionTokens = responseMessage.tokenCount;
+        await this.recordTokenUsage({
+          usage,
+          promptTokens,
+          completionTokens,
+          model: responseMessage.model,
+        });
       }
-
-      await this.recordTokenUsage({ promptTokens, completionTokens, usage });
     }
 
     if (userMessagePromise) {
@@ -792,7 +828,8 @@ 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 +838,8 @@ 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 +1131,50 @@ 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);
@@ -1151,7 +1233,7 @@ class BaseClient {
         {},
       );
 
-      await this.addImageURLs(message, files, this.visionMode);
+      await this.processAttachments(message, files, this.visionMode);
 
       this.message_file_map[message.messageId] = files;
       return message;

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 +1,7 @@
 const { google } = require('googleapis');
 const { concat } = require('@langchain/core/utils/stream');
 const { ChatVertexAI } = require('@langchain/google-vertexai');
+const { Tokenizer, getSafetySettings } = require('@librechat/api');
 const { ChatGoogleGenerativeAI } = require('@langchain/google-genai');
 const { GoogleGenerativeAI: GenAI } = require('@google/generative-ai');
 const { HumanMessage, SystemMessage } = require('@langchain/core/messages');
@@ -11,15 +12,14 @@ const {
   endpointSettings,
   parseTextParts,
   EModelEndpoint,
+  googleSettings,
   ContentTypes,
   VisionModes,
   ErrorTypes,
   Constants,
   AuthKeys,
 } = require('librechat-data-provider');
-const { getSafetySettings } = require('~/server/services/Endpoints/google/llm');
 const { encodeAndFormat } = require('~/server/services/Files/images');
-const Tokenizer = require('~/server/services/Tokenizer');
 const { spendTokens } = require('~/models/spendTokens');
 const { getModelMaxTokens } = require('~/utils');
 const { sleep } = require('~/server/utils');
@@ -34,7 +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)/;
@@ -165,6 +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({
@@ -236,11 +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);
 
@@ -257,7 +268,7 @@ class GoogleClient extends BaseClient {
     const formattedMessages = [];
     const attachments = await this.options.attachments;
     const latestMessage = { ...messages[messages.length - 1] };
-    const files = await this.addImageURLs(latestMessage, attachments, VisionModes.generative);
+    const files = await this.processAttachments(latestMessage, attachments, VisionModes.generative);
     this.options.attachments = files;
     messages[messages.length - 1] = latestMessage;
 
@@ -301,6 +312,20 @@ class GoogleClient extends BaseClient {
     return files;
   }
 
+  // eslint-disable-next-line no-unused-vars
+  async addDocuments(message, attachments) {
+    // GoogleClient doesn't support document processing yet
+    // Return empty results for consistency
+    return [];
+  }
+
+  async processAttachments(message, attachments, mode = '') {
+    // For GoogleClient, only process images
+    const imageFiles = await this.addImageURLs(message, attachments, mode);
+    const documentFiles = await this.addDocuments(message, attachments);
+    return [...imageFiles, ...documentFiles];
+  }
+
   /**
    * Builds the augmented prompt for attachments
    * TODO: Add File API Support
@@ -334,7 +359,7 @@ class GoogleClient extends BaseClient {
 
     const { prompt } = await this.buildMessagesPrompt(messages, parentMessageId);
 
-    const files = await this.addImageURLs(latestMessage, attachments);
+    const files = await this.processAttachments(latestMessage, attachments);
 
     this.options.attachments = files;
 

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 +1,21 @@
 const { OllamaClient } = require('./OllamaClient');
 const { HttpsProxyAgent } = require('https-proxy-agent');
 const { SplitStreamHandler, CustomOpenAIClient: OpenAI } = require('@librechat/agents');
+const {
+  isEnabled,
+  Tokenizer,
+  createFetch,
+  resolveHeaders,
+  constructAzureURL,
+  genAzureChatCompletion,
+  createStreamEventHandlers,
+} = require('@librechat/api');
 const {
   Constants,
   ImageDetail,
   ContentTypes,
   parseTextParts,
   EModelEndpoint,
-  resolveHeaders,
   KnownEndpoints,
   openAISettings,
   ImageDetailCost,
@@ -16,13 +24,6 @@ const {
   validateVisionModel,
   mapModelToAzureConfig,
 } = require('librechat-data-provider');
-const {
-  extractBaseURL,
-  constructAzureURL,
-  getModelMaxTokens,
-  genAzureChatCompletion,
-  getModelMaxOutputTokens,
-} = require('~/utils');
 const {
   truncateText,
   formatMessage,
@@ -30,14 +31,12 @@ const {
   titleInstruction,
   createContextHandlers,
 } = require('./prompts');
+const { extractBaseURL, getModelMaxTokens, getModelMaxOutputTokens } = require('~/utils');
 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 { addSpaceIfNeeded, sleep } = require('~/server/utils');
 const { spendTokens } = require('~/models/spendTokens');
 const { handleOpenAIErrors } = require('./tools/util');
 const { createLLM, RunManager } = require('./llm');
-const ChatGPTClient = require('./ChatGPTClient');
 const { summaryBuffer } = require('./memory');
 const { runTitleChain } = require('./chains');
 const { tokenSplit } = require('./document');
@@ -47,12 +46,6 @@ const { logger } = require('~/config');
 class OpenAIClient extends BaseClient {
   constructor(apiKey, options = {}) {
     super(apiKey, options);
-    this.ChatGPTClient = new ChatGPTClient();
-    this.buildPrompt = this.ChatGPTClient.buildPrompt.bind(this);
-    /** @type {getCompletion} */
-    this.getCompletion = this.ChatGPTClient.getCompletion.bind(this);
-    /** @type {cohereChatCompletion} */
-    this.cohereChatCompletion = this.ChatGPTClient.cohereChatCompletion.bind(this);
     this.contextStrategy = options.contextStrategy
       ? options.contextStrategy.toLowerCase()
       : 'discard';
@@ -379,23 +372,25 @@ class OpenAIClient extends BaseClient {
     return files;
   }
 
-  async buildMessages(
-    messages,
-    parentMessageId,
-    { isChatCompletion = false, promptPrefix = null },
-    opts,
-  ) {
+  async addDocuments(message, attachments) {
+    // OpenAI doesn't support native document processing yet
+    // Return empty results for consistency
+    return [];
+  }
+
+  async processAttachments(message, attachments) {
+    // For OpenAI, only process images
+    const imageFiles = await this.addImageURLs(message, attachments);
+    const documentFiles = await this.addDocuments(message, attachments);
+    return [...imageFiles, ...documentFiles];
+  }
+
+  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;
@@ -418,7 +413,7 @@ class OpenAIClient extends BaseClient {
         };
       }
 
-      const files = await this.addImageURLs(
+      const files = await this.processAttachments(
         orderedMessages[orderedMessages.length - 1],
         attachments,
       );
@@ -1159,6 +1154,7 @@ ${convo}
       logger.debug('[OpenAIClient] chatCompletion', { baseURL, modelOptions });
       const opts = {
         baseURL,
+        fetchOptions: {},
       };
 
       if (this.useOpenRouter) {
@@ -1177,7 +1173,7 @@ ${convo}
       }
 
       if (this.options.proxy) {
-        opts.httpAgent = new HttpsProxyAgent(this.options.proxy);
+        opts.fetchOptions.agent = new HttpsProxyAgent(this.options.proxy);
       }
 
       /** @type {TAzureConfig | undefined} */
@@ -1239,7 +1235,9 @@ ${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) {
@@ -1395,7 +1393,7 @@ ${convo}
           ...modelOptions,
           stream: 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/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/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/prompts/createContextHandlers.js

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

api/app/clients/prompts/formatMessages.js

@@ -3,24 +3,61 @@ const { EModelEndpoint, ContentTypes } = require('librechat-data-provider');
 const { HumanMessage, AIMessage, SystemMessage } = require('@langchain/core/messages');
 
 /**
- * Formats a message to OpenAI Vision API payload format.
+ * Formats a message with document attachments for specific endpoints.
  *
  * @param {Object} params - The parameters for formatting.
  * @param {Object} params.message - The message object to format.
- * @param {string} [params.message.role] - The role of the message sender (must be 'user').
- * @param {string} [params.message.content] - The text content of the message.
+ * @param {Array<Object>} [params.documents] - The document attachments for the message.
  * @param {EModelEndpoint} [params.endpoint] - Identifier for specific endpoint handling
+ * @returns {(Object)} - The formatted message.
+ */
+const formatDocumentMessage = ({ message, documents, endpoint }) => {
+  const contentParts = [];
+
+  // Add documents first (for Anthropic PDFs)
+  if (documents && documents.length > 0) {
+    contentParts.push(...documents);
+  }
+
+  // Add text content
+  contentParts.push({ type: ContentTypes.TEXT, text: message.content });
+
+  if (endpoint === EModelEndpoint.anthropic) {
+    message.content = contentParts;
+    return message;
+  }
+
+  // For other endpoints, might need different handling
+  message.content = contentParts;
+  return message;
+};
+
+/**
+ * Formats a message with vision capabilities (image_urls) for specific endpoints.
+ *
+ * @param {Object} params - The parameters for formatting.
+ * @param {Object} params.message - The message object to format.
  * @param {Array<string>} [params.image_urls] - The image_urls to attach to the message.
+ * @param {EModelEndpoint} [params.endpoint] - Identifier for specific endpoint handling
  * @returns {(Object)} - The formatted message.
  */
 const formatVisionMessage = ({ message, image_urls, endpoint }) => {
+  const contentParts = [];
+
+  // Add images
+  if (image_urls && image_urls.length > 0) {
+    contentParts.push(...image_urls);
+  }
+
+  // Add text content
+  contentParts.push({ type: ContentTypes.TEXT, text: message.content });
+
   if (endpoint === EModelEndpoint.anthropic) {
-    message.content = [...image_urls, { type: ContentTypes.TEXT, text: message.content }];
+    message.content = contentParts;
     return message;
   }
 
   message.content = [{ type: ContentTypes.TEXT, text: message.content }, ...image_urls];
-
   return message;
 };
 
@@ -58,7 +95,18 @@ const formatMessage = ({ message, userName, assistantName, endpoint, langChain =
     content,
   };
 
-  const { image_urls } = message;
+  const { image_urls, documents } = message;
+
+  // Handle documents
+  if (Array.isArray(documents) && documents.length > 0 && role === 'user') {
+    return formatDocumentMessage({
+      message: formattedMessage,
+      documents: message.documents,
+      endpoint,
+    });
+  }
+
+  // Handle images
   if (Array.isArray(image_urls) && image_urls.length > 0 && role === 'user') {
     return formatVisionMessage({
       message: formattedMessage,
@@ -146,7 +194,21 @@ const formatAgentMessages = (payload) => {
       message.content = [{ type: ContentTypes.TEXT, [ContentTypes.TEXT]: message.content }];
     }
     if (message.role !== 'assistant') {
-      messages.push(formatMessage({ message, langChain: true }));
+      // Check if message has documents and preserve array structure
+      const hasDocuments =
+        Array.isArray(message.content) &&
+        message.content.some((part) => part && part.type === 'document');
+
+      if (hasDocuments && message.role === 'user') {
+        // For user messages with documents, create HumanMessage directly with array content
+        messages.push(new HumanMessage({ content: message.content }));
+      } else if (hasDocuments && message.role === 'system') {
+        // For system messages with documents, create SystemMessage directly with array content
+        messages.push(new SystemMessage({ content: message.content }));
+      } else {
+        // Use regular formatting for messages without documents
+        messages.push(formatMessage({ message, langChain: true }));
+      }
       continue;
     }
 
@@ -237,41 +299,11 @@ 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,
+  formatDocumentMessage,
+  formatVisionMessage,
   formatFromLangChain,
   formatAgentMessages,
-  formatContentStrings,
   formatLangChainMessages,
 };

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

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

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

@@ -33,7 +33,9 @@ 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 {};
     }),
@@ -420,6 +422,46 @@ 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 = {

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

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

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

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

api/app/clients/tools/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/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 { FileContext, ContentTypes } = require('librechat-data-provider');
 const { getImageBasename } = require('~/server/services/Files/images');
 const extractBaseURL = require('~/utils/extractBaseURL');
-const { logger } = require('~/config');
+const logger = require('~/config/winston');
 
 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/OpenAIImageTools.js

@@ -3,13 +3,14 @@ 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 { logAxiosError } = require('@librechat/api');
+const { logger } = require('@librechat/data-schemas');
 const { ContentTypes, EImageOutputType } = require('librechat-data-provider');
 const { getStrategyFunctions } = require('~/server/services/Files/strategies');
-const { logAxiosError, extractBaseURL } = require('~/utils');
+const { extractBaseURL } = require('~/utils');
 const { getFiles } = require('~/models/File');
-const { logger } = require('~/config');
 
 /** Default descriptions for image generation tool  */
 const DEFAULT_IMAGE_GEN_DESCRIPTION = `
@@ -64,7 +65,7 @@ const DEFAULT_IMAGE_EDIT_PROMPT_DESCRIPTION = `Describe the changes, enhancement
       Always base this prompt on the most recently uploaded reference images.`;
 
 const displayMessage =
-  'The tool displayed an image. All generated images are already plainly visible, so don\'t repeat the descriptions in detail. Do not list download links as they are available in the UI already. The user may download the images by clicking on them, but do not mention anything about downloading to the user.';
+  "The tool displayed an image. All generated images are already plainly visible, so don't repeat the descriptions in detail. Do not list download links as they are available in the UI already. The user may download the images by clicking on them, but do not mention anything about downloading to the user.";
 
 /**
  * Replaces unwanted characters from the input string
@@ -106,6 +107,12 @@ const getImageEditPromptDescription = () => {
   return process.env.IMAGE_EDIT_OAI_PROMPT_DESCRIPTION || DEFAULT_IMAGE_EDIT_PROMPT_DESCRIPTION;
 };
 
+function createAbortHandler() {
+  return function () {
+    logger.debug('[ImageGenOAI] Image generation aborted');
+  };
+}
+
 /**
  * Creates OpenAI Image tools (generation and editing)
  * @param {Object} fields - Configuration fields
@@ -182,7 +189,10 @@ 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 +210,18 @@ 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 +245,10 @@ 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) {
@@ -316,7 +338,10 @@ 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 +433,17 @@ 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 +453,19 @@ 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,
@@ -466,6 +511,10 @@ Error Message: ${error.message}`);
         logAxiosError({ error, message });
         return returnValue(`Something went wrong when trying to edit the image. The OpenAI API may be unavailable:
 Error Message: ${error.message || 'Unknown error'}`);
+      } finally {
+        if (abortHandler && derivedSignal) {
+          derivedSignal.removeEventListener('abort', abortHandler);
+        }
       }
     },
     {

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

@@ -0,0 +1,94 @@
+const DALLE3 = require('../DALLE3');
+const { ProxyAgent } = require('undici');
+
+const processFileURL = jest.fn();
+
+jest.mock('~/server/services/Files/images', () => ({
+  getImageBasename: jest.fn().mockImplementation((url) => {
+    const parts = url.split('/');
+    const lastPart = parts.pop();
+    const imageExtensionRegex = /\.(jpg|jpeg|png|gif|bmp|tiff|svg)$/i;
+    if (imageExtensionRegex.test(lastPart)) {
+      return lastPart;
+    }
+    return '';
+  }),
+}));
+
+jest.mock('fs', () => {
+  return {
+    existsSync: jest.fn(),
+    mkdirSync: jest.fn(),
+    promises: {
+      writeFile: jest.fn(),
+      readFile: jest.fn(),
+      unlink: jest.fn(),
+    },
+  };
+});
+
+jest.mock('path', () => {
+  return {
+    resolve: jest.fn(),
+    join: jest.fn(),
+    relative: jest.fn(),
+    extname: jest.fn().mockImplementation((filename) => {
+      return filename.slice(filename.lastIndexOf('.'));
+    }),
+  };
+});
+
+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

@@ -1,10 +1,29 @@
 const OpenAI = require('openai');
 const DALLE3 = require('../DALLE3');
-
-const { logger } = require('~/config');
+const logger = require('~/config/winston');
 
 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', () => ({
@@ -37,6 +56,11 @@ 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/fileSearch.js

@@ -1,26 +1,35 @@
 const { z } = require('zod');
 const axios = require('axios');
 const { tool } = require('@langchain/core/tools');
+const { logger } = require('@librechat/data-schemas');
 const { Tools, EToolResources } = require('librechat-data-provider');
+const { generateShortLivedToken } = require('~/server/services/AuthService');
 const { getFiles } = require('~/models/File');
-const { logger } = require('~/config');
 
 /**
  *
  * @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);
+  const dbFiles = (
+    (await getFiles(
+      { file_id: { $in: file_ids } },
+      null,
+      { text: 0 },
+      { userId: req?.user?.id, agentId },
+    )) ?? []
+  ).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.`;
 
@@ -59,7 +68,7 @@ const createFileSearchTool = async ({ req, files, entity_id }) => {
       if (files.length === 0) {
         return 'No files to search. Instruct the user to add files for the search.';
       }
-      const jwtToken = req.headers.authorization.split(' ')[1];
+      const jwtToken = generateShortLivedToken(req.user.id);
       if (!jwtToken) {
         return 'There was an error authenticating the file search request.';
       }
@@ -135,7 +144,7 @@ const createFileSearchTool = async ({ req, files, entity_id }) => {
         query: z
           .string()
           .describe(
-            'A natural language query to search for relevant information in the files. Be specific and use keywords related to the information you\'re looking for. The query will be used for semantic similarity matching against the file contents.',
+            "A natural language query to search for relevant information in the files. Be specific and use keywords related to the information you're looking for. The query will be used for semantic similarity matching against the file contents.",
           ),
       }),
     },

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

@@ -1,14 +1,9 @@
+const { logger } = require('@librechat/data-schemas');
 const { SerpAPI } = require('@langchain/community/tools/serpapi');
 const { Calculator } = require('@langchain/community/tools/calculator');
+const { mcpToolPattern, loadWebSearchAuth } = require('@librechat/api');
 const { EnvVar, createCodeExecutionTool, createSearchTool } = require('@librechat/agents');
-const {
-  Tools,
-  Constants,
-  EToolResources,
-  loadWebSearchAuth,
-  replaceSpecialVars,
-} = require('librechat-data-provider');
-const { getUserPluginAuthValue } = require('~/server/services/PluginService');
+const { Tools, EToolResources, replaceSpecialVars } = require('librechat-data-provider');
 const {
   availableTools,
   manifestToolMap,
@@ -28,11 +23,10 @@ const {
 } = require('../');
 const { primeFiles: primeCodeFiles } = require('~/server/services/Files/Code/process');
 const { createFileSearchTool, primeFiles: primeSearchFiles } = require('./fileSearch');
+const { getUserPluginAuthValue } = require('~/server/services/PluginService');
 const { loadAuthValues } = require('~/server/services/Tools/credentials');
+const { getCachedTools } = require('~/server/services/Config');
 const { createMCPTool } = require('~/server/services/MCP');
-const { logger } = require('~/config');
-
-const mcpToolPattern = new RegExp(`^.+${Constants.mcp_delimiter}.+$`);
 
 /**
  * Validates the availability and authentication of tools for a user based on environment variables or user-specific plugin authentication values.
@@ -93,7 +87,7 @@ 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);
   }
 };
 
@@ -236,7 +230,7 @@ const loadTools = async ({
 
   /** @type {Record<string, string>} */
   const toolContextMap = {};
-  const appTools = options.req?.app?.locals?.availableTools ?? {};
+  const cachedTools = (await getCachedTools({ userId: user, includeGlobal: true })) ?? {};
 
   for (const tool of tools) {
     if (tool === Tools.execute_code) {
@@ -246,7 +240,13 @@ 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;
         }
@@ -261,7 +261,10 @@ 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;
         }
@@ -295,10 +298,11 @@ Current Date & Time: ${replaceSpecialVars({ text: '{{iso_datetime}}' })}
         });
       };
       continue;
-    } else if (tool && appTools[tool] && mcpToolPattern.test(tool)) {
+    } else if (tool && cachedTools && mcpToolPattern.test(tool)) {
       requestedTools[tool] = async () =>
         createMCPTool({
           req: options.req,
+          res: options.res,
           toolKey: tool,
           model: agent?.model ?? model,
           provider: agent?.provider ?? endpoint,

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

@@ -1,8 +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(),
@@ -10,29 +7,18 @@ const mockPluginService = {
   getUserPluginAuthValue: jest.fn(),
 };
 
-const mockModels = {
-  User: mockUser,
-};
-jest.mock('~/db/connect', () => {
-  return {
-    connectDb: jest.fn(),
-    User: mockModels.mockUser,
-  };
-});
-jest.mock('~/models/File', () => ({
-  File: jest.fn(),
-}));
-
 jest.mock('~/server/services/PluginService', () => mockPluginService);
 
 const { BaseLLM } = require('@langchain/openai');
 const { Calculator } = require('@langchain/community/tools/calculator');
 
+const { User } = require('~/db/models');
 const PluginService = require('~/server/services/PluginService');
 const { validateTools, loadTools, loadToolWithAuth } = require('./handleTools');
 const { StructuredSD, availableTools, DALLE3 } = require('../');
 
 describe('Tool Handlers', () => {
+  let mongoServer;
   let fakeUser;
   const pluginKey = 'dalle';
   const pluginKey2 = 'wolfram';
@@ -43,7 +29,9 @@ 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) => {
@@ -58,7 +46,7 @@ describe('Tool Handlers', () => {
       },
     );
 
-    fakeUser = await mockModels.User.createUser({
+    fakeUser = new User({
       name: 'Fake User',
       username: 'fakeuser',
       email: 'fakeuser@example.com',
@@ -84,9 +72,36 @@ 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,
+      );
     }
   });
 

api/cache/banViolation.js

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

api/cache/banViolation.spec.js

@@ -1,69 +1,28 @@
+const mongoose = require('mongoose');
+const { MongoMemoryServer } = require('mongodb-memory-server');
 const banViolation = require('./banViolation');
 
-const mockModels = {
-  Session: {
-    deleteAllUserSessions: jest.fn(),
-  },
-};
-
-jest.mock('~/db/connect', () => {
-  return {
-    connectDb: jest.fn(),
-    get models() {
-      return mockModels;
-    },
-  };
-});
-
-jest.mock('~/server/utils', () => ({
-  isEnabled: jest.fn(() => true), // default to false, override per test if needed
-  math: jest.fn(() => 20), // default to false, override per test if needed
-  removePorts: jest.fn(),
+// Mock deleteAllUserSessions since we're testing ban logic, not session deletion
+jest.mock('~/models', () => ({
+  ...jest.requireActual('~/models'),
+  deleteAllUserSessions: jest.fn().mockResolvedValue(true),
 }));
 
-jest.mock('keyv');
-// jest.mock('../models/Session');
-// Mocking the getLogStores function
-jest.mock('./getLogStores', () => {
-  return jest.fn().mockImplementation(() => {
-    const EventEmitter = require('events');
-    const { CacheKeys } = require('librechat-data-provider');
-    const math = require('../server/utils/math');
-    const mockGet = jest.fn();
-    const mockSet = jest.fn();
-    class KeyvMongo extends EventEmitter {
-      constructor(url = 'mongodb://127.0.0.1:27017', options) {
-        super();
-        this.ttlSupport = false;
-        url = url ?? {};
-        if (typeof url === 'string') {
-          url = { url };
-        }
-        if (url.uri) {
-          url = { url: url.uri, ...url };
-        }
-        this.opts = {
-          url,
-          collection: 'keyv',
-          ...url,
-          ...options,
-        };
-      }
-
-      get = mockGet;
-      set = mockSet;
-    }
-
-    return new KeyvMongo('', {
-      namespace: CacheKeys.BANS,
-      ttl: math(process.env.BAN_DURATION, 7200000),
-    });
-  });
-});
-
 describe('banViolation', () => {
+  let mongoServer;
   let req, res, errorMessage;
 
+  beforeAll(async () => {
+    mongoServer = await MongoMemoryServer.create();
+    const mongoUri = mongoServer.getUri();
+    await mongoose.connect(mongoUri);
+  });
+
+  afterAll(async () => {
+    await mongoose.disconnect();
+    await mongoServer.stop();
+  });
+
   beforeEach(() => {
     req = {
       ip: '127.0.0.1',
@@ -76,7 +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 +1,62 @@
+const fs = require('fs');
+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(', ')}`,
+    );
+  }
+}
+
+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: process.env.REDIS_CA ? fs.readFileSync(process.env.REDIS_CA, 'utf8') : null,
+  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'),
+
+  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 +1,157 @@
+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.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('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, STATIC_CONFIG ,MESSAGES ';
+
+      const { cacheConfig } = require('./cacheConfig');
+      expect(cacheConfig.FORCED_IN_MEMORY_CACHE_NAMESPACES).toEqual([
+        'ROLES',
+        'STATIC_CONFIG',
+        '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/getLogStores.js

@@ -1,108 +1,53 @@
+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 openIdExchangedTokensCache = isRedisEnabled
-  ? new Keyv({ store: keyvRedis, ttl: Time.TEN_MINUTES })
-  : new Keyv({ namespace: CacheKeys.OPENID_EXCHANGED_TOKENS, 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.MCP_TOOLS]: standardCache(CacheKeys.MCP_TOOLS),
+  [CacheKeys.CONFIG_STORE]: standardCache(CacheKeys.CONFIG_STORE),
+  [CacheKeys.STATIC_CONFIG]: standardCache(CacheKeys.STATIC_CONFIG),
+  [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,
-  [CacheKeys.OPENID_EXCHANGED_TOKENS]: openIdExchangedTokensCache,
 };
 
 /**
@@ -111,7 +56,10 @@ 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
   );
 }
 
@@ -147,18 +95,18 @@ 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;
       }
@@ -167,7 +115,7 @@ 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}`,
       );
@@ -208,7 +156,7 @@ async function clearAllExpiredFromCache() {
   }
 }
 
-if (!isRedisEnabled && !isEnabled(CI)) {
+if (!cacheConfig.USE_REDIS && !cacheConfig.CI) {
   /** @type {Set<NodeJS.Timeout>} */
   const cleanupIntervals = new Set();
 
@@ -219,7 +167,7 @@ if (!isRedisEnabled && !isEnabled(CI)) {
 
   cleanupIntervals.add(cleanup);
 
-  if (debugMemoryCache) {
+  if (cacheConfig.DEBUG_MEMORY_CACHE) {
     const monitor = setInterval(() => {
       const ttlStores = getTTLStores();
       const memory = process.memoryUsage();
@@ -240,13 +188,13 @@ 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/keyvRedis.js

@@ -1,109 +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 { 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 +1,204 @@
+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 true;
+      }
+      return false;
+    },
+    enableOfflineQueue: cacheConfig.REDIS_ENABLE_OFFLINE_QUEUE,
+    connectTimeout: cacheConfig.REDIS_CONNECT_TIMEOUT,
+    maxRetriesPerRequest: 3,
+  };
+
+  ioredisClient =
+    urls.length === 1
+      ? new IoRedis(cacheConfig.REDIS_URI, redisOptions)
+      : new IoRedis.Cluster(cacheConfig.REDIS_URI, {
+          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
+      ? createClient({ url: cacheConfig.REDIS_URI, ...redisOptions })
+      : createCluster({
+          rootNodes: cacheConfig.REDIS_URI.split(',').map((url) => ({ url })),
+          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

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

api/db/indexSync.js

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

api/models/Action.js

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

api/models/Agent.js

@@ -11,10 +11,10 @@ const {
   removeAgentIdsFromProject,
   removeAgentFromAllProjects,
 } = require('./Project');
+const { getCachedTools } = require('~/server/services/Config');
 const getLogStores = require('~/cache/getLogStores');
 const { getActions } = require('./Action');
-
-const Agent = require('~/db/models').Agent;
+const { Agent } = require('~/db/models');
 
 /**
  * Create an agent with the provided data.
@@ -56,12 +56,12 @@ const getAgent = async (searchParameter) => await Agent.findOne(searchParameter)
  * @param {string} params.agent_id
  * @param {string} params.endpoint
  * @param {import('@librechat/agents').ClientOptions} [params.model_parameters]
- * @returns {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 availableTools = await getCachedTools({ userId: req.user.id, includeGlobal: true });
   /** @type {TEphemeralAgent | null} */
   const ephemeralAgent = req.body.ephemeralAgent;
   const mcpServers = new Set(ephemeralAgent?.mcp);
@@ -70,6 +70,9 @@ const loadEphemeralAgent = ({ req, agent_id, endpoint, model_parameters: _m }) =
   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);
   }
@@ -87,7 +90,7 @@ const loadEphemeralAgent = ({ req, agent_id, endpoint, model_parameters: _m }) =
   }
 
   const instructions = req.body.promptPrefix;
-  return {
+  const result = {
     id: agent_id,
     instructions,
     provider: endpoint,
@@ -95,6 +98,11 @@ const loadEphemeralAgent = ({ req, agent_id, endpoint, model_parameters: _m }) =
     model,
     tools,
   };
+
+  if (ephemeralAgent?.artifacts != null && ephemeralAgent.artifacts) {
+    result.artifacts = ephemeralAgent.artifacts;
+  }
+  return result;
 };
 
 /**
@@ -112,7 +120,7 @@ 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,
@@ -171,7 +179,6 @@ const isDuplicateVersion = (updateData, currentData, versions, actionsHash = nul
     'created_at',
     'updated_at',
     '__v',
-    'agent_ids',
     'versions',
     'actionsHash', // Exclude actionsHash from direct comparison
   ];
@@ -261,11 +268,12 @@ const isDuplicateVersion = (updateData, currentData, versions, actionsHash = nul
  * @param {Object} [options] - Optional configuration object.
  * @param {string} [options.updatingUserId] - The ID of the user performing the update (used for tracking non-author updates).
  * @param {boolean} [options.forceVersion] - Force creation of a new version even if no fields changed.
+ * @param {boolean} [options.skipVersioning] - Skip version creation entirely (useful for isolated operations like sharing).
  * @returns {Promise<Agent>} The updated or newly created agent document as a plain object.
  * @throws {Error} If the update would create a duplicate version
  */
 const updateAgent = async (searchParameter, updateData, options = {}) => {
-  const { updatingUserId = null, forceVersion = false } = options;
+  const { updatingUserId = null, forceVersion = false, skipVersioning = false } = options;
   const mongoOptions = { new: true, upsert: false };
 
   const currentAgent = await Agent.findOne(searchParameter);
@@ -302,25 +310,16 @@ const updateAgent = async (searchParameter, updateData, options = {}) => {
     }
 
     const shouldCreateVersion =
-      forceVersion ||
-      (versions &&
-        versions.length > 0 &&
-        (Object.keys(directUpdates).length > 0 || $push || $pull || $addToSet));
+      !skipVersioning &&
+      (forceVersion || Object.keys(directUpdates).length > 0 || $push || $pull || $addToSet);
 
     if (shouldCreateVersion) {
       const duplicateVersion = isDuplicateVersion(updateData, versionData, versions, actionsHash);
       if (duplicateVersion && !forceVersion) {
-        const error = new Error(
-          'Duplicate version: This would create a version identical to an existing one',
-        );
-        error.statusCode = 409;
-        error.details = {
-          duplicateVersion,
-          versionIndex: versions.findIndex(
-            (v) => JSON.stringify(duplicateVersion) === JSON.stringify(v),
-          ),
-        };
-        throw error;
+        // No changes detected, return the current agent without creating a new version
+        const agentObj = currentAgent.toObject();
+        agentObj.version = versions.length;
+        return agentObj;
       }
     }
 
@@ -340,7 +339,7 @@ const updateAgent = async (searchParameter, updateData, options = {}) => {
       versionEntry.updatedBy = new mongoose.Types.ObjectId(updatingUserId);
     }
 
-    if (shouldCreateVersion || forceVersion) {
+    if (shouldCreateVersion) {
       updateData.$push = {
         ...($push || {}),
         versions: versionEntry,
@@ -551,7 +550,10 @@ const updateAgentProjects = async ({ user, agentId, projectIds, removeProjectIds
     delete updateQuery.author;
   }
 
-  const updatedAgent = await updateAgent(updateQuery, updateOps, { updatingUserId: user.id });
+  const updatedAgent = await updateAgent(updateQuery, updateOps, {
+    updatingUserId: user.id,
+    skipVersioning: true,
+  });
   if (updatedAgent) {
     return updatedAgent;
   }

api/models/Assistant.js

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

api/models/Banner.js

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

api/models/Conversation.js

@@ -1,8 +1,8 @@
-const mongoose = require('mongoose');
 const { logger } = require('@librechat/data-schemas');
+const { createTempChatExpirationDate } = require('@librechat/api');
+const { getCustomConfig } = require('~/server/services/Config/getCustomConfig');
 const { getMessages, deleteMessages } = require('./Message');
-
-const Conversation = require('~/db/models').Conversation;
+const { Conversation } = require('~/db/models');
 
 /**
  * Searches for a conversation by conversationId and returns a lean document with only conversationId and user.
@@ -100,10 +100,15 @@ module.exports = {
         update.conversationId = newConversationId;
       }
 
-      if (req.body.isTemporary) {
-        const expiredAt = new Date();
-        expiredAt.setDate(expiredAt.getDate() + 30);
-        update.expiredAt = expiredAt;
+      if (req?.body?.isTemporary) {
+        try {
+          const customConfig = await getCustomConfig();
+          update.expiredAt = createTempChatExpirationDate(customConfig);
+        } catch (err) {
+          logger.error('Error creating temporary chat expiration date:', err);
+          logger.info(`---\`saveConvo\` context: ${metadata?.context}`);
+          update.expiredAt = null;
+        }
       } else {
         update.expiredAt = null;
       }

api/models/Conversation.spec.js

@@ -0,0 +1,572 @@
+const mongoose = require('mongoose');
+const { v4: uuidv4 } = require('uuid');
+const { EModelEndpoint } = require('librechat-data-provider');
+const { MongoMemoryServer } = require('mongodb-memory-server');
+const {
+  deleteNullOrEmptyConversations,
+  searchConversation,
+  getConvosByCursor,
+  getConvosQueried,
+  getConvoFiles,
+  getConvoTitle,
+  deleteConvos,
+  saveConvo,
+  getConvo,
+} = require('./Conversation');
+jest.mock('~/server/services/Config/getCustomConfig');
+jest.mock('./Message');
+const { getCustomConfig } = require('~/server/services/Config/getCustomConfig');
+const { getMessages, deleteMessages } = require('./Message');
+
+const { Conversation } = require('~/db/models');
+
+describe('Conversation Operations', () => {
+  let mongoServer;
+  let mockReq;
+  let mockConversationData;
+
+  beforeAll(async () => {
+    mongoServer = await MongoMemoryServer.create();
+    const mongoUri = mongoServer.getUri();
+    await mongoose.connect(mongoUri);
+  });
+
+  afterAll(async () => {
+    await mongoose.disconnect();
+    await mongoServer.stop();
+  });
+
+  beforeEach(async () => {
+    // Clear database
+    await Conversation.deleteMany({});
+
+    // Reset mocks
+    jest.clearAllMocks();
+
+    // Default mock implementations
+    getMessages.mockResolvedValue([]);
+    deleteMessages.mockResolvedValue({ deletedCount: 0 });
+
+    mockReq = {
+      user: { id: 'user123' },
+      body: {},
+    };
+
+    mockConversationData = {
+      conversationId: uuidv4(),
+      title: 'Test Conversation',
+      endpoint: EModelEndpoint.openAI,
+    };
+  });
+
+  describe('saveConvo', () => {
+    it('should save a conversation for an authenticated user', async () => {
+      const result = await saveConvo(mockReq, mockConversationData);
+
+      expect(result.conversationId).toBe(mockConversationData.conversationId);
+      expect(result.user).toBe('user123');
+      expect(result.title).toBe('Test Conversation');
+      expect(result.endpoint).toBe(EModelEndpoint.openAI);
+
+      // Verify the conversation was actually saved to the database
+      const savedConvo = await Conversation.findOne({
+        conversationId: mockConversationData.conversationId,
+        user: 'user123',
+      });
+      expect(savedConvo).toBeTruthy();
+      expect(savedConvo.title).toBe('Test Conversation');
+    });
+
+    it('should query messages when saving a conversation', async () => {
+      // Mock messages as ObjectIds
+      const mongoose = require('mongoose');
+      const mockMessages = [new mongoose.Types.ObjectId(), new mongoose.Types.ObjectId()];
+      getMessages.mockResolvedValue(mockMessages);
+
+      await saveConvo(mockReq, mockConversationData);
+
+      // Verify that getMessages was called with correct parameters
+      expect(getMessages).toHaveBeenCalledWith(
+        { conversationId: mockConversationData.conversationId },
+        '_id',
+      );
+    });
+
+    it('should handle newConversationId when provided', async () => {
+      const newConversationId = uuidv4();
+      const result = await saveConvo(mockReq, {
+        ...mockConversationData,
+        newConversationId,
+      });
+
+      expect(result.conversationId).toBe(newConversationId);
+    });
+
+    it('should handle unsetFields metadata', async () => {
+      const metadata = {
+        unsetFields: { someField: 1 },
+      };
+
+      await saveConvo(mockReq, mockConversationData, metadata);
+
+      const savedConvo = await Conversation.findOne({
+        conversationId: mockConversationData.conversationId,
+      });
+      expect(savedConvo.someField).toBeUndefined();
+    });
+  });
+
+  describe('isTemporary conversation handling', () => {
+    it('should save a conversation with expiredAt when isTemporary is true', async () => {
+      // Mock custom config with 24 hour retention
+      getCustomConfig.mockResolvedValue({
+        interface: {
+          temporaryChatRetention: 24,
+        },
+      });
+
+      mockReq.body = { isTemporary: true };
+
+      const beforeSave = new Date();
+      const result = await saveConvo(mockReq, mockConversationData);
+      const afterSave = new Date();
+
+      expect(result.conversationId).toBe(mockConversationData.conversationId);
+      expect(result.expiredAt).toBeDefined();
+      expect(result.expiredAt).toBeInstanceOf(Date);
+
+      // Verify expiredAt is approximately 24 hours in the future
+      const expectedExpirationTime = new Date(beforeSave.getTime() + 24 * 60 * 60 * 1000);
+      const actualExpirationTime = new Date(result.expiredAt);
+
+      expect(actualExpirationTime.getTime()).toBeGreaterThanOrEqual(
+        expectedExpirationTime.getTime() - 1000,
+      );
+      expect(actualExpirationTime.getTime()).toBeLessThanOrEqual(
+        new Date(afterSave.getTime() + 24 * 60 * 60 * 1000 + 1000).getTime(),
+      );
+    });
+
+    it('should save a conversation without expiredAt when isTemporary is false', async () => {
+      mockReq.body = { isTemporary: false };
+
+      const result = await saveConvo(mockReq, mockConversationData);
+
+      expect(result.conversationId).toBe(mockConversationData.conversationId);
+      expect(result.expiredAt).toBeNull();
+    });
+
+    it('should save a conversation without expiredAt when isTemporary is not provided', async () => {
+      // No isTemporary in body
+      mockReq.body = {};
+
+      const result = await saveConvo(mockReq, mockConversationData);
+
+      expect(result.conversationId).toBe(mockConversationData.conversationId);
+      expect(result.expiredAt).toBeNull();
+    });
+
+    it('should use custom retention period from config', async () => {
+      // Mock custom config with 48 hour retention
+      getCustomConfig.mockResolvedValue({
+        interface: {
+          temporaryChatRetention: 48,
+        },
+      });
+
+      mockReq.body = { isTemporary: true };
+
+      const beforeSave = new Date();
+      const result = await saveConvo(mockReq, mockConversationData);
+
+      expect(result.expiredAt).toBeDefined();
+
+      // Verify expiredAt is approximately 48 hours in the future
+      const expectedExpirationTime = new Date(beforeSave.getTime() + 48 * 60 * 60 * 1000);
+      const actualExpirationTime = new Date(result.expiredAt);
+
+      expect(actualExpirationTime.getTime()).toBeGreaterThanOrEqual(
+        expectedExpirationTime.getTime() - 1000,
+      );
+      expect(actualExpirationTime.getTime()).toBeLessThanOrEqual(
+        expectedExpirationTime.getTime() + 1000,
+      );
+    });
+
+    it('should handle minimum retention period (1 hour)', async () => {
+      // Mock custom config with less than minimum retention
+      getCustomConfig.mockResolvedValue({
+        interface: {
+          temporaryChatRetention: 0.5, // Half hour - should be clamped to 1 hour
+        },
+      });
+
+      mockReq.body = { isTemporary: true };
+
+      const beforeSave = new Date();
+      const result = await saveConvo(mockReq, mockConversationData);
+
+      expect(result.expiredAt).toBeDefined();
+
+      // Verify expiredAt is approximately 1 hour in the future (minimum)
+      const expectedExpirationTime = new Date(beforeSave.getTime() + 1 * 60 * 60 * 1000);
+      const actualExpirationTime = new Date(result.expiredAt);
+
+      expect(actualExpirationTime.getTime()).toBeGreaterThanOrEqual(
+        expectedExpirationTime.getTime() - 1000,
+      );
+      expect(actualExpirationTime.getTime()).toBeLessThanOrEqual(
+        expectedExpirationTime.getTime() + 1000,
+      );
+    });
+
+    it('should handle maximum retention period (8760 hours)', async () => {
+      // Mock custom config with more than maximum retention
+      getCustomConfig.mockResolvedValue({
+        interface: {
+          temporaryChatRetention: 10000, // Should be clamped to 8760 hours
+        },
+      });
+
+      mockReq.body = { isTemporary: true };
+
+      const beforeSave = new Date();
+      const result = await saveConvo(mockReq, mockConversationData);
+
+      expect(result.expiredAt).toBeDefined();
+
+      // Verify expiredAt is approximately 8760 hours (1 year) in the future
+      const expectedExpirationTime = new Date(beforeSave.getTime() + 8760 * 60 * 60 * 1000);
+      const actualExpirationTime = new Date(result.expiredAt);
+
+      expect(actualExpirationTime.getTime()).toBeGreaterThanOrEqual(
+        expectedExpirationTime.getTime() - 1000,
+      );
+      expect(actualExpirationTime.getTime()).toBeLessThanOrEqual(
+        expectedExpirationTime.getTime() + 1000,
+      );
+    });
+
+    it('should handle getCustomConfig errors gracefully', async () => {
+      // Mock getCustomConfig to throw an error
+      getCustomConfig.mockRejectedValue(new Error('Config service unavailable'));
+
+      mockReq.body = { isTemporary: true };
+
+      const result = await saveConvo(mockReq, mockConversationData);
+
+      // Should still save the conversation but with expiredAt as null
+      expect(result.conversationId).toBe(mockConversationData.conversationId);
+      expect(result.expiredAt).toBeNull();
+    });
+
+    it('should use default retention when config is not provided', async () => {
+      // Mock getCustomConfig to return empty config
+      getCustomConfig.mockResolvedValue({});
+
+      mockReq.body = { isTemporary: true };
+
+      const beforeSave = new Date();
+      const result = await saveConvo(mockReq, mockConversationData);
+
+      expect(result.expiredAt).toBeDefined();
+
+      // Default retention is 30 days (720 hours)
+      const expectedExpirationTime = new Date(beforeSave.getTime() + 30 * 24 * 60 * 60 * 1000);
+      const actualExpirationTime = new Date(result.expiredAt);
+
+      expect(actualExpirationTime.getTime()).toBeGreaterThanOrEqual(
+        expectedExpirationTime.getTime() - 1000,
+      );
+      expect(actualExpirationTime.getTime()).toBeLessThanOrEqual(
+        expectedExpirationTime.getTime() + 1000,
+      );
+    });
+
+    it('should update expiredAt when saving existing temporary conversation', async () => {
+      // First save a temporary conversation
+      getCustomConfig.mockResolvedValue({
+        interface: {
+          temporaryChatRetention: 24,
+        },
+      });
+
+      mockReq.body = { isTemporary: true };
+      const firstSave = await saveConvo(mockReq, mockConversationData);
+      const originalExpiredAt = firstSave.expiredAt;
+
+      // Wait a bit to ensure time difference
+      await new Promise((resolve) => setTimeout(resolve, 100));
+
+      // Save again with same conversationId but different title
+      const updatedData = { ...mockConversationData, title: 'Updated Title' };
+      const secondSave = await saveConvo(mockReq, updatedData);
+
+      // Should update title and create new expiredAt
+      expect(secondSave.title).toBe('Updated Title');
+      expect(secondSave.expiredAt).toBeDefined();
+      expect(new Date(secondSave.expiredAt).getTime()).toBeGreaterThan(
+        new Date(originalExpiredAt).getTime(),
+      );
+    });
+
+    it('should not set expiredAt when updating non-temporary conversation', async () => {
+      // First save a non-temporary conversation
+      mockReq.body = { isTemporary: false };
+      const firstSave = await saveConvo(mockReq, mockConversationData);
+      expect(firstSave.expiredAt).toBeNull();
+
+      // Update without isTemporary flag
+      mockReq.body = {};
+      const updatedData = { ...mockConversationData, title: 'Updated Title' };
+      const secondSave = await saveConvo(mockReq, updatedData);
+
+      expect(secondSave.title).toBe('Updated Title');
+      expect(secondSave.expiredAt).toBeNull();
+    });
+
+    it('should filter out expired conversations in getConvosByCursor', async () => {
+      // Create some test conversations
+      const nonExpiredConvo = await Conversation.create({
+        conversationId: uuidv4(),
+        user: 'user123',
+        title: 'Non-expired',
+        endpoint: EModelEndpoint.openAI,
+        expiredAt: null,
+        updatedAt: new Date(),
+      });
+
+      await Conversation.create({
+        conversationId: uuidv4(),
+        user: 'user123',
+        title: 'Future expired',
+        endpoint: EModelEndpoint.openAI,
+        expiredAt: new Date(Date.now() + 24 * 60 * 60 * 1000), // 24 hours from now
+        updatedAt: new Date(),
+      });
+
+      // Mock Meili search
+      Conversation.meiliSearch = jest.fn().mockResolvedValue({ hits: [] });
+
+      const result = await getConvosByCursor('user123');
+
+      // Should only return conversations with null or non-existent expiredAt
+      expect(result.conversations).toHaveLength(1);
+      expect(result.conversations[0].conversationId).toBe(nonExpiredConvo.conversationId);
+    });
+
+    it('should filter out expired conversations in getConvosQueried', async () => {
+      // Create test conversations
+      const nonExpiredConvo = await Conversation.create({
+        conversationId: uuidv4(),
+        user: 'user123',
+        title: 'Non-expired',
+        endpoint: EModelEndpoint.openAI,
+        expiredAt: null,
+      });
+
+      const expiredConvo = await Conversation.create({
+        conversationId: uuidv4(),
+        user: 'user123',
+        title: 'Expired',
+        endpoint: EModelEndpoint.openAI,
+        expiredAt: new Date(Date.now() + 24 * 60 * 60 * 1000),
+      });
+
+      const convoIds = [
+        { conversationId: nonExpiredConvo.conversationId },
+        { conversationId: expiredConvo.conversationId },
+      ];
+
+      const result = await getConvosQueried('user123', convoIds);
+
+      // Should only return the non-expired conversation
+      expect(result.conversations).toHaveLength(1);
+      expect(result.conversations[0].conversationId).toBe(nonExpiredConvo.conversationId);
+      expect(result.convoMap[nonExpiredConvo.conversationId]).toBeDefined();
+      expect(result.convoMap[expiredConvo.conversationId]).toBeUndefined();
+    });
+  });
+
+  describe('searchConversation', () => {
+    it('should find a conversation by conversationId', async () => {
+      await Conversation.create({
+        conversationId: mockConversationData.conversationId,
+        user: 'user123',
+        title: 'Test',
+        endpoint: EModelEndpoint.openAI,
+      });
+
+      const result = await searchConversation(mockConversationData.conversationId);
+
+      expect(result).toBeTruthy();
+      expect(result.conversationId).toBe(mockConversationData.conversationId);
+      expect(result.user).toBe('user123');
+      expect(result.title).toBeUndefined(); // Only returns conversationId and user
+    });
+
+    it('should return null if conversation not found', async () => {
+      const result = await searchConversation('non-existent-id');
+      expect(result).toBeNull();
+    });
+  });
+
+  describe('getConvo', () => {
+    it('should retrieve a conversation for a user', async () => {
+      await Conversation.create({
+        conversationId: mockConversationData.conversationId,
+        user: 'user123',
+        title: 'Test Conversation',
+        endpoint: EModelEndpoint.openAI,
+      });
+
+      const result = await getConvo('user123', mockConversationData.conversationId);
+
+      expect(result.conversationId).toBe(mockConversationData.conversationId);
+      expect(result.user).toBe('user123');
+      expect(result.title).toBe('Test Conversation');
+    });
+
+    it('should return null if conversation not found', async () => {
+      const result = await getConvo('user123', 'non-existent-id');
+      expect(result).toBeNull();
+    });
+  });
+
+  describe('getConvoTitle', () => {
+    it('should return the conversation title', async () => {
+      await Conversation.create({
+        conversationId: mockConversationData.conversationId,
+        user: 'user123',
+        title: 'Test Title',
+        endpoint: EModelEndpoint.openAI,
+      });
+
+      const result = await getConvoTitle('user123', mockConversationData.conversationId);
+      expect(result).toBe('Test Title');
+    });
+
+    it('should return null if conversation has no title', async () => {
+      await Conversation.create({
+        conversationId: mockConversationData.conversationId,
+        user: 'user123',
+        title: null,
+        endpoint: EModelEndpoint.openAI,
+      });
+
+      const result = await getConvoTitle('user123', mockConversationData.conversationId);
+      expect(result).toBeNull();
+    });
+
+    it('should return "New Chat" if conversation not found', async () => {
+      const result = await getConvoTitle('user123', 'non-existent-id');
+      expect(result).toBe('New Chat');
+    });
+  });
+
+  describe('getConvoFiles', () => {
+    it('should return conversation files', async () => {
+      const files = ['file1', 'file2'];
+      await Conversation.create({
+        conversationId: mockConversationData.conversationId,
+        user: 'user123',
+        endpoint: EModelEndpoint.openAI,
+        files,
+      });
+
+      const result = await getConvoFiles(mockConversationData.conversationId);
+      expect(result).toEqual(files);
+    });
+
+    it('should return empty array if no files', async () => {
+      await Conversation.create({
+        conversationId: mockConversationData.conversationId,
+        user: 'user123',
+        endpoint: EModelEndpoint.openAI,
+      });
+
+      const result = await getConvoFiles(mockConversationData.conversationId);
+      expect(result).toEqual([]);
+    });
+
+    it('should return empty array if conversation not found', async () => {
+      const result = await getConvoFiles('non-existent-id');
+      expect(result).toEqual([]);
+    });
+  });
+
+  describe('deleteConvos', () => {
+    it('should delete conversations and associated messages', async () => {
+      await Conversation.create({
+        conversationId: mockConversationData.conversationId,
+        user: 'user123',
+        title: 'To Delete',
+        endpoint: EModelEndpoint.openAI,
+      });
+
+      deleteMessages.mockResolvedValue({ deletedCount: 5 });
+
+      const result = await deleteConvos('user123', {
+        conversationId: mockConversationData.conversationId,
+      });
+
+      expect(result.deletedCount).toBe(1);
+      expect(result.messages.deletedCount).toBe(5);
+      expect(deleteMessages).toHaveBeenCalledWith({
+        conversationId: { $in: [mockConversationData.conversationId] },
+      });
+
+      // Verify conversation was deleted
+      const deletedConvo = await Conversation.findOne({
+        conversationId: mockConversationData.conversationId,
+      });
+      expect(deletedConvo).toBeNull();
+    });
+
+    it('should throw error if no conversations found', async () => {
+      await expect(deleteConvos('user123', { conversationId: 'non-existent' })).rejects.toThrow(
+        'Conversation not found or already deleted.',
+      );
+    });
+  });
+
+  describe('deleteNullOrEmptyConversations', () => {
+    it('should delete conversations with null, empty, or missing conversationIds', async () => {
+      // Since conversationId is required by the schema, we can't create documents with null/missing IDs
+      // This test should verify the function works when such documents exist (e.g., from data corruption)
+
+      // For this test, let's create a valid conversation and verify the function doesn't delete it
+      await Conversation.create({
+        conversationId: mockConversationData.conversationId,
+        user: 'user4',
+        endpoint: EModelEndpoint.openAI,
+      });
+
+      deleteMessages.mockResolvedValue({ deletedCount: 0 });
+
+      const result = await deleteNullOrEmptyConversations();
+
+      expect(result.conversations.deletedCount).toBe(0); // No invalid conversations to delete
+      expect(result.messages.deletedCount).toBe(0);
+
+      // Verify valid conversation remains
+      const remainingConvos = await Conversation.find({});
+      expect(remainingConvos).toHaveLength(1);
+      expect(remainingConvos[0].conversationId).toBe(mockConversationData.conversationId);
+    });
+  });
+
+  describe('Error Handling', () => {
+    it('should handle database errors in saveConvo', async () => {
+      // Force a database error by disconnecting
+      await mongoose.disconnect();
+
+      const result = await saveConvo(mockReq, mockConversationData);
+
+      expect(result).toEqual({ message: 'Error saving conversation' });
+
+      // Reconnect for other tests
+      await mongoose.connect(mongoServer.getUri());
+    });
+  });
+});

api/models/ConversationTag.js

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

api/models/File.js

@@ -1,8 +1,8 @@
-const mongoose = require('mongoose');
 const { logger } = require('@librechat/data-schemas');
-const { EToolResources } = require('librechat-data-provider');
-
-const File = require('~/db/models').File;
+const { EToolResources, FileContext, Constants } = require('librechat-data-provider');
+const { getProjectByName } = require('./Project');
+const { getAgent } = require('./Agent');
+const { File } = require('~/db/models');
 
 /**
  * Finds a file by its file_id with additional query options.
@@ -14,17 +14,124 @@ const findFileById = async (file_id, options = {}) => {
   return await File.findOne({ file_id, ...options }).lean();
 };
 
+/**
+ * Checks if a user has access to multiple files through a shared agent (batch operation)
+ * @param {string} userId - The user ID to check access for
+ * @param {string[]} fileIds - Array of file IDs to check
+ * @param {string} agentId - The agent ID that might grant access
+ * @returns {Promise<Map<string, boolean>>} Map of fileId to access status
+ */
+const hasAccessToFilesViaAgent = async (userId, fileIds, agentId, checkCollaborative = true) => {
+  const accessMap = new Map();
+
+  // Initialize all files as no access
+  fileIds.forEach((fileId) => accessMap.set(fileId, false));
+
+  try {
+    const agent = await getAgent({ id: agentId });
+
+    if (!agent) {
+      return accessMap;
+    }
+
+    // Check if user is the author - if so, grant access to all files
+    if (agent.author.toString() === userId) {
+      fileIds.forEach((fileId) => accessMap.set(fileId, true));
+      return accessMap;
+    }
+
+    // Check if agent is shared with the user via projects
+    if (!agent.projectIds || agent.projectIds.length === 0) {
+      return accessMap;
+    }
+
+    // Check if agent is in global project
+    const globalProject = await getProjectByName(Constants.GLOBAL_PROJECT_NAME, '_id');
+    if (
+      !globalProject ||
+      !agent.projectIds.some((pid) => pid.toString() === globalProject._id.toString())
+    ) {
+      return accessMap;
+    }
+
+    // Agent is globally shared - check if it's collaborative
+    if (checkCollaborative && !agent.isCollaborative) {
+      return accessMap;
+    }
+
+    // Check which files are actually attached
+    const attachedFileIds = new Set();
+    if (agent.tool_resources) {
+      for (const [_resourceType, resource] of Object.entries(agent.tool_resources)) {
+        if (resource?.file_ids && Array.isArray(resource.file_ids)) {
+          resource.file_ids.forEach((fileId) => attachedFileIds.add(fileId));
+        }
+      }
+    }
+
+    // Grant access only to files that are attached to this agent
+    fileIds.forEach((fileId) => {
+      if (attachedFileIds.has(fileId)) {
+        accessMap.set(fileId, true);
+      }
+    });
+
+    return accessMap;
+  } catch (error) {
+    logger.error('[hasAccessToFilesViaAgent] Error checking file access:', error);
+    return accessMap;
+  }
+};
+
 /**
  * Retrieves files matching a given filter, sorted by the most recently updated.
  * @param {Object} filter - The filter criteria to apply.
  * @param {Object} [_sortOptions] - Optional sort parameters.
  * @param {Object|String} [selectFields={ text: 0 }] - Fields to include/exclude in the query results.
  *                                                   Default excludes the 'text' field.
+ * @param {Object} [options] - Additional options
+ * @param {string} [options.userId] - User ID for access control
+ * @param {string} [options.agentId] - Agent ID that might grant access to files
  * @returns {Promise<Array<MongoFile>>} A promise that resolves to an array of file documents.
  */
-const getFiles = async (filter, _sortOptions, selectFields = { text: 0 }) => {
+const getFiles = async (filter, _sortOptions, selectFields = { text: 0 }, options = {}) => {
   const sortOptions = { updatedAt: -1, ..._sortOptions };
-  return await File.find(filter).select(selectFields).sort(sortOptions).lean();
+  const files = await File.find(filter).select(selectFields).sort(sortOptions).lean();
+
+  // If userId and agentId are provided, filter files based on access
+  if (options.userId && options.agentId) {
+    // Collect file IDs that need access check
+    const filesToCheck = [];
+    const ownedFiles = [];
+
+    for (const file of files) {
+      if (file.user && file.user.toString() === options.userId) {
+        ownedFiles.push(file);
+      } else {
+        filesToCheck.push(file);
+      }
+    }
+
+    if (filesToCheck.length === 0) {
+      return ownedFiles;
+    }
+
+    // Batch check access for all non-owned files
+    const fileIds = filesToCheck.map((f) => f.file_id);
+    const accessMap = await hasAccessToFilesViaAgent(
+      options.userId,
+      fileIds,
+      options.agentId,
+      false,
+    );
+
+    // Filter files based on access
+    const accessibleFiles = filesToCheck.filter((file) => accessMap.get(file.file_id));
+
+    return [...ownedFiles, ...accessibleFiles];
+  }
+
+  return files;
 };
 
 /**
@@ -34,19 +141,19 @@ const getFiles = async (filter, _sortOptions, selectFields = { text: 0 }) => {
  * @returns {Promise<Array<MongoFile>>} Files that match the criteria
  */
 const getToolFilesByIds = async (fileIds, toolResourceSet) => {
-  if (!fileIds || !fileIds.length) {
+  if (!fileIds || !fileIds.length || !toolResourceSet?.size) {
     return [];
   }
 
   try {
     const filter = {
       file_id: { $in: fileIds },
+      $or: [],
     };
 
-    if (toolResourceSet.size) {
-      filter.$or = [];
+    if (toolResourceSet.has(EToolResources.ocr)) {
+      filter.$or.push({ text: { $exists: true, $ne: null }, context: FileContext.agents });
     }
-
     if (toolResourceSet.has(EToolResources.file_search)) {
       filter.$or.push({ embedded: true });
     }
@@ -178,4 +285,5 @@ module.exports = {
   deleteFiles,
   deleteFileByFilter,
   batchUpdateFiles,
+  hasAccessToFilesViaAgent,
 };

api/models/File.spec.js

@@ -0,0 +1,264 @@
+const mongoose = require('mongoose');
+const { v4: uuidv4 } = require('uuid');
+const { fileSchema } = require('@librechat/data-schemas');
+const { agentSchema } = require('@librechat/data-schemas');
+const { projectSchema } = require('@librechat/data-schemas');
+const { MongoMemoryServer } = require('mongodb-memory-server');
+const { GLOBAL_PROJECT_NAME } = require('librechat-data-provider').Constants;
+const { getFiles, createFile } = require('./File');
+const { getProjectByName } = require('./Project');
+const { createAgent } = require('./Agent');
+
+let File;
+let Agent;
+let Project;
+
+describe('File Access Control', () => {
+  let mongoServer;
+
+  beforeAll(async () => {
+    mongoServer = await MongoMemoryServer.create();
+    const mongoUri = mongoServer.getUri();
+    File = mongoose.models.File || mongoose.model('File', fileSchema);
+    Agent = mongoose.models.Agent || mongoose.model('Agent', agentSchema);
+    Project = mongoose.models.Project || mongoose.model('Project', projectSchema);
+    await mongoose.connect(mongoUri);
+  });
+
+  afterAll(async () => {
+    await mongoose.disconnect();
+    await mongoServer.stop();
+  });
+
+  beforeEach(async () => {
+    await File.deleteMany({});
+    await Agent.deleteMany({});
+    await Project.deleteMany({});
+  });
+
+  describe('hasAccessToFilesViaAgent', () => {
+    it('should efficiently check access for multiple files at once', async () => {
+      const userId = new mongoose.Types.ObjectId().toString();
+      const authorId = new mongoose.Types.ObjectId().toString();
+      const agentId = uuidv4();
+      const fileIds = [uuidv4(), uuidv4(), uuidv4(), uuidv4()];
+
+      // Create files
+      for (const fileId of fileIds) {
+        await createFile({
+          user: authorId,
+          file_id: fileId,
+          filename: `file-${fileId}.txt`,
+          filepath: `/uploads/${fileId}`,
+        });
+      }
+
+      // Create agent with only first two files attached
+      await createAgent({
+        id: agentId,
+        name: 'Test Agent',
+        author: authorId,
+        model: 'gpt-4',
+        provider: 'openai',
+        isCollaborative: true,
+        tool_resources: {
+          file_search: {
+            file_ids: [fileIds[0], fileIds[1]],
+          },
+        },
+      });
+
+      // Get or create global project
+      const globalProject = await getProjectByName(GLOBAL_PROJECT_NAME, '_id');
+
+      // Share agent globally
+      await Agent.updateOne({ id: agentId }, { $push: { projectIds: globalProject._id } });
+
+      // Check access for all files
+      const { hasAccessToFilesViaAgent } = require('./File');
+      const accessMap = await hasAccessToFilesViaAgent(userId, fileIds, agentId);
+
+      // Should have access only to the first two files
+      expect(accessMap.get(fileIds[0])).toBe(true);
+      expect(accessMap.get(fileIds[1])).toBe(true);
+      expect(accessMap.get(fileIds[2])).toBe(false);
+      expect(accessMap.get(fileIds[3])).toBe(false);
+    });
+
+    it('should grant access to all files when user is the agent author', async () => {
+      const authorId = new mongoose.Types.ObjectId().toString();
+      const agentId = uuidv4();
+      const fileIds = [uuidv4(), uuidv4(), uuidv4()];
+
+      // Create agent
+      await createAgent({
+        id: agentId,
+        name: 'Test Agent',
+        author: authorId,
+        model: 'gpt-4',
+        provider: 'openai',
+        tool_resources: {
+          file_search: {
+            file_ids: [fileIds[0]], // Only one file attached
+          },
+        },
+      });
+
+      // Check access as the author
+      const { hasAccessToFilesViaAgent } = require('./File');
+      const accessMap = await hasAccessToFilesViaAgent(authorId, fileIds, agentId);
+
+      // Author should have access to all files
+      expect(accessMap.get(fileIds[0])).toBe(true);
+      expect(accessMap.get(fileIds[1])).toBe(true);
+      expect(accessMap.get(fileIds[2])).toBe(true);
+    });
+
+    it('should handle non-existent agent gracefully', async () => {
+      const userId = new mongoose.Types.ObjectId().toString();
+      const fileIds = [uuidv4(), uuidv4()];
+
+      const { hasAccessToFilesViaAgent } = require('./File');
+      const accessMap = await hasAccessToFilesViaAgent(userId, fileIds, 'non-existent-agent');
+
+      // Should have no access to any files
+      expect(accessMap.get(fileIds[0])).toBe(false);
+      expect(accessMap.get(fileIds[1])).toBe(false);
+    });
+
+    it('should deny access when agent is not collaborative', async () => {
+      const userId = new mongoose.Types.ObjectId().toString();
+      const authorId = new mongoose.Types.ObjectId().toString();
+      const agentId = uuidv4();
+      const fileIds = [uuidv4(), uuidv4()];
+
+      // Create agent with files but isCollaborative: false
+      await createAgent({
+        id: agentId,
+        name: 'Non-Collaborative Agent',
+        author: authorId,
+        model: 'gpt-4',
+        provider: 'openai',
+        isCollaborative: false,
+        tool_resources: {
+          file_search: {
+            file_ids: fileIds,
+          },
+        },
+      });
+
+      // Get or create global project
+      const globalProject = await getProjectByName(GLOBAL_PROJECT_NAME, '_id');
+
+      // Share agent globally
+      await Agent.updateOne({ id: agentId }, { $push: { projectIds: globalProject._id } });
+
+      // Check access for files
+      const { hasAccessToFilesViaAgent } = require('./File');
+      const accessMap = await hasAccessToFilesViaAgent(userId, fileIds, agentId);
+
+      // Should have no access to any files when isCollaborative is false
+      expect(accessMap.get(fileIds[0])).toBe(false);
+      expect(accessMap.get(fileIds[1])).toBe(false);
+    });
+  });
+
+  describe('getFiles with agent access control', () => {
+    test('should return files owned by user and files accessible through agent', async () => {
+      const authorId = new mongoose.Types.ObjectId();
+      const userId = new mongoose.Types.ObjectId();
+      const agentId = `agent_${uuidv4()}`;
+      const ownedFileId = `file_${uuidv4()}`;
+      const sharedFileId = `file_${uuidv4()}`;
+      const inaccessibleFileId = `file_${uuidv4()}`;
+
+      // Create/get global project using getProjectByName which will upsert
+      const globalProject = await getProjectByName(GLOBAL_PROJECT_NAME);
+
+      // Create agent with shared file
+      await createAgent({
+        id: agentId,
+        name: 'Shared Agent',
+        provider: 'test',
+        model: 'test-model',
+        author: authorId,
+        projectIds: [globalProject._id],
+        isCollaborative: true,
+        tool_resources: {
+          file_search: {
+            file_ids: [sharedFileId],
+          },
+        },
+      });
+
+      // Create files
+      await createFile({
+        file_id: ownedFileId,
+        user: userId,
+        filename: 'owned.txt',
+        filepath: '/uploads/owned.txt',
+        type: 'text/plain',
+        bytes: 100,
+      });
+
+      await createFile({
+        file_id: sharedFileId,
+        user: authorId,
+        filename: 'shared.txt',
+        filepath: '/uploads/shared.txt',
+        type: 'text/plain',
+        bytes: 200,
+        embedded: true,
+      });
+
+      await createFile({
+        file_id: inaccessibleFileId,
+        user: authorId,
+        filename: 'inaccessible.txt',
+        filepath: '/uploads/inaccessible.txt',
+        type: 'text/plain',
+        bytes: 300,
+      });
+
+      // Get files with access control
+      const files = await getFiles(
+        { file_id: { $in: [ownedFileId, sharedFileId, inaccessibleFileId] } },
+        null,
+        { text: 0 },
+        { userId: userId.toString(), agentId },
+      );
+
+      expect(files).toHaveLength(2);
+      expect(files.map((f) => f.file_id)).toContain(ownedFileId);
+      expect(files.map((f) => f.file_id)).toContain(sharedFileId);
+      expect(files.map((f) => f.file_id)).not.toContain(inaccessibleFileId);
+    });
+
+    test('should return all files when no userId/agentId provided', async () => {
+      const userId = new mongoose.Types.ObjectId();
+      const fileId1 = `file_${uuidv4()}`;
+      const fileId2 = `file_${uuidv4()}`;
+
+      await createFile({
+        file_id: fileId1,
+        user: userId,
+        filename: 'file1.txt',
+        filepath: '/uploads/file1.txt',
+        type: 'text/plain',
+        bytes: 100,
+      });
+
+      await createFile({
+        file_id: fileId2,
+        user: new mongoose.Types.ObjectId(),
+        filename: 'file2.txt',
+        filepath: '/uploads/file2.txt',
+        type: 'text/plain',
+        bytes: 200,
+      });
+
+      const files = await getFiles({ file_id: { $in: [fileId1, fileId2] } });
+      expect(files).toHaveLength(2);
+    });
+  });
+});

api/models/Message.js

@@ -1,7 +1,9 @@
 const { z } = require('zod');
 const { logger } = require('@librechat/data-schemas');
+const { createTempChatExpirationDate } = require('@librechat/api');
+const { getCustomConfig } = require('~/server/services/Config/getCustomConfig');
+const { Message } = require('~/db/models');
 
-const Message = require('~/db/models').Message;
 const idSchema = z.string().uuid();
 
 /**
@@ -54,9 +56,14 @@ async function saveMessage(req, params, metadata) {
     };
 
     if (req?.body?.isTemporary) {
-      const expiredAt = new Date();
-      expiredAt.setDate(expiredAt.getDate() + 30);
-      update.expiredAt = expiredAt;
+      try {
+        const customConfig = await getCustomConfig();
+        update.expiredAt = createTempChatExpirationDate(customConfig);
+      } catch (err) {
+        logger.error('Error creating temporary chat expiration date:', err);
+        logger.info(`---\`saveMessage\` context: ${metadata?.context}`);
+        update.expiredAt = null;
+      }
     } else {
       update.expiredAt = null;
     }
@@ -253,6 +260,7 @@ async function updateMessage(req, message, metadata) {
       text: updatedMessage.text,
       isCreatedByUser: updatedMessage.isCreatedByUser,
       tokenCount: updatedMessage.tokenCount,
+      feedback: updatedMessage.feedback,
     };
   } catch (err) {
     logger.error('Error updating message:', err);

api/models/Message.spec.js

@@ -1,116 +1,115 @@
+const mongoose = require('mongoose');
 const { v4: uuidv4 } = require('uuid');
-
-jest.mock('mongoose');
-
-const mockFindQuery = {
-  select: jest.fn().mockReturnThis(),
-  sort: jest.fn().mockReturnThis(),
-  lean: jest.fn().mockReturnThis(),
-  deleteMany: jest.fn().mockResolvedValue({ deletedCount: 1 }),
-};
-
-const mockSchema = {
-  findOneAndUpdate: jest.fn(),
-  updateOne: jest.fn(),
-  findOne: jest.fn(() => ({
-    lean: jest.fn(),
-  })),
-  find: jest.fn(() => mockFindQuery),
-  deleteMany: jest.fn(),
-};
-
-jest.mock('~/config/winston', () => ({
-  error: jest.fn(),
-}));
-
-const mockModels = {
-  Message: {
-    findOneAndUpdate: mockSchema.findOneAndUpdate,
-    updateOne: mockSchema.updateOne,
-    findOne: mockSchema.findOne,
-    find: mockSchema.find,
-    deleteMany: mockSchema.deleteMany,
-  },
-};
+const { messageSchema } = require('@librechat/data-schemas');
+const { MongoMemoryServer } = require('mongodb-memory-server');
 
 const {
   saveMessage,
   getMessages,
   updateMessage,
   deleteMessages,
+  bulkSaveMessages,
   updateMessageText,
   deleteMessagesSince,
-} = require('~/models/Message');
+} = require('./Message');
+
+jest.mock('~/server/services/Config/getCustomConfig');
+const { getCustomConfig } = require('~/server/services/Config/getCustomConfig');
+
+/**
+ * @type {import('mongoose').Model<import('@librechat/data-schemas').IMessage>}
+ */
+let Message;
 
 describe('Message Operations', () => {
+  let mongoServer;
   let mockReq;
-  let mockMessage;
+  let mockMessageData;
 
-  beforeEach(() => {
-    jest.clearAllMocks();
+  beforeAll(async () => {
+    mongoServer = await MongoMemoryServer.create();
+    const mongoUri = mongoServer.getUri();
+    Message = mongoose.models.Message || mongoose.model('Message', messageSchema);
+    await mongoose.connect(mongoUri);
+  });
+
+  afterAll(async () => {
+    await mongoose.disconnect();
+    await mongoServer.stop();
+  });
+
+  beforeEach(async () => {
+    // Clear database
+    await Message.deleteMany({});
 
     mockReq = {
       user: { id: 'user123' },
     };
 
-    mockMessage = {
+    mockMessageData = {
       messageId: 'msg123',
       conversationId: uuidv4(),
       text: 'Hello, world!',
       user: 'user123',
     };
-
-    mockSchema.findOneAndUpdate.mockResolvedValue({
-      toObject: () => mockMessage,
-    });
   });
 
   describe('saveMessage', () => {
     it('should save a message for an authenticated user', async () => {
-      const result = await saveMessage(mockReq, mockMessage);
-      expect(result).toEqual(mockMessage);
-      expect(mockSchema.findOneAndUpdate).toHaveBeenCalledWith(
-        { messageId: 'msg123', user: 'user123' },
-        expect.objectContaining({ user: 'user123' }),
-        expect.any(Object),
-      );
+      const result = await saveMessage(mockReq, mockMessageData);
+
+      expect(result.messageId).toBe('msg123');
+      expect(result.user).toBe('user123');
+      expect(result.text).toBe('Hello, world!');
+
+      // Verify the message was actually saved to the database
+      const savedMessage = await Message.findOne({ messageId: 'msg123', user: 'user123' });
+      expect(savedMessage).toBeTruthy();
+      expect(savedMessage.text).toBe('Hello, world!');
     });
 
     it('should throw an error for unauthenticated user', async () => {
       mockReq.user = null;
-      await expect(saveMessage(mockReq, mockMessage)).rejects.toThrow('User not authenticated');
+      await expect(saveMessage(mockReq, mockMessageData)).rejects.toThrow('User not authenticated');
     });
 
-    it('should throw an error for invalid conversation ID', async () => {
-      mockMessage.conversationId = 'invalid-id';
-      await expect(saveMessage(mockReq, mockMessage)).resolves.toBeUndefined();
+    it('should handle invalid conversation ID gracefully', async () => {
+      mockMessageData.conversationId = 'invalid-id';
+      const result = await saveMessage(mockReq, mockMessageData);
+      expect(result).toBeUndefined();
     });
   });
 
   describe('updateMessageText', () => {
     it('should update message text for the authenticated user', async () => {
+      // First save a message
+      await saveMessage(mockReq, mockMessageData);
+
+      // Then update it
       await updateMessageText(mockReq, { messageId: 'msg123', text: 'Updated text' });
-      expect(mockSchema.updateOne).toHaveBeenCalledWith(
-        { messageId: 'msg123', user: 'user123' },
-        { text: 'Updated text' },
-      );
+
+      // Verify the update
+      const updatedMessage = await Message.findOne({ messageId: 'msg123', user: 'user123' });
+      expect(updatedMessage.text).toBe('Updated text');
     });
   });
 
   describe('updateMessage', () => {
     it('should update a message for the authenticated user', async () => {
-      mockSchema.findOneAndUpdate.mockResolvedValue(mockMessage);
+      // First save a message
+      await saveMessage(mockReq, mockMessageData);
+
       const result = await updateMessage(mockReq, { messageId: 'msg123', text: 'Updated text' });
-      expect(result).toEqual(
-        expect.objectContaining({
-          messageId: 'msg123',
-          text: 'Hello, world!',
-        }),
-      );
+
+      expect(result.messageId).toBe('msg123');
+      expect(result.text).toBe('Updated text');
+
+      // Verify in database
+      const updatedMessage = await Message.findOne({ messageId: 'msg123', user: 'user123' });
+      expect(updatedMessage.text).toBe('Updated text');
     });
 
     it('should throw an error if message is not found', async () => {
-      mockSchema.findOneAndUpdate.mockResolvedValue(null);
       await expect(
         updateMessage(mockReq, { messageId: 'nonexistent', text: 'Test' }),
       ).rejects.toThrow('Message not found or user not authorized.');
@@ -119,19 +118,45 @@ describe('Message Operations', () => {
 
   describe('deleteMessagesSince', () => {
     it('should delete messages only for the authenticated user', async () => {
-      mockSchema.findOne().lean.mockResolvedValueOnce({ createdAt: new Date() });
-      mockFindQuery.deleteMany.mockResolvedValueOnce({ deletedCount: 1 });
-      const result = await deleteMessagesSince(mockReq, {
-        messageId: 'msg123',
-        conversationId: 'convo123',
+      const conversationId = uuidv4();
+
+      // Create multiple messages in the same conversation
+      await saveMessage(mockReq, {
+        messageId: 'msg1',
+        conversationId,
+        text: 'First message',
+        user: 'user123',
       });
-      expect(mockSchema.findOne).toHaveBeenCalledWith({ messageId: 'msg123', user: 'user123' });
-      expect(mockSchema.find).not.toHaveBeenCalled();
-      expect(result).toBeUndefined();
+
+      await saveMessage(mockReq, {
+        messageId: 'msg2',
+        conversationId,
+        text: 'Second message',
+        user: 'user123',
+      });
+
+      await saveMessage(mockReq, {
+        messageId: 'msg3',
+        conversationId,
+        text: 'Third message',
+        user: 'user123',
+      });
+
+      // Delete messages since message2 (this should only delete messages created AFTER msg2)
+      await deleteMessagesSince(mockReq, {
+        messageId: 'msg2',
+        conversationId,
+      });
+
+      // Verify msg1 and msg2 remain, msg3 is deleted
+      const remainingMessages = await Message.find({ conversationId, user: 'user123' });
+      expect(remainingMessages).toHaveLength(2);
+      expect(remainingMessages.map((m) => m.messageId)).toContain('msg1');
+      expect(remainingMessages.map((m) => m.messageId)).toContain('msg2');
+      expect(remainingMessages.map((m) => m.messageId)).not.toContain('msg3');
     });
 
     it('should return undefined if no message is found', async () => {
-      mockSchema.findOne().lean.mockResolvedValueOnce(null);
       const result = await deleteMessagesSince(mockReq, {
         messageId: 'nonexistent',
         conversationId: 'convo123',
@@ -142,29 +167,71 @@ describe('Message Operations', () => {
 
   describe('getMessages', () => {
     it('should retrieve messages with the correct filter', async () => {
-      const filter = { conversationId: 'convo123' };
-      await getMessages(filter);
-      expect(mockSchema.find).toHaveBeenCalledWith(filter);
-      expect(mockFindQuery.sort).toHaveBeenCalledWith({ createdAt: 1 });
-      expect(mockFindQuery.lean).toHaveBeenCalled();
+      const conversationId = uuidv4();
+
+      // Save some messages
+      await saveMessage(mockReq, {
+        messageId: 'msg1',
+        conversationId,
+        text: 'First message',
+        user: 'user123',
+      });
+
+      await saveMessage(mockReq, {
+        messageId: 'msg2',
+        conversationId,
+        text: 'Second message',
+        user: 'user123',
+      });
+
+      const messages = await getMessages({ conversationId });
+      expect(messages).toHaveLength(2);
+      expect(messages[0].text).toBe('First message');
+      expect(messages[1].text).toBe('Second message');
     });
   });
 
   describe('deleteMessages', () => {
     it('should delete messages with the correct filter', async () => {
+      // Save some messages for different users
+      await saveMessage(mockReq, mockMessageData);
+      await saveMessage(
+        { user: { id: 'user456' } },
+        {
+          messageId: 'msg456',
+          conversationId: uuidv4(),
+          text: 'Other user message',
+          user: 'user456',
+        },
+      );
+
       await deleteMessages({ user: 'user123' });
-      expect(mockSchema.deleteMany).toHaveBeenCalledWith({ user: 'user123' });
+
+      // Verify only user123's messages were deleted
+      const user123Messages = await Message.find({ user: 'user123' });
+      const user456Messages = await Message.find({ user: 'user456' });
+
+      expect(user123Messages).toHaveLength(0);
+      expect(user456Messages).toHaveLength(1);
     });
   });
 
   describe('Conversation Hijacking Prevention', () => {
     it("should not allow editing a message in another user's conversation", async () => {
       const attackerReq = { user: { id: 'attacker123' } };
-      const victimConversationId = 'victim-convo-123';
+      const victimConversationId = uuidv4();
       const victimMessageId = 'victim-msg-123';
 
-      mockSchema.findOneAndUpdate.mockResolvedValue(null);
+      // First, save a message as the victim (but we'll try to edit as attacker)
+      const victimReq = { user: { id: 'victim123' } };
+      await saveMessage(victimReq, {
+        messageId: victimMessageId,
+        conversationId: victimConversationId,
+        text: 'Victim message',
+        user: 'victim123',
+      });
 
+      // Attacker tries to edit the victim's message
       await expect(
         updateMessage(attackerReq, {
           messageId: victimMessageId,
@@ -173,71 +240,343 @@ describe('Message Operations', () => {
         }),
       ).rejects.toThrow('Message not found or user not authorized.');
 
-      expect(mockSchema.findOneAndUpdate).toHaveBeenCalledWith(
-        { messageId: victimMessageId, user: 'attacker123' },
-        expect.anything(),
-        expect.anything(),
-      );
+      // Verify the original message is unchanged
+      const originalMessage = await Message.findOne({
+        messageId: victimMessageId,
+        user: 'victim123',
+      });
+      expect(originalMessage.text).toBe('Victim message');
     });
 
     it("should not allow deleting messages from another user's conversation", async () => {
       const attackerReq = { user: { id: 'attacker123' } };
-      const victimConversationId = 'victim-convo-123';
+      const victimConversationId = uuidv4();
       const victimMessageId = 'victim-msg-123';
 
-      mockSchema.findOne().lean.mockResolvedValueOnce(null); // Simulating message not found for this user
+      // Save a message as the victim
+      const victimReq = { user: { id: 'victim123' } };
+      await saveMessage(victimReq, {
+        messageId: victimMessageId,
+        conversationId: victimConversationId,
+        text: 'Victim message',
+        user: 'victim123',
+      });
+
+      // Attacker tries to delete from victim's conversation
       const result = await deleteMessagesSince(attackerReq, {
         messageId: victimMessageId,
         conversationId: victimConversationId,
       });
 
       expect(result).toBeUndefined();
-      expect(mockSchema.findOne).toHaveBeenCalledWith({
+
+      // Verify the victim's message still exists
+      const victimMessage = await Message.findOne({
         messageId: victimMessageId,
-        user: 'attacker123',
+        user: 'victim123',
       });
+      expect(victimMessage).toBeTruthy();
+      expect(victimMessage.text).toBe('Victim message');
     });
 
     it("should not allow inserting a new message into another user's conversation", async () => {
       const attackerReq = { user: { id: 'attacker123' } };
-      const victimConversationId = uuidv4(); // Use a valid UUID
+      const victimConversationId = uuidv4();
 
-      await expect(
-        saveMessage(attackerReq, {
-          conversationId: victimConversationId,
-          text: 'Inserted malicious message',
-          messageId: 'new-msg-123',
-        }),
-      ).resolves.not.toThrow(); // It should not throw an error
+      // Attacker tries to save a message - this should succeed but with attacker's user ID
+      const result = await saveMessage(attackerReq, {
+        conversationId: victimConversationId,
+        text: 'Inserted malicious message',
+        messageId: 'new-msg-123',
+        user: 'attacker123',
+      });
 
-      // Check that the message was saved with the attacker's user ID
-      expect(mockSchema.findOneAndUpdate).toHaveBeenCalledWith(
-        { messageId: 'new-msg-123', user: 'attacker123' },
-        expect.objectContaining({
-          user: 'attacker123',
-          conversationId: victimConversationId,
-        }),
-        expect.anything(),
-      );
+      expect(result).toBeTruthy();
+      expect(result.user).toBe('attacker123');
+
+      // Verify the message was saved with the attacker's user ID, not as an anonymous message
+      const savedMessage = await Message.findOne({ messageId: 'new-msg-123' });
+      expect(savedMessage.user).toBe('attacker123');
+      expect(savedMessage.conversationId).toBe(victimConversationId);
     });
 
     it('should allow retrieving messages from any conversation', async () => {
-      const victimConversationId = 'victim-convo-123';
+      const victimConversationId = uuidv4();
 
-      await getMessages({ conversationId: victimConversationId });
-
-      expect(mockSchema.find).toHaveBeenCalledWith({
+      // Save a message in the victim's conversation
+      const victimReq = { user: { id: 'victim123' } };
+      await saveMessage(victimReq, {
+        messageId: 'victim-msg',
         conversationId: victimConversationId,
+        text: 'Victim message',
+        user: 'victim123',
       });
 
-      mockSchema.find.mockReturnValueOnce({
-        select: jest.fn().mockReturnThis(),
-        sort: jest.fn().mockReturnThis(),
-        lean: jest.fn().mockResolvedValue([{ text: 'Test message' }]),
+      // Anyone should be able to retrieve messages by conversation ID
+      const messages = await getMessages({ conversationId: victimConversationId });
+      expect(messages).toHaveLength(1);
+      expect(messages[0].text).toBe('Victim message');
+    });
+  });
+
+  describe('isTemporary message handling', () => {
+    beforeEach(() => {
+      // Reset mocks before each test
+      jest.clearAllMocks();
+    });
+
+    it('should save a message with expiredAt when isTemporary is true', async () => {
+      // Mock custom config with 24 hour retention
+      getCustomConfig.mockResolvedValue({
+        interface: {
+          temporaryChatRetention: 24,
+        },
       });
 
-      const result = await getMessages({ conversationId: victimConversationId });
-      expect(result).toEqual([{ text: 'Test message' }]);
+      mockReq.body = { isTemporary: true };
+
+      const beforeSave = new Date();
+      const result = await saveMessage(mockReq, mockMessageData);
+      const afterSave = new Date();
+
+      expect(result.messageId).toBe('msg123');
+      expect(result.expiredAt).toBeDefined();
+      expect(result.expiredAt).toBeInstanceOf(Date);
+
+      // Verify expiredAt is approximately 24 hours in the future
+      const expectedExpirationTime = new Date(beforeSave.getTime() + 24 * 60 * 60 * 1000);
+      const actualExpirationTime = new Date(result.expiredAt);
+
+      expect(actualExpirationTime.getTime()).toBeGreaterThanOrEqual(
+        expectedExpirationTime.getTime() - 1000,
+      );
+      expect(actualExpirationTime.getTime()).toBeLessThanOrEqual(
+        new Date(afterSave.getTime() + 24 * 60 * 60 * 1000 + 1000).getTime(),
+      );
+    });
+
+    it('should save a message without expiredAt when isTemporary is false', async () => {
+      mockReq.body = { isTemporary: false };
+
+      const result = await saveMessage(mockReq, mockMessageData);
+
+      expect(result.messageId).toBe('msg123');
+      expect(result.expiredAt).toBeNull();
+    });
+
+    it('should save a message without expiredAt when isTemporary is not provided', async () => {
+      // No isTemporary in body
+      mockReq.body = {};
+
+      const result = await saveMessage(mockReq, mockMessageData);
+
+      expect(result.messageId).toBe('msg123');
+      expect(result.expiredAt).toBeNull();
+    });
+
+    it('should use custom retention period from config', async () => {
+      // Mock custom config with 48 hour retention
+      getCustomConfig.mockResolvedValue({
+        interface: {
+          temporaryChatRetention: 48,
+        },
+      });
+
+      mockReq.body = { isTemporary: true };
+
+      const beforeSave = new Date();
+      const result = await saveMessage(mockReq, mockMessageData);
+
+      expect(result.expiredAt).toBeDefined();
+
+      // Verify expiredAt is approximately 48 hours in the future
+      const expectedExpirationTime = new Date(beforeSave.getTime() + 48 * 60 * 60 * 1000);
+      const actualExpirationTime = new Date(result.expiredAt);
+
+      expect(actualExpirationTime.getTime()).toBeGreaterThanOrEqual(
+        expectedExpirationTime.getTime() - 1000,
+      );
+      expect(actualExpirationTime.getTime()).toBeLessThanOrEqual(
+        expectedExpirationTime.getTime() + 1000,
+      );
+    });
+
+    it('should handle minimum retention period (1 hour)', async () => {
+      // Mock custom config with less than minimum retention
+      getCustomConfig.mockResolvedValue({
+        interface: {
+          temporaryChatRetention: 0.5, // Half hour - should be clamped to 1 hour
+        },
+      });
+
+      mockReq.body = { isTemporary: true };
+
+      const beforeSave = new Date();
+      const result = await saveMessage(mockReq, mockMessageData);
+
+      expect(result.expiredAt).toBeDefined();
+
+      // Verify expiredAt is approximately 1 hour in the future (minimum)
+      const expectedExpirationTime = new Date(beforeSave.getTime() + 1 * 60 * 60 * 1000);
+      const actualExpirationTime = new Date(result.expiredAt);
+
+      expect(actualExpirationTime.getTime()).toBeGreaterThanOrEqual(
+        expectedExpirationTime.getTime() - 1000,
+      );
+      expect(actualExpirationTime.getTime()).toBeLessThanOrEqual(
+        expectedExpirationTime.getTime() + 1000,
+      );
+    });
+
+    it('should handle maximum retention period (8760 hours)', async () => {
+      // Mock custom config with more than maximum retention
+      getCustomConfig.mockResolvedValue({
+        interface: {
+          temporaryChatRetention: 10000, // Should be clamped to 8760 hours
+        },
+      });
+
+      mockReq.body = { isTemporary: true };
+
+      const beforeSave = new Date();
+      const result = await saveMessage(mockReq, mockMessageData);
+
+      expect(result.expiredAt).toBeDefined();
+
+      // Verify expiredAt is approximately 8760 hours (1 year) in the future
+      const expectedExpirationTime = new Date(beforeSave.getTime() + 8760 * 60 * 60 * 1000);
+      const actualExpirationTime = new Date(result.expiredAt);
+
+      expect(actualExpirationTime.getTime()).toBeGreaterThanOrEqual(
+        expectedExpirationTime.getTime() - 1000,
+      );
+      expect(actualExpirationTime.getTime()).toBeLessThanOrEqual(
+        expectedExpirationTime.getTime() + 1000,
+      );
+    });
+
+    it('should handle getCustomConfig errors gracefully', async () => {
+      // Mock getCustomConfig to throw an error
+      getCustomConfig.mockRejectedValue(new Error('Config service unavailable'));
+
+      mockReq.body = { isTemporary: true };
+
+      const result = await saveMessage(mockReq, mockMessageData);
+
+      // Should still save the message but with expiredAt as null
+      expect(result.messageId).toBe('msg123');
+      expect(result.expiredAt).toBeNull();
+    });
+
+    it('should use default retention when config is not provided', async () => {
+      // Mock getCustomConfig to return empty config
+      getCustomConfig.mockResolvedValue({});
+
+      mockReq.body = { isTemporary: true };
+
+      const beforeSave = new Date();
+      const result = await saveMessage(mockReq, mockMessageData);
+
+      expect(result.expiredAt).toBeDefined();
+
+      // Default retention is 30 days (720 hours)
+      const expectedExpirationTime = new Date(beforeSave.getTime() + 30 * 24 * 60 * 60 * 1000);
+      const actualExpirationTime = new Date(result.expiredAt);
+
+      expect(actualExpirationTime.getTime()).toBeGreaterThanOrEqual(
+        expectedExpirationTime.getTime() - 1000,
+      );
+      expect(actualExpirationTime.getTime()).toBeLessThanOrEqual(
+        expectedExpirationTime.getTime() + 1000,
+      );
+    });
+
+    it('should not update expiredAt on message update', async () => {
+      // First save a temporary message
+      getCustomConfig.mockResolvedValue({
+        interface: {
+          temporaryChatRetention: 24,
+        },
+      });
+
+      mockReq.body = { isTemporary: true };
+      const savedMessage = await saveMessage(mockReq, mockMessageData);
+      const originalExpiredAt = savedMessage.expiredAt;
+
+      // Now update the message without isTemporary flag
+      mockReq.body = {};
+      const updatedMessage = await updateMessage(mockReq, {
+        messageId: 'msg123',
+        text: 'Updated text',
+      });
+
+      // expiredAt should not be in the returned updated message object
+      expect(updatedMessage.expiredAt).toBeUndefined();
+
+      // Verify in database that expiredAt wasn't changed
+      const dbMessage = await Message.findOne({ messageId: 'msg123', user: 'user123' });
+      expect(dbMessage.expiredAt).toEqual(originalExpiredAt);
+    });
+
+    it('should preserve expiredAt when saving existing temporary message', async () => {
+      // First save a temporary message
+      getCustomConfig.mockResolvedValue({
+        interface: {
+          temporaryChatRetention: 24,
+        },
+      });
+
+      mockReq.body = { isTemporary: true };
+      const firstSave = await saveMessage(mockReq, mockMessageData);
+      const originalExpiredAt = firstSave.expiredAt;
+
+      // Wait a bit to ensure time difference
+      await new Promise((resolve) => setTimeout(resolve, 100));
+
+      // Save again with same messageId but different text
+      const updatedData = { ...mockMessageData, text: 'Updated text' };
+      const secondSave = await saveMessage(mockReq, updatedData);
+
+      // Should update text but create new expiredAt
+      expect(secondSave.text).toBe('Updated text');
+      expect(secondSave.expiredAt).toBeDefined();
+      expect(new Date(secondSave.expiredAt).getTime()).toBeGreaterThan(
+        new Date(originalExpiredAt).getTime(),
+      );
+    });
+
+    it('should handle bulk operations with temporary messages', async () => {
+      // This test verifies bulkSaveMessages doesn't interfere with expiredAt
+      const messages = [
+        {
+          messageId: 'bulk1',
+          conversationId: uuidv4(),
+          text: 'Bulk message 1',
+          user: 'user123',
+          expiredAt: new Date(Date.now() + 24 * 60 * 60 * 1000),
+        },
+        {
+          messageId: 'bulk2',
+          conversationId: uuidv4(),
+          text: 'Bulk message 2',
+          user: 'user123',
+          expiredAt: null,
+        },
+      ];
+
+      await bulkSaveMessages(messages);
+
+      const savedMessages = await Message.find({
+        messageId: { $in: ['bulk1', 'bulk2'] },
+      }).lean();
+
+      expect(savedMessages).toHaveLength(2);
+
+      const bulk1 = savedMessages.find((m) => m.messageId === 'bulk1');
+      const bulk2 = savedMessages.find((m) => m.messageId === 'bulk2');
+
+      expect(bulk1.expiredAt).toBeDefined();
+      expect(bulk2.expiredAt).toBeNull();
     });
   });
 });

api/models/Preset.js

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

api/models/Project.js

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

api/models/Prompt.js

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

api/models/Role.js

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

api/models/Role.spec.js

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

api/models/Share.js

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

api/models/Token.js

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

api/models/ToolCall.js

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

api/models/Transaction.js

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

api/models/Transaction.spec.js

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

api/models/balanceMethods.js

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

api/models/convoStructure.spec.js

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

api/models/inviteUser.js

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

api/models/spendTokens.spec.js

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

api/models/tx.js

@@ -1,4 +1,4 @@
-const { matchModelName } = require('../utils');
+const { matchModelName } = require('../utils/tokens');
 const defaultRate = 6;
 
 /**
@@ -78,7 +78,7 @@ const tokenValues = Object.assign(
     'gpt-3.5-turbo-1106': { prompt: 1, completion: 2 },
     'o4-mini': { prompt: 1.1, completion: 4.4 },
     'o3-mini': { prompt: 1.1, completion: 4.4 },
-    o3: { prompt: 10, completion: 40 },
+    o3: { prompt: 2, completion: 8 },
     'o1-mini': { prompt: 1.1, completion: 4.4 },
     'o1-preview': { prompt: 15, completion: 60 },
     o1: { prompt: 15, completion: 60 },
@@ -87,6 +87,9 @@ const tokenValues = Object.assign(
     'gpt-4.1': { prompt: 2, completion: 8 },
     'gpt-4.5': { prompt: 75, completion: 150 },
     'gpt-4o-mini': { prompt: 0.15, completion: 0.6 },
+    'gpt-5': { prompt: 1.25, completion: 10 },
+    'gpt-5-mini': { prompt: 0.25, completion: 2 },
+    'gpt-5-nano': { prompt: 0.05, completion: 0.4 },
     'gpt-4o': { prompt: 2.5, completion: 10 },
     'gpt-4o-2024-05-13': { prompt: 5, completion: 15 },
     'gpt-4-1106': { prompt: 10, completion: 30 },
@@ -135,10 +138,11 @@ const tokenValues = Object.assign(
     'grok-2-1212': { prompt: 2.0, completion: 10.0 },
     'grok-2-latest': { prompt: 2.0, completion: 10.0 },
     'grok-2': { prompt: 2.0, completion: 10.0 },
-    'grok-3-mini-fast': { prompt: 0.4, completion: 4 },
+    'grok-3-mini-fast': { prompt: 0.6, completion: 4 },
     'grok-3-mini': { prompt: 0.3, completion: 0.5 },
     'grok-3-fast': { prompt: 5.0, completion: 25.0 },
     'grok-3': { prompt: 3.0, completion: 15.0 },
+    'grok-4': { prompt: 3.0, completion: 15.0 },
     'grok-beta': { prompt: 5.0, completion: 15.0 },
     'mistral-large': { prompt: 2.0, completion: 6.0 },
     'pixtral-large': { prompt: 2.0, completion: 6.0 },
@@ -146,6 +150,9 @@ const tokenValues = Object.assign(
     codestral: { prompt: 0.3, completion: 0.9 },
     'ministral-8b': { prompt: 0.1, completion: 0.1 },
     'ministral-3b': { prompt: 0.04, completion: 0.04 },
+    // GPT-OSS models
+    'gpt-oss-20b': { prompt: 0.05, completion: 0.2 },
+    'gpt-oss-120b': { prompt: 0.15, completion: 0.6 },
   },
   bedrockValues,
 );
@@ -213,6 +220,12 @@ const getValueKey = (model, endpoint) => {
     return 'gpt-4.1';
   } else if (modelName.includes('gpt-4o-2024-05-13')) {
     return 'gpt-4o-2024-05-13';
+  } else if (modelName.includes('gpt-5-nano')) {
+    return 'gpt-5-nano';
+  } else if (modelName.includes('gpt-5-mini')) {
+    return 'gpt-5-mini';
+  } else if (modelName.includes('gpt-5')) {
+    return 'gpt-5';
   } else if (modelName.includes('gpt-4o-mini')) {
     return 'gpt-4o-mini';
   } else if (modelName.includes('gpt-4o')) {

api/models/tx.spec.js

@@ -25,8 +25,14 @@ describe('getValueKey', () => {
     expect(getValueKey('gpt-4-some-other-info')).toBe('8k');
   });
 
-  it('should return undefined for model names that do not match any known patterns', () => {
-    expect(getValueKey('gpt-5-some-other-info')).toBeUndefined();
+  it('should return "gpt-5" for model name containing "gpt-5"', () => {
+    expect(getValueKey('gpt-5-some-other-info')).toBe('gpt-5');
+    expect(getValueKey('gpt-5-2025-01-30')).toBe('gpt-5');
+    expect(getValueKey('gpt-5-2025-01-30-0130')).toBe('gpt-5');
+    expect(getValueKey('openai/gpt-5')).toBe('gpt-5');
+    expect(getValueKey('openai/gpt-5-2025-01-30')).toBe('gpt-5');
+    expect(getValueKey('gpt-5-turbo')).toBe('gpt-5');
+    expect(getValueKey('gpt-5-0130')).toBe('gpt-5');
   });
 
   it('should return "gpt-3.5-turbo-1106" for model name containing "gpt-3.5-turbo-1106"', () => {
@@ -84,6 +90,29 @@ describe('getValueKey', () => {
     expect(getValueKey('gpt-4.1-nano-0125')).toBe('gpt-4.1-nano');
   });
 
+  it('should return "gpt-5" for model type of "gpt-5"', () => {
+    expect(getValueKey('gpt-5-2025-01-30')).toBe('gpt-5');
+    expect(getValueKey('gpt-5-2025-01-30-0130')).toBe('gpt-5');
+    expect(getValueKey('openai/gpt-5')).toBe('gpt-5');
+    expect(getValueKey('openai/gpt-5-2025-01-30')).toBe('gpt-5');
+    expect(getValueKey('gpt-5-turbo')).toBe('gpt-5');
+    expect(getValueKey('gpt-5-0130')).toBe('gpt-5');
+  });
+
+  it('should return "gpt-5-mini" for model type of "gpt-5-mini"', () => {
+    expect(getValueKey('gpt-5-mini-2025-01-30')).toBe('gpt-5-mini');
+    expect(getValueKey('openai/gpt-5-mini')).toBe('gpt-5-mini');
+    expect(getValueKey('gpt-5-mini-0130')).toBe('gpt-5-mini');
+    expect(getValueKey('gpt-5-mini-2025-01-30-0130')).toBe('gpt-5-mini');
+  });
+
+  it('should return "gpt-5-nano" for model type of "gpt-5-nano"', () => {
+    expect(getValueKey('gpt-5-nano-2025-01-30')).toBe('gpt-5-nano');
+    expect(getValueKey('openai/gpt-5-nano')).toBe('gpt-5-nano');
+    expect(getValueKey('gpt-5-nano-0130')).toBe('gpt-5-nano');
+    expect(getValueKey('gpt-5-nano-2025-01-30-0130')).toBe('gpt-5-nano');
+  });
+
   it('should return "gpt-4o" for model type of "gpt-4o"', () => {
     expect(getValueKey('gpt-4o-2024-08-06')).toBe('gpt-4o');
     expect(getValueKey('gpt-4o-2024-08-06-0718')).toBe('gpt-4o');
@@ -207,6 +236,48 @@ describe('getMultiplier', () => {
     );
   });
 
+  it('should return the correct multiplier for gpt-5', () => {
+    const valueKey = getValueKey('gpt-5-2025-01-30');
+    expect(getMultiplier({ valueKey, tokenType: 'prompt' })).toBe(tokenValues['gpt-5'].prompt);
+    expect(getMultiplier({ valueKey, tokenType: 'completion' })).toBe(
+      tokenValues['gpt-5'].completion,
+    );
+    expect(getMultiplier({ model: 'gpt-5-preview', tokenType: 'prompt' })).toBe(
+      tokenValues['gpt-5'].prompt,
+    );
+    expect(getMultiplier({ model: 'openai/gpt-5', tokenType: 'completion' })).toBe(
+      tokenValues['gpt-5'].completion,
+    );
+  });
+
+  it('should return the correct multiplier for gpt-5-mini', () => {
+    const valueKey = getValueKey('gpt-5-mini-2025-01-30');
+    expect(getMultiplier({ valueKey, tokenType: 'prompt' })).toBe(tokenValues['gpt-5-mini'].prompt);
+    expect(getMultiplier({ valueKey, tokenType: 'completion' })).toBe(
+      tokenValues['gpt-5-mini'].completion,
+    );
+    expect(getMultiplier({ model: 'gpt-5-mini-preview', tokenType: 'prompt' })).toBe(
+      tokenValues['gpt-5-mini'].prompt,
+    );
+    expect(getMultiplier({ model: 'openai/gpt-5-mini', tokenType: 'completion' })).toBe(
+      tokenValues['gpt-5-mini'].completion,
+    );
+  });
+
+  it('should return the correct multiplier for gpt-5-nano', () => {
+    const valueKey = getValueKey('gpt-5-nano-2025-01-30');
+    expect(getMultiplier({ valueKey, tokenType: 'prompt' })).toBe(tokenValues['gpt-5-nano'].prompt);
+    expect(getMultiplier({ valueKey, tokenType: 'completion' })).toBe(
+      tokenValues['gpt-5-nano'].completion,
+    );
+    expect(getMultiplier({ model: 'gpt-5-nano-preview', tokenType: 'prompt' })).toBe(
+      tokenValues['gpt-5-nano'].prompt,
+    );
+    expect(getMultiplier({ model: 'openai/gpt-5-nano', tokenType: 'completion' })).toBe(
+      tokenValues['gpt-5-nano'].completion,
+    );
+  });
+
   it('should return the correct multiplier for gpt-4o', () => {
     const valueKey = getValueKey('gpt-4o-2024-08-06');
     expect(getMultiplier({ valueKey, tokenType: 'prompt' })).toBe(tokenValues['gpt-4o'].prompt);
@@ -307,10 +378,22 @@ describe('getMultiplier', () => {
   });
 
   it('should return defaultRate if derived valueKey does not match any known patterns', () => {
-    expect(getMultiplier({ tokenType: 'prompt', model: 'gpt-5-some-other-info' })).toBe(
+    expect(getMultiplier({ tokenType: 'prompt', model: 'gpt-10-some-other-info' })).toBe(
       defaultRate,
     );
   });
+
+  it('should return correct multipliers for GPT-OSS models', () => {
+    const models = ['gpt-oss-20b', 'gpt-oss-120b'];
+    models.forEach((key) => {
+      const expectedPrompt = tokenValues[key].prompt;
+      const expectedCompletion = tokenValues[key].completion;
+      expect(getMultiplier({ valueKey: key, tokenType: 'prompt' })).toBe(expectedPrompt);
+      expect(getMultiplier({ valueKey: key, tokenType: 'completion' })).toBe(expectedCompletion);
+      expect(getMultiplier({ model: key, tokenType: 'prompt' })).toBe(expectedPrompt);
+      expect(getMultiplier({ model: key, tokenType: 'completion' })).toBe(expectedCompletion);
+    });
+  });
 });
 
 describe('AWS Bedrock Model Tests', () => {
@@ -636,6 +719,15 @@ describe('Grok Model Tests - Pricing', () => {
       );
     });
 
+    test('should return correct prompt and completion rates for Grok 4 model', () => {
+      expect(getMultiplier({ model: 'grok-4-0709', tokenType: 'prompt' })).toBe(
+        tokenValues['grok-4'].prompt,
+      );
+      expect(getMultiplier({ model: 'grok-4-0709', tokenType: 'completion' })).toBe(
+        tokenValues['grok-4'].completion,
+      );
+    });
+
     test('should return correct prompt and completion rates for Grok 3 models with prefixes', () => {
       expect(getMultiplier({ model: 'xai/grok-3', tokenType: 'prompt' })).toBe(
         tokenValues['grok-3'].prompt,
@@ -662,6 +754,15 @@ describe('Grok Model Tests - Pricing', () => {
         tokenValues['grok-3-mini-fast'].completion,
       );
     });
+
+    test('should return correct prompt and completion rates for Grok 4 model with prefixes', () => {
+      expect(getMultiplier({ model: 'xai/grok-4-0709', tokenType: 'prompt' })).toBe(
+        tokenValues['grok-4'].prompt,
+      );
+      expect(getMultiplier({ model: 'xai/grok-4-0709', tokenType: 'completion' })).toBe(
+        tokenValues['grok-4'].completion,
+      );
+    });
   });
 });
 

api/models/userMethods.js

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

api/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@librechat/backend",
-  "version": "v0.7.8",
+  "version": "v0.8.0-rc2",
   "description": "",
   "scripts": {
     "start": "echo 'please run this from the root directory'",
@@ -34,29 +34,31 @@
   },
   "homepage": "https://librechat.ai",
   "dependencies": {
-    "@anthropic-ai/sdk": "^0.37.0",
+    "@anthropic-ai/sdk": "^0.52.0",
     "@aws-sdk/client-s3": "^3.758.0",
     "@aws-sdk/s3-request-presigner": "^3.758.0",
     "@azure/identity": "^4.7.0",
     "@azure/search-documents": "^12.0.0",
     "@azure/storage-blob": "^12.27.0",
-    "@google/generative-ai": "^0.23.0",
+    "@google/generative-ai": "^0.24.0",
     "@googleapis/youtube": "^20.0.0",
     "@keyv/redis": "^4.3.3",
-    "@langchain/community": "^0.3.44",
-    "@langchain/core": "^0.3.57",
-    "@langchain/google-genai": "^0.2.9",
-    "@langchain/google-vertexai": "^0.2.9",
+    "@langchain/community": "^0.3.47",
+    "@langchain/core": "^0.3.62",
+    "@langchain/google-genai": "^0.2.13",
+    "@langchain/google-vertexai": "^0.2.13",
+    "@langchain/openai": "^0.5.18",
     "@langchain/textsplitters": "^0.1.0",
-    "@librechat/agents": "^2.4.37",
+    "@librechat/agents": "^2.4.75",
+    "@librechat/api": "*",
     "@librechat/data-schemas": "*",
-    "@node-saml/passport-saml": "^5.0.0",
+    "@modelcontextprotocol/sdk": "^1.17.1",
+    "@node-saml/passport-saml": "^5.1.0",
     "@waylaidwanderer/fetch-event-source": "^3.0.1",
     "axios": "^1.8.2",
     "bcryptjs": "^2.4.3",
-    "cohere-ai": "^7.9.1",
-    "compression": "^1.7.4",
-    "connect-redis": "^7.1.0",
+    "compression": "^1.8.1",
+    "connect-redis": "^8.1.0",
     "cookie": "^0.7.2",
     "cookie-parser": "^1.4.7",
     "cors": "^2.8.5",
@@ -66,10 +68,11 @@
     "express": "^4.21.2",
     "express-mongo-sanitize": "^2.2.0",
     "express-rate-limit": "^7.4.1",
-    "express-session": "^1.18.1",
+    "express-session": "^1.18.2",
     "express-static-gzip": "^2.2.0",
     "file-type": "^18.7.0",
     "firebase": "^11.0.2",
+    "form-data": "^4.0.4",
     "googleapis": "^126.0.1",
     "handlebars": "^4.7.7",
     "https-proxy-agent": "^7.0.6",
@@ -81,18 +84,18 @@
     "keyv-file": "^5.1.2",
     "klona": "^2.0.6",
     "librechat-data-provider": "*",
-    "librechat-mcp": "*",
     "lodash": "^4.17.21",
     "meilisearch": "^0.38.0",
     "memorystore": "^1.6.7",
     "mime": "^3.0.0",
     "module-alias": "^2.2.3",
     "mongoose": "^8.12.1",
-    "multer": "^2.0.0",
+    "multer": "^2.0.2",
     "nanoid": "^3.3.7",
+    "node-fetch": "^2.7.0",
     "nodemailer": "^6.9.15",
     "ollama": "^0.5.0",
-    "openai": "^4.96.2",
+    "openai": "^5.10.1",
     "openai-chat-tokens": "^0.2.8",
     "openid-client": "^6.5.0",
     "passport": "^0.6.0",
@@ -109,14 +112,15 @@
     "tiktoken": "^1.0.15",
     "traverse": "^0.6.7",
     "ua-parser-js": "^1.0.36",
+    "undici": "^7.10.0",
     "winston": "^3.11.0",
-    "winston-daily-rotate-file": "^4.7.1",
+    "winston-daily-rotate-file": "^5.0.0",
     "youtube-transcript": "^1.2.1",
     "zod": "^3.22.4"
   },
   "devDependencies": {
     "jest": "^29.7.0",
-    "mongodb-memory-server": "^10.1.3",
+    "mongodb-memory-server": "^10.1.4",
     "nodemon": "^3.0.3",
     "supertest": "^7.1.0"
   }

api/server/cleanup.js

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

api/server/controllers/AskController.js

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

api/server/controllers/AuthController.js

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

api/server/controllers/Balance.js

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

api/server/controllers/EditController.js

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

api/server/controllers/ErrorController.js

@@ -1,40 +0,0 @@
-const { logger } = require('~/config');
-
-//handle duplicates
-const handleDuplicateKeyError = (err, res) => {
-  logger.error('Duplicate key error:', err.keyValue);
-  const field = `${JSON.stringify(Object.keys(err.keyValue))}`;
-  const code = 409;
-  res
-    .status(code)
-    .send({ messages: `An document with that ${field} already exists.`, fields: field });
-};
-
-//handle validation errors
-const handleValidationError = (err, res) => {
-  logger.error('Validation error:', err.errors);
-  let errors = Object.values(err.errors).map((el) => el.message);
-  let fields = `${JSON.stringify(Object.values(err.errors).map((el) => el.path))}`;
-  let code = 400;
-  if (errors.length > 1) {
-    errors = errors.join(' ');
-    res.status(code).send({ messages: `${JSON.stringify(errors)}`, fields: fields });
-  } else {
-    res.status(code).send({ messages: `${JSON.stringify(errors)}`, fields: fields });
-  }
-};
-
-// eslint-disable-next-line no-unused-vars
-module.exports = (err, req, res, next) => {
-  try {
-    if (err.name === 'ValidationError') {
-      return (err = handleValidationError(err, res));
-    }
-    if (err.code && err.code == 11000) {
-      return (err = handleDuplicateKeyError(err, res));
-    }
-  } catch (err) {
-    logger.error('ErrorController => error', err);
-    res.status(500).send('An unknown error occurred.');
-  }
-};

api/server/controllers/ModelController.js

@@ -5,6 +5,7 @@ const { logger } = require('~/config');
 
 /**
  * @param {ServerRequest} req
+ * @returns {Promise<TModelsConfig>} The models config.
  */
 const getModelsConfig = async (req) => {
   const cache = getLogStores(CacheKeys.CONFIG_STORE);

api/server/controllers/PluginController.js

@@ -1,53 +1,16 @@
-const { CacheKeys, AuthType } = require('librechat-data-provider');
-const { getToolkitKey } = require('~/server/services/ToolService');
-const { getCustomConfig } = require('~/server/services/Config');
-const { availableTools } = require('~/app/clients/tools');
-const { getMCPManager } = require('~/config');
+const { logger } = require('@librechat/data-schemas');
+const { CacheKeys, Constants } = require('librechat-data-provider');
+const {
+  getToolkitKey,
+  checkPluginAuth,
+  filterUniquePlugins,
+  convertMCPToolsToPlugins,
+} = require('@librechat/api');
+const { getCustomConfig, getCachedTools } = require('~/server/services/Config');
+const { availableTools, toolkits } = require('~/app/clients/tools');
+const { getMCPManager, getFlowStateManager } = require('~/config');
 const { getLogStores } = require('~/cache');
 
-/**
- * Filters out duplicate plugins from the list of plugins.
- *
- * @param {TPlugin[]} plugins The list of plugins to filter.
- * @returns {TPlugin[]} The list of plugins with duplicates removed.
- */
-const filterUniquePlugins = (plugins) => {
-  const seen = new Set();
-  return plugins.filter((plugin) => {
-    const duplicate = seen.has(plugin.pluginKey);
-    seen.add(plugin.pluginKey);
-    return !duplicate;
-  });
-};
-
-/**
- * Determines if a plugin is authenticated by checking if all required authentication fields have non-empty values.
- * Supports alternate authentication fields, allowing validation against multiple possible environment variables.
- *
- * @param {TPlugin} plugin The plugin object containing the authentication configuration.
- * @returns {boolean} True if the plugin is authenticated for all required fields, false otherwise.
- */
-const checkPluginAuth = (plugin) => {
-  if (!plugin.authConfig || plugin.authConfig.length === 0) {
-    return false;
-  }
-
-  return plugin.authConfig.every((authFieldObj) => {
-    const authFieldOptions = authFieldObj.authField.split('||');
-    let isFieldAuthenticated = false;
-
-    for (const fieldOption of authFieldOptions) {
-      const envValue = process.env[fieldOption];
-      if (envValue && envValue.trim() !== '' && envValue !== AuthType.USER_PROVIDED) {
-        isFieldAuthenticated = true;
-        break;
-      }
-    }
-
-    return isFieldAuthenticated;
-  });
-};
-
 const getAvailablePluginsController = async (req, res) => {
   try {
     const cache = getLogStores(CacheKeys.CONFIG_STORE);
@@ -84,6 +47,45 @@ const getAvailablePluginsController = async (req, res) => {
   }
 };
 
+function createServerToolsCallback() {
+  /**
+   * @param {string} serverName
+   * @param {TPlugin[] | null} serverTools
+   */
+  return async function (serverName, serverTools) {
+    try {
+      const mcpToolsCache = getLogStores(CacheKeys.MCP_TOOLS);
+      if (!serverName || !mcpToolsCache) {
+        return;
+      }
+      await mcpToolsCache.set(serverName, serverTools);
+      logger.debug(`MCP tools for ${serverName} added to cache.`);
+    } catch (error) {
+      logger.error('Error retrieving MCP tools from cache:', error);
+    }
+  };
+}
+
+function createGetServerTools() {
+  /**
+   * Retrieves cached server tools
+   * @param {string} serverName
+   * @returns {Promise<TPlugin[] | null>}
+   */
+  return async function (serverName) {
+    try {
+      const mcpToolsCache = getLogStores(CacheKeys.MCP_TOOLS);
+      if (!mcpToolsCache) {
+        return null;
+      }
+      return await mcpToolsCache.get(serverName);
+    } catch (error) {
+      logger.error('Error retrieving MCP tools from cache:', error);
+      return null;
+    }
+  };
+}
+
 /**
  * Retrieves and returns a list of available tools, either from a cache or by reading a plugin manifest file.
  *
@@ -98,18 +100,33 @@ const getAvailablePluginsController = async (req, res) => {
  */
 const getAvailableTools = async (req, res) => {
   try {
+    const userId = req.user?.id;
+    const customConfig = await getCustomConfig();
     const cache = getLogStores(CacheKeys.CONFIG_STORE);
-    const cachedTools = await cache.get(CacheKeys.TOOLS);
-    if (cachedTools) {
-      res.status(200).json(cachedTools);
+    const cachedToolsArray = await cache.get(CacheKeys.TOOLS);
+    const cachedUserTools = await getCachedTools({ userId });
+    const userPlugins = convertMCPToolsToPlugins({ functionTools: cachedUserTools, customConfig });
+
+    if (cachedToolsArray != null && userPlugins != null) {
+      const dedupedTools = filterUniquePlugins([...userPlugins, ...cachedToolsArray]);
+      res.status(200).json(dedupedTools);
       return;
     }
 
+    // If not in cache, build from manifest
     let pluginManifest = availableTools;
-    const customConfig = await getCustomConfig();
     if (customConfig?.mcpServers != null) {
       const mcpManager = getMCPManager();
-      pluginManifest = await mcpManager.loadManifestTools(pluginManifest);
+      const flowsCache = getLogStores(CacheKeys.FLOWS);
+      const flowManager = flowsCache ? getFlowStateManager(flowsCache) : null;
+      const serverToolsCallback = createServerToolsCallback();
+      const getServerTools = createGetServerTools();
+      const mcpTools = await mcpManager.loadManifestTools({
+        flowManager,
+        serverToolsCallback,
+        getServerTools,
+      });
+      pluginManifest = [...mcpTools, ...pluginManifest];
     }
 
     /** @type {TPlugin[]} */
@@ -123,17 +140,61 @@ const getAvailableTools = async (req, res) => {
       }
     });
 
-    const toolDefinitions = req.app.locals.availableTools;
-    const tools = authenticatedPlugins.filter(
-      (plugin) =>
-        toolDefinitions[plugin.pluginKey] !== undefined ||
-        (plugin.toolkit === true &&
-          Object.keys(toolDefinitions).some((key) => getToolkitKey(key) === plugin.pluginKey)),
-    );
+    const toolDefinitions = (await getCachedTools({ includeGlobal: true })) || {};
 
-    await cache.set(CacheKeys.TOOLS, tools);
-    res.status(200).json(tools);
+    const toolsOutput = [];
+    for (const plugin of authenticatedPlugins) {
+      const isToolDefined = toolDefinitions[plugin.pluginKey] !== undefined;
+      const isToolkit =
+        plugin.toolkit === true &&
+        Object.keys(toolDefinitions).some(
+          (key) => getToolkitKey({ toolkits, toolName: key }) === plugin.pluginKey,
+        );
+
+      if (!isToolDefined && !isToolkit) {
+        continue;
+      }
+
+      const toolToAdd = { ...plugin };
+
+      if (!plugin.pluginKey.includes(Constants.mcp_delimiter)) {
+        toolsOutput.push(toolToAdd);
+        continue;
+      }
+
+      const parts = plugin.pluginKey.split(Constants.mcp_delimiter);
+      const serverName = parts[parts.length - 1];
+      const serverConfig = customConfig?.mcpServers?.[serverName];
+
+      if (!serverConfig?.customUserVars) {
+        toolsOutput.push(toolToAdd);
+        continue;
+      }
+
+      const customVarKeys = Object.keys(serverConfig.customUserVars);
+
+      if (customVarKeys.length === 0) {
+        toolToAdd.authConfig = [];
+        toolToAdd.authenticated = true;
+      } else {
+        toolToAdd.authConfig = Object.entries(serverConfig.customUserVars).map(([key, value]) => ({
+          authField: key,
+          label: value.title || key,
+          description: value.description || '',
+        }));
+        toolToAdd.authenticated = false;
+      }
+
+      toolsOutput.push(toolToAdd);
+    }
+    const finalTools = filterUniquePlugins(toolsOutput);
+    await cache.set(CacheKeys.TOOLS, finalTools);
+
+    const dedupedTools = filterUniquePlugins([...userPlugins, ...finalTools]);
+
+    res.status(200).json(dedupedTools);
   } catch (error) {
+    logger.error('[getAvailableTools]', error);
     res.status(500).json({ message: error.message });
   }
 };

api/server/controllers/PluginController.spec.js

@@ -0,0 +1,520 @@
+const { Constants } = require('librechat-data-provider');
+const { getCustomConfig, getCachedTools } = require('~/server/services/Config');
+const { getLogStores } = require('~/cache');
+
+// Mock the dependencies
+jest.mock('@librechat/data-schemas', () => ({
+  logger: {
+    debug: jest.fn(),
+    error: jest.fn(),
+  },
+}));
+
+jest.mock('~/server/services/Config', () => ({
+  getCustomConfig: jest.fn(),
+  getCachedTools: jest.fn(),
+}));
+
+jest.mock('~/server/services/ToolService', () => ({
+  getToolkitKey: jest.fn(),
+}));
+
+jest.mock('~/config', () => ({
+  getMCPManager: jest.fn(() => ({
+    loadManifestTools: jest.fn().mockResolvedValue([]),
+  })),
+  getFlowStateManager: jest.fn(),
+}));
+
+jest.mock('~/app/clients/tools', () => ({
+  availableTools: [],
+  toolkits: [],
+}));
+
+jest.mock('~/cache', () => ({
+  getLogStores: jest.fn(),
+}));
+
+jest.mock('@librechat/api', () => ({
+  getToolkitKey: jest.fn(),
+  checkPluginAuth: jest.fn(),
+  filterUniquePlugins: jest.fn(),
+  convertMCPToolsToPlugins: jest.fn(),
+}));
+
+// Import the actual module with the function we want to test
+const { getAvailableTools, getAvailablePluginsController } = require('./PluginController');
+const {
+  filterUniquePlugins,
+  checkPluginAuth,
+  convertMCPToolsToPlugins,
+  getToolkitKey,
+} = require('@librechat/api');
+
+describe('PluginController', () => {
+  let mockReq, mockRes, mockCache;
+
+  beforeEach(() => {
+    jest.clearAllMocks();
+    mockReq = { user: { id: 'test-user-id' } };
+    mockRes = { status: jest.fn().mockReturnThis(), json: jest.fn() };
+    mockCache = { get: jest.fn(), set: jest.fn() };
+    getLogStores.mockReturnValue(mockCache);
+  });
+
+  describe('getAvailablePluginsController', () => {
+    beforeEach(() => {
+      mockReq.app = { locals: { filteredTools: [], includedTools: [] } };
+    });
+
+    it('should use filterUniquePlugins to remove duplicate plugins', async () => {
+      const mockPlugins = [
+        { name: 'Plugin1', pluginKey: 'key1', description: 'First' },
+        { name: 'Plugin2', pluginKey: 'key2', description: 'Second' },
+      ];
+
+      mockCache.get.mockResolvedValue(null);
+      filterUniquePlugins.mockReturnValue(mockPlugins);
+      checkPluginAuth.mockReturnValue(true);
+
+      await getAvailablePluginsController(mockReq, mockRes);
+
+      expect(filterUniquePlugins).toHaveBeenCalled();
+      expect(mockRes.status).toHaveBeenCalledWith(200);
+      // The response includes authenticated: true for each plugin when checkPluginAuth returns true
+      expect(mockRes.json).toHaveBeenCalledWith([
+        { name: 'Plugin1', pluginKey: 'key1', description: 'First', authenticated: true },
+        { name: 'Plugin2', pluginKey: 'key2', description: 'Second', authenticated: true },
+      ]);
+    });
+
+    it('should use checkPluginAuth to verify plugin authentication', async () => {
+      const mockPlugin = { name: 'Plugin1', pluginKey: 'key1', description: 'First' };
+
+      mockCache.get.mockResolvedValue(null);
+      filterUniquePlugins.mockReturnValue([mockPlugin]);
+      checkPluginAuth.mockReturnValueOnce(true);
+
+      await getAvailablePluginsController(mockReq, mockRes);
+
+      expect(checkPluginAuth).toHaveBeenCalledWith(mockPlugin);
+      const responseData = mockRes.json.mock.calls[0][0];
+      expect(responseData[0].authenticated).toBe(true);
+    });
+
+    it('should return cached plugins when available', async () => {
+      const cachedPlugins = [
+        { name: 'CachedPlugin', pluginKey: 'cached', description: 'Cached plugin' },
+      ];
+
+      mockCache.get.mockResolvedValue(cachedPlugins);
+
+      await getAvailablePluginsController(mockReq, mockRes);
+
+      expect(filterUniquePlugins).not.toHaveBeenCalled();
+      expect(checkPluginAuth).not.toHaveBeenCalled();
+      expect(mockRes.json).toHaveBeenCalledWith(cachedPlugins);
+    });
+
+    it('should filter plugins based on includedTools', async () => {
+      const mockPlugins = [
+        { name: 'Plugin1', pluginKey: 'key1', description: 'First' },
+        { name: 'Plugin2', pluginKey: 'key2', description: 'Second' },
+      ];
+
+      mockReq.app.locals.includedTools = ['key1'];
+      mockCache.get.mockResolvedValue(null);
+      filterUniquePlugins.mockReturnValue(mockPlugins);
+      checkPluginAuth.mockReturnValue(false);
+
+      await getAvailablePluginsController(mockReq, mockRes);
+
+      const responseData = mockRes.json.mock.calls[0][0];
+      expect(responseData).toHaveLength(1);
+      expect(responseData[0].pluginKey).toBe('key1');
+    });
+  });
+
+  describe('getAvailableTools', () => {
+    it('should use convertMCPToolsToPlugins for user-specific MCP tools', async () => {
+      const mockUserTools = {
+        [`tool1${Constants.mcp_delimiter}server1`]: {
+          function: { name: 'tool1', description: 'Tool 1' },
+        },
+      };
+      const mockConvertedPlugins = [
+        {
+          name: 'tool1',
+          pluginKey: `tool1${Constants.mcp_delimiter}server1`,
+          description: 'Tool 1',
+        },
+      ];
+
+      mockCache.get.mockResolvedValue(null);
+      getCachedTools.mockResolvedValueOnce(mockUserTools);
+      convertMCPToolsToPlugins.mockReturnValue(mockConvertedPlugins);
+      filterUniquePlugins.mockImplementation((plugins) => plugins);
+      getCustomConfig.mockResolvedValue(null);
+
+      await getAvailableTools(mockReq, mockRes);
+
+      expect(convertMCPToolsToPlugins).toHaveBeenCalledWith({
+        functionTools: mockUserTools,
+        customConfig: null,
+      });
+    });
+
+    it('should use filterUniquePlugins to deduplicate combined tools', async () => {
+      const mockUserPlugins = [
+        { name: 'UserTool', pluginKey: 'user-tool', description: 'User tool' },
+      ];
+      const mockManifestPlugins = [
+        { name: 'ManifestTool', pluginKey: 'manifest-tool', description: 'Manifest tool' },
+      ];
+
+      mockCache.get.mockResolvedValue(mockManifestPlugins);
+      getCachedTools.mockResolvedValueOnce({});
+      convertMCPToolsToPlugins.mockReturnValue(mockUserPlugins);
+      filterUniquePlugins.mockReturnValue([...mockUserPlugins, ...mockManifestPlugins]);
+      getCustomConfig.mockResolvedValue(null);
+
+      await getAvailableTools(mockReq, mockRes);
+
+      // Should be called to deduplicate the combined array
+      expect(filterUniquePlugins).toHaveBeenLastCalledWith([
+        ...mockUserPlugins,
+        ...mockManifestPlugins,
+      ]);
+    });
+
+    it('should use checkPluginAuth to verify authentication status', async () => {
+      const mockPlugin = { name: 'Tool1', pluginKey: 'tool1', description: 'Tool 1' };
+
+      mockCache.get.mockResolvedValue(null);
+      getCachedTools.mockResolvedValue({});
+      convertMCPToolsToPlugins.mockReturnValue([]);
+      filterUniquePlugins.mockReturnValue([mockPlugin]);
+      checkPluginAuth.mockReturnValue(true);
+      getCustomConfig.mockResolvedValue(null);
+
+      // Mock getCachedTools second call to return tool definitions
+      getCachedTools.mockResolvedValueOnce({}).mockResolvedValueOnce({ tool1: true });
+
+      await getAvailableTools(mockReq, mockRes);
+
+      expect(checkPluginAuth).toHaveBeenCalledWith(mockPlugin);
+    });
+
+    it('should use getToolkitKey for toolkit validation', async () => {
+      const mockToolkit = {
+        name: 'Toolkit1',
+        pluginKey: 'toolkit1',
+        description: 'Toolkit 1',
+        toolkit: true,
+      };
+
+      mockCache.get.mockResolvedValue(null);
+      getCachedTools.mockResolvedValue({});
+      convertMCPToolsToPlugins.mockReturnValue([]);
+      filterUniquePlugins.mockReturnValue([mockToolkit]);
+      checkPluginAuth.mockReturnValue(false);
+      getToolkitKey.mockReturnValue('toolkit1');
+      getCustomConfig.mockResolvedValue(null);
+
+      // Mock getCachedTools second call to return tool definitions
+      getCachedTools.mockResolvedValueOnce({}).mockResolvedValueOnce({
+        toolkit1_function: true,
+      });
+
+      await getAvailableTools(mockReq, mockRes);
+
+      expect(getToolkitKey).toHaveBeenCalled();
+    });
+  });
+
+  describe('plugin.icon behavior', () => {
+    const callGetAvailableToolsWithMCPServer = async (mcpServers) => {
+      mockCache.get.mockResolvedValue(null);
+      getCustomConfig.mockResolvedValue({ mcpServers });
+
+      const functionTools = {
+        [`test-tool${Constants.mcp_delimiter}test-server`]: {
+          function: { name: 'test-tool', description: 'A test tool' },
+        },
+      };
+
+      const mockConvertedPlugin = {
+        name: 'test-tool',
+        pluginKey: `test-tool${Constants.mcp_delimiter}test-server`,
+        description: 'A test tool',
+        icon: mcpServers['test-server']?.iconPath,
+        authenticated: true,
+        authConfig: [],
+      };
+
+      getCachedTools.mockResolvedValueOnce(functionTools);
+      convertMCPToolsToPlugins.mockReturnValue([mockConvertedPlugin]);
+      filterUniquePlugins.mockImplementation((plugins) => plugins);
+      checkPluginAuth.mockReturnValue(true);
+      getToolkitKey.mockReturnValue(undefined);
+
+      getCachedTools.mockResolvedValueOnce({
+        [`test-tool${Constants.mcp_delimiter}test-server`]: true,
+      });
+
+      await getAvailableTools(mockReq, mockRes);
+      const responseData = mockRes.json.mock.calls[0][0];
+      return responseData.find((tool) => tool.name === 'test-tool');
+    };
+
+    it('should set plugin.icon when iconPath is defined', async () => {
+      const mcpServers = {
+        'test-server': {
+          iconPath: '/path/to/icon.png',
+        },
+      };
+      const testTool = await callGetAvailableToolsWithMCPServer(mcpServers);
+      expect(testTool.icon).toBe('/path/to/icon.png');
+    });
+
+    it('should set plugin.icon to undefined when iconPath is not defined', async () => {
+      const mcpServers = {
+        'test-server': {},
+      };
+      const testTool = await callGetAvailableToolsWithMCPServer(mcpServers);
+      expect(testTool.icon).toBeUndefined();
+    });
+  });
+
+  describe('helper function integration', () => {
+    it('should properly handle MCP tools with custom user variables', async () => {
+      const customConfig = {
+        mcpServers: {
+          'test-server': {
+            customUserVars: {
+              API_KEY: { title: 'API Key', description: 'Your API key' },
+            },
+          },
+        },
+      };
+
+      // We need to test the actual flow where MCP manager tools are included
+      const mcpManagerTools = [
+        {
+          name: 'tool1',
+          pluginKey: `tool1${Constants.mcp_delimiter}test-server`,
+          description: 'Tool 1',
+          authenticated: true,
+        },
+      ];
+
+      // Mock the MCP manager to return tools
+      const mockMCPManager = {
+        loadManifestTools: jest.fn().mockResolvedValue(mcpManagerTools),
+      };
+      require('~/config').getMCPManager.mockReturnValue(mockMCPManager);
+
+      mockCache.get.mockResolvedValue(null);
+      getCustomConfig.mockResolvedValue(customConfig);
+
+      // First call returns user tools (empty in this case)
+      getCachedTools.mockResolvedValueOnce({});
+
+      // Mock convertMCPToolsToPlugins to return empty array for user tools
+      convertMCPToolsToPlugins.mockReturnValue([]);
+
+      // Mock filterUniquePlugins to pass through
+      filterUniquePlugins.mockImplementation((plugins) => plugins || []);
+
+      // Mock checkPluginAuth
+      checkPluginAuth.mockReturnValue(true);
+
+      // Second call returns tool definitions
+      getCachedTools.mockResolvedValueOnce({
+        [`tool1${Constants.mcp_delimiter}test-server`]: true,
+      });
+
+      await getAvailableTools(mockReq, mockRes);
+
+      const responseData = mockRes.json.mock.calls[0][0];
+
+      // Find the MCP tool in the response
+      const mcpTool = responseData.find(
+        (tool) => tool.pluginKey === `tool1${Constants.mcp_delimiter}test-server`,
+      );
+
+      // The actual implementation adds authConfig and sets authenticated to false when customUserVars exist
+      expect(mcpTool).toBeDefined();
+      expect(mcpTool.authConfig).toEqual([
+        { authField: 'API_KEY', label: 'API Key', description: 'Your API key' },
+      ]);
+      expect(mcpTool.authenticated).toBe(false);
+    });
+
+    it('should handle error cases gracefully', async () => {
+      mockCache.get.mockRejectedValue(new Error('Cache error'));
+
+      await getAvailableTools(mockReq, mockRes);
+
+      expect(mockRes.status).toHaveBeenCalledWith(500);
+      expect(mockRes.json).toHaveBeenCalledWith({ message: 'Cache error' });
+    });
+  });
+
+  describe('edge cases with undefined/null values', () => {
+    it('should handle undefined cache gracefully', async () => {
+      getLogStores.mockReturnValue(undefined);
+
+      await getAvailableTools(mockReq, mockRes);
+
+      expect(mockRes.status).toHaveBeenCalledWith(500);
+    });
+
+    it('should handle null cachedTools and cachedUserTools', async () => {
+      mockCache.get.mockResolvedValue(null);
+      getCachedTools.mockResolvedValue(null);
+      convertMCPToolsToPlugins.mockReturnValue(undefined);
+      filterUniquePlugins.mockImplementation((plugins) => plugins || []);
+      getCustomConfig.mockResolvedValue(null);
+
+      await getAvailableTools(mockReq, mockRes);
+
+      expect(convertMCPToolsToPlugins).toHaveBeenCalledWith({
+        functionTools: null,
+        customConfig: null,
+      });
+    });
+
+    it('should handle when getCachedTools returns undefined', async () => {
+      mockCache.get.mockResolvedValue(null);
+      getCachedTools.mockResolvedValue(undefined);
+      convertMCPToolsToPlugins.mockReturnValue(undefined);
+      filterUniquePlugins.mockImplementation((plugins) => plugins || []);
+      getCustomConfig.mockResolvedValue(null);
+      checkPluginAuth.mockReturnValue(false);
+
+      // Mock getCachedTools to return undefined for both calls
+      getCachedTools.mockReset();
+      getCachedTools.mockResolvedValueOnce(undefined).mockResolvedValueOnce(undefined);
+
+      await getAvailableTools(mockReq, mockRes);
+
+      expect(convertMCPToolsToPlugins).toHaveBeenCalledWith({
+        functionTools: undefined,
+        customConfig: null,
+      });
+    });
+
+    it('should handle cachedToolsArray and userPlugins both being defined', async () => {
+      const cachedTools = [{ name: 'CachedTool', pluginKey: 'cached-tool', description: 'Cached' }];
+      const userTools = {
+        'user-tool': { function: { name: 'user-tool', description: 'User tool' } },
+      };
+      const userPlugins = [{ name: 'UserTool', pluginKey: 'user-tool', description: 'User tool' }];
+
+      mockCache.get.mockResolvedValue(cachedTools);
+      getCachedTools.mockResolvedValue(userTools);
+      convertMCPToolsToPlugins.mockReturnValue(userPlugins);
+      filterUniquePlugins.mockReturnValue([...userPlugins, ...cachedTools]);
+
+      await getAvailableTools(mockReq, mockRes);
+
+      expect(mockRes.status).toHaveBeenCalledWith(200);
+      expect(mockRes.json).toHaveBeenCalledWith([...userPlugins, ...cachedTools]);
+    });
+
+    it('should handle empty toolDefinitions object', async () => {
+      mockCache.get.mockResolvedValue(null);
+      getCachedTools.mockResolvedValueOnce({}).mockResolvedValueOnce({});
+      convertMCPToolsToPlugins.mockReturnValue([]);
+      filterUniquePlugins.mockImplementation((plugins) => plugins || []);
+      getCustomConfig.mockResolvedValue(null);
+      checkPluginAuth.mockReturnValue(true);
+
+      await getAvailableTools(mockReq, mockRes);
+
+      // With empty tool definitions, no tools should be in the final output
+      expect(mockRes.json).toHaveBeenCalledWith([]);
+    });
+
+    it('should handle MCP tools without customUserVars', async () => {
+      const customConfig = {
+        mcpServers: {
+          'test-server': {
+            // No customUserVars defined
+          },
+        },
+      };
+
+      const mockUserTools = {
+        [`tool1${Constants.mcp_delimiter}test-server`]: {
+          function: { name: 'tool1', description: 'Tool 1' },
+        },
+      };
+
+      mockCache.get.mockResolvedValue(null);
+      getCustomConfig.mockResolvedValue(customConfig);
+      getCachedTools.mockResolvedValueOnce(mockUserTools);
+
+      const mockPlugin = {
+        name: 'tool1',
+        pluginKey: `tool1${Constants.mcp_delimiter}test-server`,
+        description: 'Tool 1',
+        authenticated: true,
+        authConfig: [],
+      };
+
+      convertMCPToolsToPlugins.mockReturnValue([mockPlugin]);
+      filterUniquePlugins.mockImplementation((plugins) => plugins);
+      checkPluginAuth.mockReturnValue(true);
+
+      getCachedTools.mockResolvedValueOnce({
+        [`tool1${Constants.mcp_delimiter}test-server`]: true,
+      });
+
+      await getAvailableTools(mockReq, mockRes);
+
+      const responseData = mockRes.json.mock.calls[0][0];
+      expect(responseData[0].authenticated).toBe(true);
+      // The actual implementation doesn't set authConfig on tools without customUserVars
+      expect(responseData[0].authConfig).toEqual([]);
+    });
+
+    it('should handle req.app.locals with undefined filteredTools and includedTools', async () => {
+      mockReq.app = { locals: {} };
+      mockCache.get.mockResolvedValue(null);
+      filterUniquePlugins.mockReturnValue([]);
+      checkPluginAuth.mockReturnValue(false);
+
+      await getAvailablePluginsController(mockReq, mockRes);
+
+      expect(mockRes.status).toHaveBeenCalledWith(200);
+      expect(mockRes.json).toHaveBeenCalledWith([]);
+    });
+
+    it('should handle toolkit with undefined toolDefinitions keys', async () => {
+      const mockToolkit = {
+        name: 'Toolkit1',
+        pluginKey: 'toolkit1',
+        description: 'Toolkit 1',
+        toolkit: true,
+      };
+
+      mockCache.get.mockResolvedValue(null);
+      getCachedTools.mockResolvedValue({});
+      convertMCPToolsToPlugins.mockReturnValue([]);
+      filterUniquePlugins.mockReturnValue([mockToolkit]);
+      checkPluginAuth.mockReturnValue(false);
+      getToolkitKey.mockReturnValue(undefined);
+      getCustomConfig.mockResolvedValue(null);
+
+      // Mock getCachedTools second call to return null
+      getCachedTools.mockResolvedValueOnce({}).mockResolvedValueOnce(null);
+
+      await getAvailableTools(mockReq, mockRes);
+
+      // Should handle null toolDefinitions gracefully
+      expect(mockRes.status).toHaveBeenCalledWith(200);
+    });
+  });
+});

api/server/controllers/TwoFactorController.js

@@ -1,3 +1,4 @@
+const { encryptV3 } = require('@librechat/api');
 const { logger } = require('@librechat/data-schemas');
 const {
   verifyTOTP,
@@ -7,7 +8,6 @@ const {
   generateBackupCodes,
 } = require('~/server/services/twoFactorService');
 const { getUserById, updateUser } = require('~/models');
-const { encryptV3 } = require('~/server/utils/crypto');
 
 const safeAppTitle = (process.env.APP_TITLE || 'LibreChat').replace(/\s+/g, '');
 
@@ -99,10 +99,36 @@ const confirm2FA = async (req, res) => {
 
 /**
  * Disable 2FA by clearing the stored secret and backup codes.
+ * Requires verification with either TOTP token or backup code if 2FA is fully enabled.
  */
 const disable2FA = async (req, res) => {
   try {
     const userId = req.user.id;
+    const { token, backupCode } = req.body;
+    const user = await getUserById(userId);
+
+    if (!user || !user.totpSecret) {
+      return res.status(400).json({ message: '2FA is not setup for this user' });
+    }
+
+    if (user.twoFactorEnabled) {
+      const secret = await getTOTPSecret(user.totpSecret);
+      let isVerified = false;
+
+      if (token) {
+        isVerified = await verifyTOTP(secret, token);
+      } else if (backupCode) {
+        isVerified = await verifyBackupCode({ user, backupCode });
+      } else {
+        return res
+          .status(400)
+          .json({ message: 'Either token or backup code is required to disable 2FA' });
+      }
+
+      if (!isVerified) {
+        return res.status(401).json({ message: 'Invalid token or backup code' });
+      }
+    }
     await updateUser(userId, { totpSecret: null, backupCodes: [], twoFactorEnabled: false });
     return res.status(200).json();
   } catch (err) {

api/server/controllers/UserController.js

@@ -1,10 +1,5 @@
-const {
-  Tools,
-  FileSources,
-  webSearchKeys,
-  extractWebSearchEnvVars,
-} = require('librechat-data-provider');
 const { logger } = require('@librechat/data-schemas');
+const { webSearchKeys, extractWebSearchEnvVars, normalizeHttpError } = require('@librechat/api');
 const {
   getFiles,
   updateUser,
@@ -19,13 +14,12 @@ const { updateUserPluginAuth, deleteUserPluginAuth } = require('~/server/service
 const { updateUserPluginsService, deleteUserKey } = require('~/server/services/UserService');
 const { verifyEmail, resendVerificationEmail } = require('~/server/services/AuthService');
 const { needsRefresh, getNewS3URL } = require('~/server/services/Files/S3/crud');
+const { Tools, Constants, FileSources } = require('librechat-data-provider');
 const { processDeleteRequest } = require('~/server/services/Files/process');
-const { deleteAllSharedLinks } = require('~/models/Share');
+const { Transaction, Balance, User } = require('~/db/models');
 const { deleteToolCalls } = require('~/models/ToolCall');
-
-const Transaction = require('~/db/models').Transaction;
-const Balance = require('~/db/models').Balance;
-const User = require('~/db/models').User;
+const { deleteAllSharedLinks } = require('~/models');
+const { getMCPManager } = require('~/config');
 
 const getUserController = async (req, res) => {
   /** @type {MongoUser} */
@@ -95,8 +89,8 @@ const updateUserPluginsController = async (req, res) => {
 
       if (userPluginsService instanceof Error) {
         logger.error('[userPluginsService]', userPluginsService);
-        const { status, message } = userPluginsService;
-        res.status(status).send({ message });
+        const { status, message } = normalizeHttpError(userPluginsService);
+        return res.status(status).send({ message });
       }
     }
 
@@ -105,10 +99,22 @@ const updateUserPluginsController = async (req, res) => {
     }
 
     let keys = Object.keys(auth);
-    if (keys.length === 0 && pluginKey !== Tools.web_search) {
+    const values = Object.values(auth); // Used in 'install' block
+
+    const isMCPTool = pluginKey.startsWith('mcp_') || pluginKey.includes(Constants.mcp_delimiter);
+
+    // Early exit condition:
+    // If keys are empty (meaning auth: {} was likely sent for uninstall, or auth was empty for install)
+    // AND it's not web_search (which has special key handling to populate `keys` for uninstall)
+    // AND it's NOT (an uninstall action FOR an MCP tool - we need to proceed for this case to clear all its auth)
+    // THEN return.
+    if (
+      keys.length === 0 &&
+      pluginKey !== Tools.web_search &&
+      !(action === 'uninstall' && isMCPTool)
+    ) {
       return res.status(200).send();
     }
-    const values = Object.values(auth);
 
     /** @type {number} */
     let status = 200;
@@ -131,24 +137,64 @@ const updateUserPluginsController = async (req, res) => {
         authService = await updateUserPluginAuth(user.id, keys[i], pluginKey, values[i]);
         if (authService instanceof Error) {
           logger.error('[authService]', authService);
-          ({ status, message } = authService);
+          ({ status, message } = normalizeHttpError(authService));
         }
       }
     } else if (action === 'uninstall') {
-      for (let i = 0; i < keys.length; i++) {
-        authService = await deleteUserPluginAuth(user.id, keys[i]);
+      // const isMCPTool was defined earlier
+      if (isMCPTool && keys.length === 0) {
+        // This handles the case where auth: {} is sent for an MCP tool uninstall.
+        // It means "delete all credentials associated with this MCP pluginKey".
+        authService = await deleteUserPluginAuth(user.id, null, true, pluginKey);
         if (authService instanceof Error) {
-          logger.error('[authService]', authService);
-          ({ status, message } = authService);
+          logger.error(
+            `[authService] Error deleting all auth for MCP tool ${pluginKey}:`,
+            authService,
+          );
+          ({ status, message } = normalizeHttpError(authService));
+        }
+      } else {
+        // This handles:
+        // 1. Web_search uninstall (keys will be populated with all webSearchKeys if auth was {}).
+        // 2. Other tools uninstall (if keys were provided).
+        // 3. MCP tool uninstall if specific keys were provided in `auth` (not current frontend behavior).
+        // If keys is empty for non-MCP tools (and not web_search), this loop won't run, and nothing is deleted.
+        for (let i = 0; i < keys.length; i++) {
+          authService = await deleteUserPluginAuth(user.id, keys[i]); // Deletes by authField name
+          if (authService instanceof Error) {
+            logger.error('[authService] Error deleting specific auth key:', authService);
+            ({ status, message } = normalizeHttpError(authService));
+          }
         }
       }
     }
 
     if (status === 200) {
+      // If auth was updated successfully, disconnect MCP sessions as they might use these credentials
+      if (pluginKey.startsWith(Constants.mcp_prefix)) {
+        try {
+          const mcpManager = getMCPManager(user.id);
+          if (mcpManager) {
+            // Extract server name from pluginKey (format: "mcp_<serverName>")
+            const serverName = pluginKey.replace(Constants.mcp_prefix, '');
+            logger.info(
+              `[updateUserPluginsController] Disconnecting MCP server ${serverName} for user ${user.id} after plugin auth update for ${pluginKey}.`,
+            );
+            await mcpManager.disconnectUserConnection(user.id, serverName);
+          }
+        } catch (disconnectError) {
+          logger.error(
+            `[updateUserPluginsController] Error disconnecting MCP connection for user ${user.id} after plugin auth update:`,
+            disconnectError,
+          );
+          // Do not fail the request for this, but log it.
+        }
+      }
       return res.status(status).send();
     }
 
-    res.status(status).send({ message });
+    const normalized = normalizeHttpError({ status, message });
+    return res.status(normalized.status).send({ message: normalized.message });
   } catch (err) {
     logger.error('[updateUserPluginsController]', err);
     return res.status(500).json({ message: 'Something went wrong.' });
@@ -166,7 +212,11 @@ const deleteUserController = async (req, res) => {
     await Balance.deleteMany({ user: user._id }); // delete user balances
     await deletePresets(user.id); // delete user presets
     /* TODO: Delete Assistant Threads */
-    await deleteConvos(user.id); // delete user convos
+    try {
+      await deleteConvos(user.id); // delete user convos
+    } catch (error) {
+      logger.error('[deleteUserController] Error deleting user convos, likely no convos', error);
+    }
     await deleteUserPluginAuth(user.id, null, true); // delete user plugin auth
     await deleteUserById(user.id); // delete user
     await deleteAllSharedLinks(user.id); // delete user shared links

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

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

api/server/controllers/agents/callbacks.js

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

api/server/controllers/agents/client.js

@@ -1,43 +1,66 @@
-// const { HttpsProxyAgent } = require('https-proxy-agent');
-// const {
-// Constants,
-// ImageDetail,
-// EModelEndpoint,
-// resolveHeaders,
-// validateVisionModel,
-// mapModelToAzureConfig,
-// } = require('librechat-data-provider');
 require('events').EventEmitter.defaultMaxListeners = 100;
+const { logger } = require('@librechat/data-schemas');
+const { DynamicStructuredTool } = require('@langchain/core/tools');
+const { getBufferString, HumanMessage } = require('@langchain/core/messages');
+const {
+  sendEvent,
+  createRun,
+  Tokenizer,
+  checkAccess,
+  memoryInstructions,
+  formatContentStrings,
+  createMemoryProcessor,
+} = require('@librechat/api');
 const {
   Callback,
+  Providers,
   GraphEvents,
+  TitleMethod,
   formatMessage,
   formatAgentMessages,
-  formatContentStrings,
   getTokenCountForMessage,
   createMetadataAggregator,
 } = require('@librechat/agents');
 const {
   Constants,
+  Permissions,
   VisionModes,
   ContentTypes,
   EModelEndpoint,
-  KnownEndpoints,
+  PermissionTypes,
   isAgentsEndpoint,
   AgentCapabilities,
   bedrockInputSchema,
   removeNullishValues,
 } = require('librechat-data-provider');
-const { getCustomEndpointConfig, checkCapability } = require('~/server/services/Config');
+const {
+  findPluginAuthsByKeys,
+  getFormattedMemories,
+  deleteMemory,
+  setMemory,
+} = require('~/models');
+const { getMCPAuthMap, checkCapability, hasCustomUserVars } = require('~/server/services/Config');
 const { addCacheControl, createContextHandlers } = require('~/app/clients/prompts');
+const { initializeAgent } = require('~/server/services/Endpoints/agents/agent');
 const { spendTokens, spendStructuredTokens } = require('~/models/spendTokens');
-const { getBufferString, HumanMessage } = require('@langchain/core/messages');
 const { encodeAndFormat } = require('~/server/services/Files/images/encode');
-const initOpenAI = require('~/server/services/Endpoints/openAI/initialize');
-const Tokenizer = require('~/server/services/Tokenizer');
+const { getProviderConfig } = require('~/server/services/Endpoints');
 const BaseClient = require('~/app/clients/BaseClient');
-const { logger, sendEvent } = require('~/config');
-const { createRun } = require('./run');
+const { getRoleByName } = require('~/models/Role');
+const { loadAgent } = require('~/models/Agent');
+const { getMCPManager } = require('~/config');
+
+const omitTitleOptions = new Set([
+  'stream',
+  'thinking',
+  'streaming',
+  'clientOptions',
+  'thinkingConfig',
+  'thinkingBudget',
+  'includeThoughts',
+  'maxOutputTokens',
+  'additionalModelRequestFields',
+]);
 
 /**
  * @param {ServerRequest} req
@@ -48,21 +71,19 @@ const payloadParser = ({ req, agent, endpoint }) => {
   if (isAgentsEndpoint(endpoint)) {
     return { model: undefined };
   } else if (endpoint === EModelEndpoint.bedrock) {
-    return bedrockInputSchema.parse(agent.model_parameters);
+    const parsedValues = bedrockInputSchema.parse(agent.model_parameters);
+    if (parsedValues.thinking == null) {
+      parsedValues.thinking = false;
+    }
+    return parsedValues;
   }
   return req.body.endpointOption.model_parameters;
 };
 
-const legacyContentEndpoints = new Set([KnownEndpoints.groq, KnownEndpoints.deepseek]);
-
 const noSystemModelRegex = [/\b(o1-preview|o1-mini|amazon\.titan-text)\b/gi];
 
-// const { processMemory, memoryInstructions } = require('~/server/services/Endpoints/agents/memory');
-// const { getFormattedMemories } = require('~/models/Memory');
-// const { getCurrentDateTime } = require('~/utils');
-
 function createTokenCounter(encoding) {
-  return (message) => {
+  return function (message) {
     const countTokens = (text) => Tokenizer.getTokenCount(text, encoding);
     return getTokenCountForMessage(message, countTokens);
   };
@@ -123,6 +144,8 @@ class AgentClient extends BaseClient {
     this.usage;
     /** @type {Record<string, number>} */
     this.indexTokenCountMap = {};
+    /** @type {(messages: BaseMessage[]) => Promise<void>} */
+    this.processMemory;
   }
 
   /**
@@ -137,55 +160,10 @@ class AgentClient extends BaseClient {
   }
 
   /**
-   *
-   * Checks if the model is a vision model based on request attachments and sets the appropriate options:
-   * - Sets `this.modelOptions.model` to `gpt-4-vision-preview` if the request is a vision request.
-   * - Sets `this.isVisionModel` to `true` if vision request.
-   * - Deletes `this.modelOptions.stop` if vision request.
+   * `AgentClient` is not opinionated about vision requests, so we don't do anything here
    * @param {MongoFile[]} attachments
    */
-  checkVisionRequest(attachments) {
-    // if (!attachments) {
-    //   return;
-    // }
-    // const availableModels = this.options.modelsConfig?.[this.options.endpoint];
-    // if (!availableModels) {
-    //   return;
-    // }
-    // let visionRequestDetected = false;
-    // for (const file of attachments) {
-    //   if (file?.type?.includes('image')) {
-    //     visionRequestDetected = true;
-    //     break;
-    //   }
-    // }
-    // if (!visionRequestDetected) {
-    //   return;
-    // }
-    // this.isVisionModel = validateVisionModel({ model: this.modelOptions.model, availableModels });
-    // if (this.isVisionModel) {
-    //   delete this.modelOptions.stop;
-    //   return;
-    // }
-    // for (const model of availableModels) {
-    //   if (!validateVisionModel({ model, availableModels })) {
-    //     continue;
-    //   }
-    //   this.modelOptions.model = model;
-    //   this.isVisionModel = true;
-    //   delete this.modelOptions.stop;
-    //   return;
-    // }
-    // if (!availableModels.includes(this.defaultVisionModel)) {
-    //   return;
-    // }
-    // if (!validateVisionModel({ model: this.defaultVisionModel, availableModels })) {
-    //   return;
-    // }
-    // this.modelOptions.model = this.defaultVisionModel;
-    // this.isVisionModel = true;
-    // delete this.modelOptions.stop;
-  }
+  checkVisionRequest() {}
 
   getSaveOptions() {
     // TODO:
@@ -248,6 +226,42 @@ class AgentClient extends BaseClient {
     return files;
   }
 
+  async addDocuments(message, attachments) {
+    const documentResult =
+      await require('~/server/services/Files/documents').encodeAndFormatDocuments(
+        this.options.req,
+        attachments,
+        this.options.agent.provider,
+      );
+    message.documents =
+      documentResult.documents && documentResult.documents.length
+        ? documentResult.documents
+        : undefined;
+    return documentResult.files;
+  }
+
+  async processAttachments(message, attachments) {
+    const [imageFiles, documentFiles] = await Promise.all([
+      this.addImageURLs(message, attachments),
+      this.addDocuments(message, attachments),
+    ]);
+
+    const allFiles = [...imageFiles, ...documentFiles];
+    const seenFileIds = new Set();
+    const uniqueFiles = [];
+
+    for (const file of allFiles) {
+      if (file.file_id && !seenFileIds.has(file.file_id)) {
+        seenFileIds.add(file.file_id);
+        uniqueFiles.push(file);
+      } else if (!file.file_id) {
+        uniqueFiles.push(file);
+      }
+    }
+
+    return uniqueFiles;
+  }
+
   async buildMessages(
     messages,
     parentMessageId,
@@ -269,24 +283,6 @@ class AgentClient extends BaseClient {
       .filter(Boolean)
       .join('\n')
       .trim();
-    // this.systemMessage = getCurrentDateTime();
-    // const { withKeys, withoutKeys } = await getFormattedMemories({
-    //   userId: this.options.req.user.id,
-    // });
-    // processMemory({
-    //   userId: this.options.req.user.id,
-    //   message: this.options.req.body.text,
-    //   parentMessageId,
-    //   memory: withKeys,
-    //   thread_id: this.conversationId,
-    // }).catch((error) => {
-    //   logger.error('Memory Agent failed to process memory', error);
-    // });
-
-    // this.systemMessage += '\n\n' + memoryInstructions;
-    // if (withoutKeys) {
-    //   this.systemMessage += `\n\n# Existing memory about the user:\n${withoutKeys}`;
-    // }
 
     if (this.options.attachments) {
       const attachments = await this.options.attachments;
@@ -299,7 +295,7 @@ class AgentClient extends BaseClient {
         };
       }
 
-      const files = await this.addImageURLs(
+      const files = await this.processAttachments(
         orderedMessages[orderedMessages.length - 1],
         attachments,
       );
@@ -322,6 +318,23 @@ class AgentClient extends BaseClient {
         assistantName: this.options?.modelLabel,
       });
 
+      if (
+        message.documents &&
+        message.documents.length > 0 &&
+        message.role === 'user' &&
+        this.options.agent.provider === EModelEndpoint.anthropic
+      ) {
+        const contentParts = [];
+        contentParts.push(...message.documents);
+        if (message.image_urls && message.image_urls.length > 0) {
+          contentParts.push(...message.image_urls);
+        }
+        const textContent =
+          typeof formattedMessage.content === 'string' ? formattedMessage.content : '';
+        contentParts.push({ type: 'text', text: textContent });
+        formattedMessage.content = contentParts;
+      }
+
       if (message.ocr && i !== orderedMessages.length - 1) {
         if (typeof formattedMessage.content === 'string') {
           formattedMessage.content = message.ocr + '\n' + formattedMessage.content;
@@ -370,6 +383,37 @@ class AgentClient extends BaseClient {
       systemContent = this.augmentedPrompt + systemContent;
     }
 
+    // Inject MCP server instructions if available
+    const ephemeralAgent = this.options.req.body.ephemeralAgent;
+    let mcpServers = [];
+
+    // Check for ephemeral agent MCP servers
+    if (ephemeralAgent && ephemeralAgent.mcp && ephemeralAgent.mcp.length > 0) {
+      mcpServers = ephemeralAgent.mcp;
+    }
+    // Check for regular agent MCP tools
+    else if (this.options.agent && this.options.agent.tools) {
+      mcpServers = this.options.agent.tools
+        .filter(
+          (tool) =>
+            tool instanceof DynamicStructuredTool && tool.name.includes(Constants.mcp_delimiter),
+        )
+        .map((tool) => tool.name.split(Constants.mcp_delimiter).pop())
+        .filter(Boolean);
+    }
+
+    if (mcpServers.length > 0) {
+      try {
+        const mcpInstructions = getMCPManager().formatInstructionsForContext(mcpServers);
+        if (mcpInstructions) {
+          systemContent = [systemContent, mcpInstructions].filter(Boolean).join('\n\n');
+          logger.debug('[AgentClient] Injected MCP instructions for servers:', mcpServers);
+        }
+      } catch (error) {
+        logger.error('[AgentClient] Failed to inject MCP instructions:', error);
+      }
+    }
+
     if (systemContent) {
       this.options.agent.instructions = systemContent;
     }
@@ -399,9 +443,226 @@ class AgentClient extends BaseClient {
       opts.getReqData({ promptTokens });
     }
 
+    const withoutKeys = await this.useMemory();
+    if (withoutKeys) {
+      systemContent += `${memoryInstructions}\n\n# Existing memory about the user:\n${withoutKeys}`;
+    }
+
+    if (systemContent) {
+      this.options.agent.instructions = systemContent;
+    }
+
     return result;
   }
 
+  /**
+   * Creates a promise that resolves with the memory promise result or undefined after a timeout
+   * @param {Promise<(TAttachment | null)[] | undefined>} memoryPromise - The memory promise to await
+   * @param {number} timeoutMs - Timeout in milliseconds (default: 3000)
+   * @returns {Promise<(TAttachment | null)[] | undefined>}
+   */
+  async awaitMemoryWithTimeout(memoryPromise, timeoutMs = 3000) {
+    if (!memoryPromise) {
+      return;
+    }
+
+    try {
+      const timeoutPromise = new Promise((_, reject) =>
+        setTimeout(() => reject(new Error('Memory processing timeout')), timeoutMs),
+      );
+
+      const attachments = await Promise.race([memoryPromise, timeoutPromise]);
+      return attachments;
+    } catch (error) {
+      if (error.message === 'Memory processing timeout') {
+        logger.warn('[AgentClient] Memory processing timed out after 3 seconds');
+      } else {
+        logger.error('[AgentClient] Error processing memory:', error);
+      }
+      return;
+    }
+  }
+
+  /**
+   * @returns {Promise<string | undefined>}
+   */
+  async useMemory() {
+    const user = this.options.req.user;
+    if (user.personalization?.memories === false) {
+      return;
+    }
+    const hasAccess = await checkAccess({
+      user,
+      permissionType: PermissionTypes.MEMORIES,
+      permissions: [Permissions.USE],
+      getRoleByName,
+    });
+
+    if (!hasAccess) {
+      logger.debug(
+        `[api/server/controllers/agents/client.js #useMemory] User ${user.id} does not have USE permission for memories`,
+      );
+      return;
+    }
+    /** @type {TCustomConfig['memory']} */
+    const memoryConfig = this.options.req?.app?.locals?.memory;
+    if (!memoryConfig || memoryConfig.disabled === true) {
+      return;
+    }
+
+    /** @type {Agent} */
+    let prelimAgent;
+    const allowedProviders = new Set(
+      this.options.req?.app?.locals?.[EModelEndpoint.agents]?.allowedProviders,
+    );
+    try {
+      if (memoryConfig.agent?.id != null && memoryConfig.agent.id !== this.options.agent.id) {
+        prelimAgent = await loadAgent({
+          req: this.options.req,
+          agent_id: memoryConfig.agent.id,
+          endpoint: EModelEndpoint.agents,
+        });
+      } else if (
+        memoryConfig.agent?.id == null &&
+        memoryConfig.agent?.model != null &&
+        memoryConfig.agent?.provider != null
+      ) {
+        prelimAgent = { id: Constants.EPHEMERAL_AGENT_ID, ...memoryConfig.agent };
+      }
+    } catch (error) {
+      logger.error(
+        '[api/server/controllers/agents/client.js #useMemory] Error loading agent for memory',
+        error,
+      );
+    }
+
+    const agent = await initializeAgent({
+      req: this.options.req,
+      res: this.options.res,
+      agent: prelimAgent,
+      allowedProviders,
+      endpointOption: {
+        endpoint:
+          prelimAgent.id !== Constants.EPHEMERAL_AGENT_ID
+            ? EModelEndpoint.agents
+            : memoryConfig.agent?.provider,
+      },
+    });
+
+    if (!agent) {
+      logger.warn(
+        '[api/server/controllers/agents/client.js #useMemory] No agent found for memory',
+        memoryConfig,
+      );
+      return;
+    }
+
+    const llmConfig = Object.assign(
+      {
+        provider: agent.provider,
+        model: agent.model,
+      },
+      agent.model_parameters,
+    );
+
+    /** @type {import('@librechat/api').MemoryConfig} */
+    const config = {
+      validKeys: memoryConfig.validKeys,
+      instructions: agent.instructions,
+      llmConfig,
+      tokenLimit: memoryConfig.tokenLimit,
+    };
+
+    const userId = this.options.req.user.id + '';
+    const messageId = this.responseMessageId + '';
+    const conversationId = this.conversationId + '';
+    const [withoutKeys, processMemory] = await createMemoryProcessor({
+      userId,
+      config,
+      messageId,
+      conversationId,
+      memoryMethods: {
+        setMemory,
+        deleteMemory,
+        getFormattedMemories,
+      },
+      res: this.options.res,
+    });
+
+    this.processMemory = processMemory;
+    return withoutKeys;
+  }
+
+  /**
+   * Filters out image URLs from message content
+   * @param {BaseMessage} message - The message to filter
+   * @returns {BaseMessage} - A new message with image URLs removed
+   */
+  filterImageUrls(message) {
+    if (!message.content || typeof message.content === 'string') {
+      return message;
+    }
+
+    if (Array.isArray(message.content)) {
+      const filteredContent = message.content.filter(
+        (part) => part.type !== ContentTypes.IMAGE_URL,
+      );
+
+      if (filteredContent.length === 1 && filteredContent[0].type === ContentTypes.TEXT) {
+        const MessageClass = message.constructor;
+        return new MessageClass({
+          content: filteredContent[0].text,
+          additional_kwargs: message.additional_kwargs,
+        });
+      }
+
+      const MessageClass = message.constructor;
+      return new MessageClass({
+        content: filteredContent,
+        additional_kwargs: message.additional_kwargs,
+      });
+    }
+
+    return message;
+  }
+
+  /**
+   * @param {BaseMessage[]} messages
+   * @returns {Promise<void | (TAttachment | null)[]>}
+   */
+  async runMemory(messages) {
+    try {
+      if (this.processMemory == null) {
+        return;
+      }
+      /** @type {TCustomConfig['memory']} */
+      const memoryConfig = this.options.req?.app?.locals?.memory;
+      const messageWindowSize = memoryConfig?.messageWindowSize ?? 5;
+
+      let messagesToProcess = [...messages];
+      if (messages.length > messageWindowSize) {
+        for (let i = messages.length - messageWindowSize; i >= 0; i--) {
+          const potentialWindow = messages.slice(i, i + messageWindowSize);
+          if (potentialWindow[0]?.role === 'user') {
+            messagesToProcess = [...potentialWindow];
+            break;
+          }
+        }
+
+        if (messagesToProcess.length === messages.length) {
+          messagesToProcess = [...messages.slice(-messageWindowSize)];
+        }
+      }
+
+      const filteredMessages = messagesToProcess.map((msg) => this.filterImageUrls(msg));
+      const bufferString = getBufferString(filteredMessages);
+      const bufferMessage = new HumanMessage(`# Current Chat:\n\n${bufferString}`);
+      return await this.processMemory([bufferMessage]);
+    } catch (error) {
+      logger.error('Memory Agent failed to process memory', error);
+    }
+  }
+
   /** @type {sendCompletion} */
   async sendCompletion(payload, opts = {}) {
     await this.chatCompletion({
@@ -544,100 +805,13 @@ class AgentClient extends BaseClient {
     let config;
     /** @type {ReturnType<createRun>} */
     let run;
+    /** @type {Promise<(TAttachment | null)[] | undefined>} */
+    let memoryPromise;
     try {
       if (!abortController) {
         abortController = new AbortController();
       }
 
-      // if (this.options.headers) {
-      //   opts.defaultHeaders = { ...opts.defaultHeaders, ...this.options.headers };
-      // }
-
-      // if (this.options.proxy) {
-      //   opts.httpAgent = new HttpsProxyAgent(this.options.proxy);
-      // }
-
-      // if (this.isVisionModel) {
-      //   modelOptions.max_tokens = 4000;
-      // }
-
-      // /** @type {TAzureConfig | undefined} */
-      // const azureConfig = this.options?.req?.app?.locals?.[EModelEndpoint.azureOpenAI];
-
-      // if (
-      //   (this.azure && this.isVisionModel && azureConfig) ||
-      //   (azureConfig && this.isVisionModel && this.options.endpoint === EModelEndpoint.azureOpenAI)
-      // ) {
-      //   const { modelGroupMap, groupMap } = azureConfig;
-      //   const {
-      //     azureOptions,
-      //     baseURL,
-      //     headers = {},
-      //     serverless,
-      //   } = mapModelToAzureConfig({
-      //     modelName: modelOptions.model,
-      //     modelGroupMap,
-      //     groupMap,
-      //   });
-      //   opts.defaultHeaders = resolveHeaders(headers);
-      //   this.langchainProxy = extractBaseURL(baseURL);
-      //   this.apiKey = azureOptions.azureOpenAIApiKey;
-
-      //   const groupName = modelGroupMap[modelOptions.model].group;
-      //   this.options.addParams = azureConfig.groupMap[groupName].addParams;
-      //   this.options.dropParams = azureConfig.groupMap[groupName].dropParams;
-      //   // Note: `forcePrompt` not re-assigned as only chat models are vision models
-
-      //   this.azure = !serverless && azureOptions;
-      //   this.azureEndpoint =
-      //     !serverless && genAzureChatCompletion(this.azure, modelOptions.model, this);
-      // }
-
-      // if (this.azure || this.options.azure) {
-      //   /* Azure Bug, extremely short default `max_tokens` response */
-      //   if (!modelOptions.max_tokens && modelOptions.model === 'gpt-4-vision-preview') {
-      //     modelOptions.max_tokens = 4000;
-      //   }
-
-      //   /* Azure does not accept `model` in the body, so we need to remove it. */
-      //   delete modelOptions.model;
-
-      //   opts.baseURL = this.langchainProxy
-      //     ? constructAzureURL({
-      //       baseURL: this.langchainProxy,
-      //       azureOptions: this.azure,
-      //     })
-      //     : this.azureEndpoint.split(/(?<!\/)\/(chat|completion)\//)[0];
-
-      //   opts.defaultQuery = { 'api-version': this.azure.azureOpenAIApiVersion };
-      //   opts.defaultHeaders = { ...opts.defaultHeaders, 'api-key': this.apiKey };
-      // }
-
-      // if (process.env.OPENAI_ORGANIZATION) {
-      //   opts.organization = process.env.OPENAI_ORGANIZATION;
-      // }
-
-      // if (this.options.addParams && typeof this.options.addParams === 'object') {
-      //   modelOptions = {
-      //     ...modelOptions,
-      //     ...this.options.addParams,
-      //   };
-      //   logger.debug('[api/server/controllers/agents/client.js #chatCompletion] added params', {
-      //     addParams: this.options.addParams,
-      //     modelOptions,
-      //   });
-      // }
-
-      // if (this.options.dropParams && Array.isArray(this.options.dropParams)) {
-      //   this.options.dropParams.forEach((param) => {
-      //     delete modelOptions[param];
-      //   });
-      //   logger.debug('[api/server/controllers/agents/client.js #chatCompletion] dropped params', {
-      //     dropParams: this.options.dropParams,
-      //     modelOptions,
-      //   });
-      // }
-
       /** @type {TCustomConfig['endpoints']['agents']} */
       const agentsEConfig = this.options.req.app.locals[EModelEndpoint.agents];
 
@@ -647,22 +821,65 @@ class AgentClient extends BaseClient {
           last_agent_index: this.agentConfigs?.size ?? 0,
           user_id: this.user ?? this.options.req.user?.id,
           hide_sequential_outputs: this.options.agent.hide_sequential_outputs,
+          user: this.options.req.user,
         },
-        recursionLimit: agentsEConfig?.recursionLimit,
+        recursionLimit: agentsEConfig?.recursionLimit ?? 25,
         signal: abortController.signal,
         streamMode: 'values',
         version: 'v2',
       };
 
       const toolSet = new Set((this.options.agent.tools ?? []).map((tool) => tool && tool.name));
+
+      if (
+        this.options.agent.provider === EModelEndpoint.anthropic &&
+        payload &&
+        Array.isArray(payload)
+      ) {
+        let userMessageWithDocs = null;
+
+        if (this.userMessage?.documents) {
+          userMessageWithDocs = this.userMessage;
+        } else if (this.currentMessages?.length > 0) {
+          const lastMessage = this.currentMessages[this.currentMessages.length - 1];
+          if (lastMessage.documents?.length > 0) {
+            userMessageWithDocs = lastMessage;
+          }
+        } else if (this.messages?.length > 0) {
+          const lastMessage = this.messages[this.messages.length - 1];
+          if (lastMessage.documents?.length > 0) {
+            userMessageWithDocs = lastMessage;
+          }
+        }
+
+        if (userMessageWithDocs) {
+          for (const payloadMessage of payload) {
+            if (
+              payloadMessage.role === 'user' &&
+              userMessageWithDocs.text === payloadMessage.content
+            ) {
+              if (typeof payloadMessage.content === 'string') {
+                payloadMessage.content = [
+                  ...userMessageWithDocs.documents,
+                  { type: 'text', text: payloadMessage.content },
+                ];
+              } else if (Array.isArray(payloadMessage.content)) {
+                payloadMessage.content = [
+                  ...userMessageWithDocs.documents,
+                  ...payloadMessage.content,
+                ];
+              }
+              break;
+            }
+          }
+        }
+      }
+
       let { messages: initialMessages, indexTokenCountMap } = formatAgentMessages(
         payload,
         this.indexTokenCountMap,
         toolSet,
       );
-      if (legacyContentEndpoints.has(this.options.agent.endpoint?.toLowerCase())) {
-        initialMessages = formatContentStrings(initialMessages);
-      }
 
       /**
        *
@@ -678,6 +895,9 @@ class AgentClient extends BaseClient {
         if (i > 0) {
           this.model = agent.model_parameters.model;
         }
+        if (i > 0 && config.signal == null) {
+          config.signal = abortController.signal;
+        }
         if (agent.recursion_limit && typeof agent.recursion_limit === 'number') {
           config.recursionLimit = agent.recursion_limit;
         }
@@ -726,6 +946,9 @@ class AgentClient extends BaseClient {
         }
 
         let messages = _messages;
+        if (agent.useLegacyContent === true) {
+          messages = formatContentStrings(messages);
+        }
         if (
           agent.model_parameters?.clientOptions?.defaultHeaders?.['anthropic-beta']?.includes(
             'prompt-caching',
@@ -734,6 +957,10 @@ class AgentClient extends BaseClient {
           messages = addCacheControl(messages);
         }
 
+        if (i === 0) {
+          memoryPromise = this.runMemory(messages);
+        }
+
         run = await createRun({
           agent,
           req: this.options.req,
@@ -769,10 +996,24 @@ class AgentClient extends BaseClient {
           run.Graph.contentData = contentData;
         }
 
-        const encoding = this.getEncoding();
+        try {
+          if (await hasCustomUserVars()) {
+            config.configurable.userMCPAuthMap = await getMCPAuthMap({
+              tools: agent.tools,
+              userId: this.options.req.user.id,
+              findPluginAuthsByKeys,
+            });
+          }
+        } catch (err) {
+          logger.error(
+            `[api/server/controllers/agents/client.js #chatCompletion] Error getting custom user vars for agent ${agent.id}`,
+            err,
+          );
+        }
+
         await run.processStream({ messages }, config, {
           keepContent: i !== 0,
-          tokenCounter: createTokenCounter(encoding),
+          tokenCounter: createTokenCounter(this.getEncoding()),
           indexTokenCountMap: currentIndexCountMap,
           maxContextTokens: agent.maxContextTokens,
           callbacks: {
@@ -887,6 +1128,10 @@ class AgentClient extends BaseClient {
       });
 
       try {
+        const attachments = await this.awaitMemoryWithTimeout(memoryPromise);
+        if (attachments && attachments.length > 0) {
+          this.artifactPromises.push(...attachments);
+        }
         await this.recordCollectedUsage({ context: 'message' });
       } catch (err) {
         logger.error(
@@ -895,6 +1140,10 @@ class AgentClient extends BaseClient {
         );
       }
     } catch (err) {
+      const attachments = await this.awaitMemoryWithTimeout(memoryPromise);
+      if (attachments && attachments.length > 0) {
+        this.artifactPromises.push(...attachments);
+      }
       logger.error(
         '[api/server/controllers/agents/client.js #sendCompletion] Operation aborted',
         err,
@@ -923,23 +1172,41 @@ class AgentClient extends BaseClient {
       throw new Error('Run not initialized');
     }
     const { handleLLMEnd, collected: collectedMetadata } = createMetadataAggregator();
-    const endpoint = this.options.agent.endpoint;
-    const { req, res } = this.options;
+    const { req, res, agent } = this.options;
+    let endpoint = agent.endpoint;
+
     /** @type {import('@librechat/agents').ClientOptions} */
     let clientOptions = {
       maxTokens: 75,
+      model: agent.model || agent.model_parameters.model,
     };
-    let endpointConfig = req.app.locals[endpoint];
+
+    let titleProviderConfig = await getProviderConfig(endpoint);
+
+    /** @type {TEndpoint | undefined} */
+    const endpointConfig =
+      req.app.locals.all ?? req.app.locals[endpoint] ?? titleProviderConfig.customEndpointConfig;
     if (!endpointConfig) {
+      logger.warn(
+        '[api/server/controllers/agents/client.js #titleConvo] Error getting endpoint config',
+      );
+    }
+
+    if (endpointConfig?.titleEndpoint && endpointConfig.titleEndpoint !== endpoint) {
       try {
-        endpointConfig = await getCustomEndpointConfig(endpoint);
-      } catch (err) {
-        logger.error(
-          '[api/server/controllers/agents/client.js #titleConvo] Error getting custom endpoint config',
-          err,
+        titleProviderConfig = await getProviderConfig(endpointConfig.titleEndpoint);
+        endpoint = endpointConfig.titleEndpoint;
+      } catch (error) {
+        logger.warn(
+          `[api/server/controllers/agents/client.js #titleConvo] Error getting title endpoint config for ${endpointConfig.titleEndpoint}, falling back to default`,
+          error,
         );
+        // Fall back to original provider config
+        endpoint = agent.endpoint;
+        titleProviderConfig = await getProviderConfig(endpoint);
       }
     }
+
     if (
       endpointConfig &&
       endpointConfig.titleModel &&
@@ -947,33 +1214,71 @@ class AgentClient extends BaseClient {
     ) {
       clientOptions.model = endpointConfig.titleModel;
     }
+
+    const options = await titleProviderConfig.getOptions({
+      req,
+      res,
+      optionsOnly: true,
+      overrideEndpoint: endpoint,
+      overrideModel: clientOptions.model,
+      endpointOption: { model_parameters: clientOptions },
+    });
+
+    let provider = options.provider ?? titleProviderConfig.overrideProvider ?? agent.provider;
     if (
       endpoint === EModelEndpoint.azureOpenAI &&
-      clientOptions.model &&
-      this.options.agent.model_parameters.model !== clientOptions.model
+      options.llmConfig?.azureOpenAIApiInstanceName == null
     ) {
-      clientOptions =
-        (
-          await initOpenAI({
-            req,
-            res,
-            optionsOnly: true,
-            overrideModel: clientOptions.model,
-            overrideEndpoint: endpoint,
-            endpointOption: {
-              model_parameters: clientOptions,
-            },
-          })
-        )?.llmConfig ?? clientOptions;
+      provider = Providers.OPENAI;
+    } else if (
+      endpoint === EModelEndpoint.azureOpenAI &&
+      options.llmConfig?.azureOpenAIApiInstanceName != null &&
+      provider !== Providers.AZURE
+    ) {
+      provider = Providers.AZURE;
     }
-    if (/\b(o\d)\b/i.test(clientOptions.model) && clientOptions.maxTokens != null) {
+
+    /** @type {import('@librechat/agents').ClientOptions} */
+    clientOptions = { ...options.llmConfig };
+    if (options.configOptions) {
+      clientOptions.configuration = options.configOptions;
+    }
+
+    const shouldRemoveMaxTokens = /\b(o\d|gpt-[5-9])\b/i.test(clientOptions.model);
+    if (shouldRemoveMaxTokens && clientOptions.maxTokens != null) {
       delete clientOptions.maxTokens;
+    } else if (!shouldRemoveMaxTokens && !clientOptions.maxTokens) {
+      clientOptions.maxTokens = 75;
     }
+    if (shouldRemoveMaxTokens && clientOptions?.modelKwargs?.max_completion_tokens != null) {
+      delete clientOptions.modelKwargs.max_completion_tokens;
+    } else if (shouldRemoveMaxTokens && clientOptions?.modelKwargs?.max_output_tokens != null) {
+      delete clientOptions.modelKwargs.max_output_tokens;
+    }
+
+    clientOptions = Object.assign(
+      Object.fromEntries(
+        Object.entries(clientOptions).filter(([key]) => !omitTitleOptions.has(key)),
+      ),
+    );
+
+    if (
+      provider === Providers.GOOGLE &&
+      (endpointConfig?.titleMethod === TitleMethod.FUNCTIONS ||
+        endpointConfig?.titleMethod === TitleMethod.STRUCTURED)
+    ) {
+      clientOptions.json = true;
+    }
+
     try {
       const titleResult = await this.run.generateTitle({
+        provider,
+        clientOptions,
         inputText: text,
         contentParts: this.contentParts,
-        clientOptions,
+        titleMethod: endpointConfig?.titleMethod,
+        titlePrompt: endpointConfig?.titlePrompt,
+        titlePromptTemplate: endpointConfig?.titlePromptTemplate,
         chainOptions: {
           signal: abortController.signal,
           callbacks: [
@@ -988,8 +1293,10 @@ class AgentClient extends BaseClient {
         let input_tokens, output_tokens;
 
         if (item.usage) {
-          input_tokens = item.usage.input_tokens || item.usage.inputTokens;
-          output_tokens = item.usage.output_tokens || item.usage.outputTokens;
+          input_tokens =
+            item.usage.prompt_tokens || item.usage.input_tokens || item.usage.inputTokens;
+          output_tokens =
+            item.usage.completion_tokens || item.usage.output_tokens || item.usage.outputTokens;
         } else if (item.tokenUsage) {
           input_tokens = item.tokenUsage.promptTokens;
           output_tokens = item.tokenUsage.completionTokens;
@@ -1019,8 +1326,52 @@ class AgentClient extends BaseClient {
     }
   }
 
-  /** Silent method, as `recordCollectedUsage` is used instead */
-  async recordTokenUsage() {}
+  /**
+   * @param {object} params
+   * @param {number} params.promptTokens
+   * @param {number} params.completionTokens
+   * @param {OpenAIUsageMetadata} [params.usage]
+   * @param {string} [params.model]
+   * @param {string} [params.context='message']
+   * @returns {Promise<void>}
+   */
+  async recordTokenUsage({ model, promptTokens, completionTokens, usage, context = 'message' }) {
+    try {
+      await spendTokens(
+        {
+          model,
+          context,
+          conversationId: this.conversationId,
+          user: this.user ?? this.options.req.user?.id,
+          endpointTokenConfig: this.options.endpointTokenConfig,
+        },
+        { promptTokens, completionTokens },
+      );
+
+      if (
+        usage &&
+        typeof usage === 'object' &&
+        'reasoning_tokens' in usage &&
+        typeof usage.reasoning_tokens === 'number'
+      ) {
+        await spendTokens(
+          {
+            model,
+            context: 'reasoning',
+            conversationId: this.conversationId,
+            user: this.user ?? this.options.req.user?.id,
+            endpointTokenConfig: this.options.endpointTokenConfig,
+          },
+          { completionTokens: usage.reasoning_tokens },
+        );
+      }
+    } catch (error) {
+      logger.error(
+        '[api/server/controllers/agents/client.js #recordTokenUsage] Error recording token usage',
+        error,
+      );
+    }
+  }
 
   getEncoding() {
     return 'o200k_base';