.

Co-authored-by: Claude <noreply@anthropic.com>

.github/actions/build-backend-image/action.yml

@@ -1,73 +0,0 @@
-name: "Build Backend Image"
-description: "Builds and pushes the backend Docker image with cache reuse"
-inputs:
-  runs-on-ecr-cache:
-    description: "ECR cache registry from runs-on/action"
-    required: true
-  ref-name:
-    description: "Git ref name used for cache suffix fallback"
-    required: true
-  pr-number:
-    description: "Optional PR number for cache suffix"
-    required: false
-    default: ""
-  github-sha:
-    description: "Commit SHA used for cache keys"
-    required: true
-  run-id:
-    description: "GitHub run ID used in output image tag"
-    required: true
-  docker-username:
-    description: "Docker Hub username"
-    required: true
-  docker-token:
-    description: "Docker Hub token"
-    required: true
-  docker-no-cache:
-    description: "Set to 'true' to disable docker build cache"
-    required: false
-    default: "false"
-runs:
-  using: "composite"
-  steps:
-    - name: Format branch name for cache
-      id: format-branch
-      shell: bash
-      env:
-        PR_NUMBER: ${{ inputs.pr-number }}
-        REF_NAME: ${{ inputs.ref-name }}
-      run: |
-        if [ -n "${PR_NUMBER}" ]; then
-          CACHE_SUFFIX="${PR_NUMBER}"
-        else
-          # shellcheck disable=SC2001
-          CACHE_SUFFIX=$(echo "${REF_NAME}" | sed 's/[^A-Za-z0-9._-]/-/g')
-        fi
-        echo "cache-suffix=${CACHE_SUFFIX}" >> "$GITHUB_OUTPUT"
-
-    - name: Set up Docker Buildx
-      uses: docker/setup-buildx-action@8d2750c68a42422c14e847fe6c8ac0403b4cbd6f # ratchet:docker/setup-buildx-action@v3
-
-    - name: Login to Docker Hub
-      uses: docker/login-action@c94ce9fb468520275223c153574b00df6fe4bcc9 # ratchet:docker/login-action@v3
-      with:
-        username: ${{ inputs.docker-username }}
-        password: ${{ inputs.docker-token }}
-
-    - name: Build and push Backend Docker image
-      uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
-      with:
-        context: ./backend
-        file: ./backend/Dockerfile
-        push: true
-        tags: ${{ inputs.runs-on-ecr-cache }}:nightly-llm-it-backend-${{ inputs.run-id }}
-        cache-from: |
-          type=registry,ref=${{ inputs.runs-on-ecr-cache }}:backend-cache-${{ inputs.github-sha }}
-          type=registry,ref=${{ inputs.runs-on-ecr-cache }}:backend-cache-${{ steps.format-branch.outputs.cache-suffix }}
-          type=registry,ref=${{ inputs.runs-on-ecr-cache }}:backend-cache
-          type=registry,ref=onyxdotapp/onyx-backend:latest
-        cache-to: |
-          type=registry,ref=${{ inputs.runs-on-ecr-cache }}:backend-cache-${{ inputs.github-sha }},mode=max
-          type=registry,ref=${{ inputs.runs-on-ecr-cache }}:backend-cache-${{ steps.format-branch.outputs.cache-suffix }},mode=max
-          type=registry,ref=${{ inputs.runs-on-ecr-cache }}:backend-cache,mode=max
-        no-cache: ${{ inputs.docker-no-cache == 'true' }}

.github/actions/build-integration-image/action.yml

@@ -1,75 +0,0 @@
-name: "Build Integration Image"
-description: "Builds and pushes the integration test image with docker bake"
-inputs:
-  runs-on-ecr-cache:
-    description: "ECR cache registry from runs-on/action"
-    required: true
-  ref-name:
-    description: "Git ref name used for cache suffix fallback"
-    required: true
-  pr-number:
-    description: "Optional PR number for cache suffix"
-    required: false
-    default: ""
-  github-sha:
-    description: "Commit SHA used for cache keys"
-    required: true
-  run-id:
-    description: "GitHub run ID used in output image tag"
-    required: true
-  docker-username:
-    description: "Docker Hub username"
-    required: true
-  docker-token:
-    description: "Docker Hub token"
-    required: true
-runs:
-  using: "composite"
-  steps:
-    - name: Set up Docker Buildx
-      uses: docker/setup-buildx-action@8d2750c68a42422c14e847fe6c8ac0403b4cbd6f # ratchet:docker/setup-buildx-action@v3
-
-    - name: Login to Docker Hub
-      uses: docker/login-action@c94ce9fb468520275223c153574b00df6fe4bcc9 # ratchet:docker/login-action@v3
-      with:
-        username: ${{ inputs.docker-username }}
-        password: ${{ inputs.docker-token }}
-
-    - name: Format branch name for cache
-      id: format-branch
-      shell: bash
-      env:
-        PR_NUMBER: ${{ inputs.pr-number }}
-        REF_NAME: ${{ inputs.ref-name }}
-      run: |
-        if [ -n "${PR_NUMBER}" ]; then
-          CACHE_SUFFIX="${PR_NUMBER}"
-        else
-          # shellcheck disable=SC2001
-          CACHE_SUFFIX=$(echo "${REF_NAME}" | sed 's/[^A-Za-z0-9._-]/-/g')
-        fi
-        echo "cache-suffix=${CACHE_SUFFIX}" >> "$GITHUB_OUTPUT"
-
-    - name: Build and push integration test image with Docker Bake
-      shell: bash
-      env:
-        RUNS_ON_ECR_CACHE: ${{ inputs.runs-on-ecr-cache }}
-        TAG: nightly-llm-it-${{ inputs.run-id }}
-        CACHE_SUFFIX: ${{ steps.format-branch.outputs.cache-suffix }}
-        HEAD_SHA: ${{ inputs.github-sha }}
-      run: |
-        docker buildx bake --push \
-          --set backend.cache-from=type=registry,ref=${RUNS_ON_ECR_CACHE}:backend-cache-${HEAD_SHA} \
-          --set backend.cache-from=type=registry,ref=${RUNS_ON_ECR_CACHE}:backend-cache-${CACHE_SUFFIX} \
-          --set backend.cache-from=type=registry,ref=${RUNS_ON_ECR_CACHE}:backend-cache \
-          --set backend.cache-from=type=registry,ref=onyxdotapp/onyx-backend:latest \
-          --set backend.cache-to=type=registry,ref=${RUNS_ON_ECR_CACHE}:backend-cache-${HEAD_SHA},mode=max \
-          --set backend.cache-to=type=registry,ref=${RUNS_ON_ECR_CACHE}:backend-cache-${CACHE_SUFFIX},mode=max \
-          --set backend.cache-to=type=registry,ref=${RUNS_ON_ECR_CACHE}:backend-cache,mode=max \
-          --set integration.cache-from=type=registry,ref=${RUNS_ON_ECR_CACHE}:integration-cache-${HEAD_SHA} \
-          --set integration.cache-from=type=registry,ref=${RUNS_ON_ECR_CACHE}:integration-cache-${CACHE_SUFFIX} \
-          --set integration.cache-from=type=registry,ref=${RUNS_ON_ECR_CACHE}:integration-cache \
-          --set integration.cache-to=type=registry,ref=${RUNS_ON_ECR_CACHE}:integration-cache-${HEAD_SHA},mode=max \
-          --set integration.cache-to=type=registry,ref=${RUNS_ON_ECR_CACHE}:integration-cache-${CACHE_SUFFIX},mode=max \
-          --set integration.cache-to=type=registry,ref=${RUNS_ON_ECR_CACHE}:integration-cache,mode=max \
-          integration

.github/actions/build-model-server-image/action.yml

@@ -1,68 +0,0 @@
-name: "Build Model Server Image"
-description: "Builds and pushes the model server Docker image with cache reuse"
-inputs:
-  runs-on-ecr-cache:
-    description: "ECR cache registry from runs-on/action"
-    required: true
-  ref-name:
-    description: "Git ref name used for cache suffix fallback"
-    required: true
-  pr-number:
-    description: "Optional PR number for cache suffix"
-    required: false
-    default: ""
-  github-sha:
-    description: "Commit SHA used for cache keys"
-    required: true
-  run-id:
-    description: "GitHub run ID used in output image tag"
-    required: true
-  docker-username:
-    description: "Docker Hub username"
-    required: true
-  docker-token:
-    description: "Docker Hub token"
-    required: true
-runs:
-  using: "composite"
-  steps:
-    - name: Format branch name for cache
-      id: format-branch
-      shell: bash
-      env:
-        PR_NUMBER: ${{ inputs.pr-number }}
-        REF_NAME: ${{ inputs.ref-name }}
-      run: |
-        if [ -n "${PR_NUMBER}" ]; then
-          CACHE_SUFFIX="${PR_NUMBER}"
-        else
-          # shellcheck disable=SC2001
-          CACHE_SUFFIX=$(echo "${REF_NAME}" | sed 's/[^A-Za-z0-9._-]/-/g')
-        fi
-        echo "cache-suffix=${CACHE_SUFFIX}" >> "$GITHUB_OUTPUT"
-
-    - name: Set up Docker Buildx
-      uses: docker/setup-buildx-action@8d2750c68a42422c14e847fe6c8ac0403b4cbd6f # ratchet:docker/setup-buildx-action@v3
-
-    - name: Login to Docker Hub
-      uses: docker/login-action@c94ce9fb468520275223c153574b00df6fe4bcc9 # ratchet:docker/login-action@v3
-      with:
-        username: ${{ inputs.docker-username }}
-        password: ${{ inputs.docker-token }}
-
-    - name: Build and push Model Server Docker image
-      uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # ratchet:docker/build-push-action@v6
-      with:
-        context: ./backend
-        file: ./backend/Dockerfile.model_server
-        push: true
-        tags: ${{ inputs.runs-on-ecr-cache }}:nightly-llm-it-model-server-${{ inputs.run-id }}
-        cache-from: |
-          type=registry,ref=${{ inputs.runs-on-ecr-cache }}:model-server-cache-${{ inputs.github-sha }}
-          type=registry,ref=${{ inputs.runs-on-ecr-cache }}:model-server-cache-${{ steps.format-branch.outputs.cache-suffix }}
-          type=registry,ref=${{ inputs.runs-on-ecr-cache }}:model-server-cache
-          type=registry,ref=onyxdotapp/onyx-model-server:latest
-        cache-to: |
-          type=registry,ref=${{ inputs.runs-on-ecr-cache }}:model-server-cache-${{ inputs.github-sha }},mode=max
-          type=registry,ref=${{ inputs.runs-on-ecr-cache }}:model-server-cache-${{ steps.format-branch.outputs.cache-suffix }},mode=max
-          type=registry,ref=${{ inputs.runs-on-ecr-cache }}:model-server-cache,mode=max

.github/actions/run-nightly-provider-chat-test/action.yml

@@ -1,120 +0,0 @@
-name: "Run Nightly Provider Chat Test"
-description: "Starts required compose services and runs nightly provider integration test"
-inputs:
-  provider:
-    description: "Provider slug for NIGHTLY_LLM_PROVIDER"
-    required: true
-  models:
-    description: "Comma-separated model list for NIGHTLY_LLM_MODELS"
-    required: true
-  provider-api-key:
-    description: "API key for NIGHTLY_LLM_API_KEY"
-    required: true
-  strict:
-    description: "String true/false for NIGHTLY_LLM_STRICT"
-    required: true
-  api-base:
-    description: "Optional NIGHTLY_LLM_API_BASE"
-    required: false
-    default: ""
-  custom-config-json:
-    description: "Optional NIGHTLY_LLM_CUSTOM_CONFIG_JSON"
-    required: false
-    default: ""
-  runs-on-ecr-cache:
-    description: "ECR cache registry from runs-on/action"
-    required: true
-  run-id:
-    description: "GitHub run ID used in image tags"
-    required: true
-  docker-username:
-    description: "Docker Hub username"
-    required: true
-  docker-token:
-    description: "Docker Hub token"
-    required: true
-runs:
-  using: "composite"
-  steps:
-    - name: Login to Docker Hub
-      uses: docker/login-action@c94ce9fb468520275223c153574b00df6fe4bcc9 # ratchet:docker/login-action@v3
-      with:
-        username: ${{ inputs.docker-username }}
-        password: ${{ inputs.docker-token }}
-
-    - name: Create .env file for Docker Compose
-      shell: bash
-      env:
-        ECR_CACHE: ${{ inputs.runs-on-ecr-cache }}
-        RUN_ID: ${{ inputs.run-id }}
-      run: |
-        cat <<EOF2 > deployment/docker_compose/.env
-        COMPOSE_PROFILES=s3-filestore
-        ENABLE_PAID_ENTERPRISE_EDITION_FEATURES=true
-        LICENSE_ENFORCEMENT_ENABLED=false
-        AUTH_TYPE=basic
-        POSTGRES_POOL_PRE_PING=true
-        POSTGRES_USE_NULL_POOL=true
-        REQUIRE_EMAIL_VERIFICATION=false
-        DISABLE_TELEMETRY=true
-        INTEGRATION_TESTS_MODE=true
-        AUTO_LLM_UPDATE_INTERVAL_SECONDS=10
-        ONYX_BACKEND_IMAGE=${ECR_CACHE}:nightly-llm-it-backend-${RUN_ID}
-        ONYX_MODEL_SERVER_IMAGE=${ECR_CACHE}:nightly-llm-it-model-server-${RUN_ID}
-        EOF2
-
-    - name: Start Docker containers
-      shell: bash
-      run: |
-        cd deployment/docker_compose
-        docker compose -f docker-compose.yml -f docker-compose.dev.yml up -d --wait \
-          relational_db \
-          index \
-          cache \
-          minio \
-          api_server \
-          inference_model_server
-
-    - name: Run nightly provider integration test
-      uses: nick-fields/retry@ce71cc2ab81d554ebbe88c79ab5975992d79ba08 # ratchet:nick-fields/retry@v3
-      env:
-        MODELS: ${{ inputs.models }}
-        NIGHTLY_LLM_PROVIDER: ${{ inputs.provider }}
-        NIGHTLY_LLM_API_KEY: ${{ inputs.provider-api-key }}
-        NIGHTLY_LLM_API_BASE: ${{ inputs.api-base }}
-        NIGHTLY_LLM_CUSTOM_CONFIG_JSON: ${{ inputs.custom-config-json }}
-        NIGHTLY_LLM_STRICT: ${{ inputs.strict }}
-        RUNS_ON_ECR_CACHE: ${{ inputs.runs-on-ecr-cache }}
-        RUN_ID: ${{ inputs.run-id }}
-      with:
-        timeout_minutes: 20
-        max_attempts: 2
-        retry_wait_seconds: 10
-        command: |
-          if [ -z "${MODELS}" ]; then
-            echo "Input 'models' must be non-empty for provider '${NIGHTLY_LLM_PROVIDER}'."
-            exit 1
-          fi
-
-          docker run --rm --network onyx_default \
-            --name test-runner \
-            -e POSTGRES_HOST=relational_db \
-            -e POSTGRES_USER=postgres \
-            -e POSTGRES_PASSWORD=password \
-            -e POSTGRES_DB=postgres \
-            -e DB_READONLY_USER=db_readonly_user \
-            -e DB_READONLY_PASSWORD=password \
-            -e POSTGRES_POOL_PRE_PING=true \
-            -e POSTGRES_USE_NULL_POOL=true \
-            -e VESPA_HOST=index \
-            -e REDIS_HOST=cache \
-            -e API_SERVER_HOST=api_server \
-            -e TEST_WEB_HOSTNAME=test-runner \
-            -e NIGHTLY_LLM_PROVIDER="${NIGHTLY_LLM_PROVIDER}" \
-            -e NIGHTLY_LLM_MODELS="${MODELS}" \
-            -e NIGHTLY_LLM_API_KEY="${NIGHTLY_LLM_API_KEY}" \
-            -e NIGHTLY_LLM_API_BASE="${NIGHTLY_LLM_API_BASE}" \
-            -e NIGHTLY_LLM_CUSTOM_CONFIG_JSON="${NIGHTLY_LLM_CUSTOM_CONFIG_JSON}" \
-            -e NIGHTLY_LLM_STRICT="${NIGHTLY_LLM_STRICT}" \
-            ${RUNS_ON_ECR_CACHE}:nightly-llm-it-${RUN_ID} \
-            /app/tests/integration/tests/llm_workflows/test_nightly_provider_chat_workflow.py

.github/workflows/nightly-llm-provider-chat-openai.yml

@@ -1,44 +0,0 @@
-name: Nightly LLM Provider Chat Tests (OpenAI)
-concurrency:
-  group: Nightly-LLM-Provider-Chat-OpenAI-${{ github.workflow }}-${{ github.ref_name }}
-  cancel-in-progress: true
-
-on:
-  schedule:
-    # Runs daily at 10:30 UTC (2:30 AM PST / 3:30 AM PDT)
-    - cron: "30 10 * * *"
-  workflow_dispatch:
-
-permissions:
-  contents: read
-
-jobs:
-  openai-provider-chat-test:
-    uses: ./.github/workflows/reusable-nightly-llm-provider-chat.yml
-    with:
-      provider: openai
-      models: ${{ vars.NIGHTLY_LLM_OPENAI_MODELS }}
-      strict: true
-    secrets:
-      provider_api_key: ${{ secrets.OPENAI_API_KEY }}
-      DOCKER_USERNAME: ${{ secrets.DOCKER_USERNAME }}
-      DOCKER_TOKEN: ${{ secrets.DOCKER_TOKEN }}
-
-  notify-slack-on-failure:
-    needs: [openai-provider-chat-test]
-    if: failure() && github.event_name == 'schedule'
-    runs-on: ubuntu-slim
-    timeout-minutes: 5
-    steps:
-      - name: Checkout
-        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v6
-        with:
-          persist-credentials: false
-
-      - name: Send Slack notification
-        uses: ./.github/actions/slack-notify
-        with:
-          webhook-url: ${{ secrets.SLACK_WEBHOOK }}
-          failed-jobs: openai-provider-chat-test
-          title: "🚨 Scheduled OpenAI Provider Chat Tests failed!"
-          ref-name: ${{ github.ref_name }}

.github/workflows/post-merge-beta-cherry-pick.yml

@@ -11,11 +11,6 @@ permissions:
 
 jobs:
   cherry-pick-to-latest-release:
-    outputs:
-      should_cherrypick: ${{ steps.gate.outputs.should_cherrypick }}
-      pr_number: ${{ steps.gate.outputs.pr_number }}
-      cherry_pick_reason: ${{ steps.run_cherry_pick.outputs.reason }}
-      cherry_pick_details: ${{ steps.run_cherry_pick.outputs.details }}
     runs-on: ubuntu-latest
     timeout-minutes: 45
     steps:
@@ -41,13 +36,9 @@ jobs:
             exit 0
           fi
 
-          # Read the PR once so we can gate behavior and infer preferred actor.
-          pr_json="$(gh api "repos/${GITHUB_REPOSITORY}/pulls/${pr_number}")"
-          pr_body="$(printf '%s' "$pr_json" | jq -r '.body // ""')"
-          merged_by="$(printf '%s' "$pr_json" | jq -r '.merged_by.login // ""')"
-
+          # Read the PR body and check whether the helper checkbox is checked.
+          pr_body="$(gh api "repos/${GITHUB_REPOSITORY}/pulls/${pr_number}" --jq '.body // ""')"
           echo "pr_number=$pr_number" >> "$GITHUB_OUTPUT"
-          echo "merged_by=$merged_by" >> "$GITHUB_OUTPUT"
 
           if echo "$pr_body" | grep -qiE "\\[x\\][[:space:]]*(\\[[^]]+\\][[:space:]]*)?Please cherry-pick this PR to the latest release version"; then
             echo "should_cherrypick=true" >> "$GITHUB_OUTPUT"
@@ -80,82 +71,9 @@ jobs:
           git config user.email "github-actions[bot]@users.noreply.github.com"
 
       - name: Create cherry-pick PR to latest release
-        id: run_cherry_pick
         if: steps.gate.outputs.should_cherrypick == 'true'
-        continue-on-error: true
         env:
           GH_TOKEN: ${{ github.token }}
           GITHUB_TOKEN: ${{ github.token }}
-          CHERRY_PICK_ASSIGNEE: ${{ steps.gate.outputs.merged_by }}
         run: |
-          set -o pipefail
-          output_file="$(mktemp)"
-          uv run --no-sync --with onyx-devtools ods cherry-pick "${GITHUB_SHA}" --yes --no-verify 2>&1 | tee "$output_file"
-          exit_code="${PIPESTATUS[0]}"
-
-          if [ "${exit_code}" -eq 0 ]; then
-            echo "status=success" >> "$GITHUB_OUTPUT"
-            exit 0
-          fi
-
-          echo "status=failure" >> "$GITHUB_OUTPUT"
-
-          reason="command-failed"
-          if grep -qiE "merge conflict during cherry-pick|CONFLICT|could not apply|cherry-pick in progress with staged changes" "$output_file"; then
-            reason="merge-conflict"
-          fi
-          echo "reason=${reason}" >> "$GITHUB_OUTPUT"
-
-          {
-            echo "details<<EOF"
-            tail -n 40 "$output_file"
-            echo "EOF"
-          } >> "$GITHUB_OUTPUT"
-
-      - name: Mark workflow as failed if cherry-pick failed
-        if: steps.gate.outputs.should_cherrypick == 'true' && steps.run_cherry_pick.outputs.status == 'failure'
-        run: |
-          echo "::error::Automated cherry-pick failed (${{ steps.run_cherry_pick.outputs.reason }})."
-          exit 1
-
-  notify-slack-on-cherry-pick-failure:
-    needs:
-      - cherry-pick-to-latest-release
-    if: always() && needs.cherry-pick-to-latest-release.outputs.should_cherrypick == 'true' && needs.cherry-pick-to-latest-release.result != 'success'
-    runs-on: ubuntu-slim
-    timeout-minutes: 10
-    steps:
-      - name: Checkout
-        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v6
-        with:
-          persist-credentials: false
-
-      - name: Build cherry-pick failure summary
-        id: failure-summary
-        env:
-          SOURCE_PR_NUMBER: ${{ needs.cherry-pick-to-latest-release.outputs.pr_number }}
-          CHERRY_PICK_REASON: ${{ needs.cherry-pick-to-latest-release.outputs.cherry_pick_reason }}
-          CHERRY_PICK_DETAILS: ${{ needs.cherry-pick-to-latest-release.outputs.cherry_pick_details }}
-        run: |
-          source_pr_url="https://github.com/${GITHUB_REPOSITORY}/pull/${SOURCE_PR_NUMBER}"
-
-          reason_text="cherry-pick command failed"
-          if [ "${CHERRY_PICK_REASON}" = "merge-conflict" ]; then
-            reason_text="merge conflict during cherry-pick"
-          fi
-
-          details_excerpt="$(printf '%s' "${CHERRY_PICK_DETAILS}" | tail -n 8 | tr '\n' ' ' | sed "s/[[:space:]]\\+/ /g" | sed "s/\"/'/g" | cut -c1-350)"
-          failed_jobs="• cherry-pick-to-latest-release\\n• source PR: ${source_pr_url}\\n• reason: ${reason_text}"
-          if [ -n "${details_excerpt}" ]; then
-            failed_jobs="${failed_jobs}\\n• excerpt: ${details_excerpt}"
-          fi
-
-          echo "jobs=${failed_jobs}" >> "$GITHUB_OUTPUT"
-
-      - name: Notify #cherry-pick-prs about cherry-pick failure
-        uses: ./.github/actions/slack-notify
-        with:
-          webhook-url: ${{ secrets.CHERRY_PICK_PRS_WEBHOOK }}
-          failed-jobs: ${{ steps.failure-summary.outputs.jobs }}
-          title: "🚨 Automated Cherry-Pick Failed"
-          ref-name: ${{ github.ref_name }}
+          uv run --no-sync --with onyx-devtools ods cherry-pick "${GITHUB_SHA}" --yes --no-verify

.github/workflows/pr-external-dependency-unit-tests.yml

@@ -116,6 +116,7 @@ jobs:
         run: |
           cat <<EOF > deployment/docker_compose/.env
           COMPOSE_PROFILES=s3-filestore,opensearch-enabled
+          CODE_INTERPRETER_BETA_ENABLED=true
           DISABLE_TELEMETRY=true
           OPENSEARCH_FOR_ONYX_ENABLED=true
           EOF

.github/workflows/pr-integration-tests.yml

@@ -20,7 +20,6 @@ env:
   # Test Environment Variables
   OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
   SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
-  SLACK_BOT_TOKEN_TEST_SPACE: ${{ secrets.SLACK_BOT_TOKEN_TEST_SPACE }}
   CONFLUENCE_TEST_SPACE_URL: ${{ vars.CONFLUENCE_TEST_SPACE_URL }}
   CONFLUENCE_USER_NAME: ${{ vars.CONFLUENCE_USER_NAME }}
   CONFLUENCE_ACCESS_TOKEN: ${{ secrets.CONFLUENCE_ACCESS_TOKEN }}
@@ -424,7 +423,6 @@ jobs:
               -e OPENAI_API_KEY=${OPENAI_API_KEY} \
               -e EXA_API_KEY=${EXA_API_KEY} \
               -e SLACK_BOT_TOKEN=${SLACK_BOT_TOKEN} \
-              -e SLACK_BOT_TOKEN_TEST_SPACE=${SLACK_BOT_TOKEN_TEST_SPACE} \
               -e CONFLUENCE_TEST_SPACE_URL=${CONFLUENCE_TEST_SPACE_URL} \
               -e CONFLUENCE_USER_NAME=${CONFLUENCE_USER_NAME} \
               -e CONFLUENCE_ACCESS_TOKEN=${CONFLUENCE_ACCESS_TOKEN} \
@@ -445,7 +443,6 @@ jobs:
               -e TEST_WEB_HOSTNAME=test-runner \
               -e MOCK_CONNECTOR_SERVER_HOST=mock_connector_server \
               -e MOCK_CONNECTOR_SERVER_PORT=8001 \
-              -e ENABLE_PAID_ENTERPRISE_EDITION_FEATURES=${{ matrix.edition == 'ee' && 'true' || 'false' }} \
               ${{ env.RUNS_ON_ECR_CACHE }}:integration-test-${{ github.run_id }} \
               /app/tests/integration/${{ matrix.test-dir.path }}
 
@@ -704,7 +701,6 @@ jobs:
             -e OPENAI_API_KEY=${OPENAI_API_KEY} \
             -e EXA_API_KEY=${EXA_API_KEY} \
             -e SLACK_BOT_TOKEN=${SLACK_BOT_TOKEN} \
-            -e SLACK_BOT_TOKEN_TEST_SPACE=${SLACK_BOT_TOKEN_TEST_SPACE} \
             -e TEST_WEB_HOSTNAME=test-runner \
             -e AUTH_TYPE=cloud \
             -e MULTI_TENANT=true \

.github/workflows/reusable-nightly-llm-provider-chat.yml

@@ -1,206 +0,0 @@
-name: Reusable Nightly LLM Provider Chat Tests
-
-on:
-  workflow_call:
-    inputs:
-      provider:
-        description: "Provider slug passed to NIGHTLY_LLM_PROVIDER (e.g. openai, anthropic)"
-        required: true
-        type: string
-      models:
-        description: "Comma-separated model list passed to NIGHTLY_LLM_MODELS"
-        required: true
-        type: string
-      strict:
-        description: "Pass-through value for NIGHTLY_LLM_STRICT"
-        required: false
-        default: true
-        type: boolean
-      api_base:
-        description: "Optional NIGHTLY_LLM_API_BASE override"
-        required: false
-        default: ""
-        type: string
-      custom_config_json:
-        description: "Optional NIGHTLY_LLM_CUSTOM_CONFIG_JSON override"
-        required: false
-        default: ""
-        type: string
-    secrets:
-      provider_api_key:
-        description: "Provider API key passed to NIGHTLY_LLM_API_KEY"
-        required: true
-      DOCKER_USERNAME:
-        required: true
-      DOCKER_TOKEN:
-        required: true
-
-permissions:
-  contents: read
-
-env:
-  NIGHTLY_LLM_PROVIDER: ${{ inputs.provider }}
-  NIGHTLY_LLM_MODELS: ${{ inputs.models }}
-  NIGHTLY_LLM_API_KEY: ${{ secrets.provider_api_key }}
-  NIGHTLY_LLM_API_BASE: ${{ inputs.api_base }}
-  NIGHTLY_LLM_CUSTOM_CONFIG_JSON: ${{ inputs.custom_config_json }}
-  NIGHTLY_LLM_STRICT: ${{ inputs.strict && 'true' || 'false' }}
-
-jobs:
-  validate-inputs:
-    # NOTE: Keep this cheap and fail before image builds if required inputs are missing.
-    runs-on: ubuntu-slim
-    timeout-minutes: 5
-    steps:
-      - name: Validate required nightly provider inputs
-        run: |
-          if [ -z "${NIGHTLY_LLM_MODELS}" ]; then
-            echo "Input 'models' must be non-empty for provider '${NIGHTLY_LLM_PROVIDER}'."
-            exit 1
-          fi
-
-  build-backend-image:
-    needs: [validate-inputs]
-    runs-on:
-      [
-        runs-on,
-        runner=1cpu-linux-arm64,
-        "run-id=${{ github.run_id }}-build-backend-image",
-        "extras=ecr-cache",
-      ]
-    timeout-minutes: 45
-    steps:
-      - uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
-
-      - name: Checkout code
-        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v6
-        with:
-          persist-credentials: false
-
-      - name: Build backend image
-        uses: ./.github/actions/build-backend-image
-        with:
-          runs-on-ecr-cache: ${{ env.RUNS_ON_ECR_CACHE }}
-          ref-name: ${{ github.ref_name }}
-          pr-number: ${{ github.event.pull_request.number }}
-          github-sha: ${{ github.sha }}
-          run-id: ${{ github.run_id }}
-          docker-username: ${{ secrets.DOCKER_USERNAME }}
-          docker-token: ${{ secrets.DOCKER_TOKEN }}
-          docker-no-cache: ${{ vars.DOCKER_NO_CACHE == 'true' && 'true' || 'false' }}
-
-  build-model-server-image:
-    needs: [validate-inputs]
-    runs-on:
-      [
-        runs-on,
-        runner=1cpu-linux-arm64,
-        "run-id=${{ github.run_id }}-build-model-server-image",
-        "extras=ecr-cache",
-      ]
-    timeout-minutes: 45
-    steps:
-      - uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
-
-      - name: Checkout code
-        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v6
-        with:
-          persist-credentials: false
-
-      - name: Build model server image
-        uses: ./.github/actions/build-model-server-image
-        with:
-          runs-on-ecr-cache: ${{ env.RUNS_ON_ECR_CACHE }}
-          ref-name: ${{ github.ref_name }}
-          pr-number: ${{ github.event.pull_request.number }}
-          github-sha: ${{ github.sha }}
-          run-id: ${{ github.run_id }}
-          docker-username: ${{ secrets.DOCKER_USERNAME }}
-          docker-token: ${{ secrets.DOCKER_TOKEN }}
-
-  build-integration-image:
-    needs: [validate-inputs]
-    runs-on:
-      [
-        runs-on,
-        runner=2cpu-linux-arm64,
-        "run-id=${{ github.run_id }}-build-integration-image",
-        "extras=ecr-cache",
-      ]
-    timeout-minutes: 45
-    steps:
-      - uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
-
-      - name: Checkout code
-        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v6
-        with:
-          persist-credentials: false
-
-      - name: Build integration image
-        uses: ./.github/actions/build-integration-image
-        with:
-          runs-on-ecr-cache: ${{ env.RUNS_ON_ECR_CACHE }}
-          ref-name: ${{ github.ref_name }}
-          pr-number: ${{ github.event.pull_request.number }}
-          github-sha: ${{ github.sha }}
-          run-id: ${{ github.run_id }}
-          docker-username: ${{ secrets.DOCKER_USERNAME }}
-          docker-token: ${{ secrets.DOCKER_TOKEN }}
-
-  provider-chat-test:
-    needs:
-      [build-backend-image, build-model-server-image, build-integration-image]
-    runs-on:
-      - runs-on
-      - runner=4cpu-linux-arm64
-      - "run-id=${{ github.run_id }}-nightly-${{ inputs.provider }}-provider-chat-test"
-      - extras=ecr-cache
-    timeout-minutes: 45
-    steps:
-      - uses: runs-on/action@cd2b598b0515d39d78c38a02d529db87d2196d1e # ratchet:runs-on/action@v2
-
-      - name: Checkout code
-        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v6
-        with:
-          persist-credentials: false
-
-      - name: Run nightly provider chat test
-        uses: ./.github/actions/run-nightly-provider-chat-test
-        with:
-          provider: ${{ env.NIGHTLY_LLM_PROVIDER }}
-          models: ${{ env.NIGHTLY_LLM_MODELS }}
-          provider-api-key: ${{ secrets.provider_api_key }}
-          strict: ${{ env.NIGHTLY_LLM_STRICT }}
-          api-base: ${{ env.NIGHTLY_LLM_API_BASE }}
-          custom-config-json: ${{ env.NIGHTLY_LLM_CUSTOM_CONFIG_JSON }}
-          runs-on-ecr-cache: ${{ env.RUNS_ON_ECR_CACHE }}
-          run-id: ${{ github.run_id }}
-          docker-username: ${{ secrets.DOCKER_USERNAME }}
-          docker-token: ${{ secrets.DOCKER_TOKEN }}
-
-      - name: Dump API server logs
-        if: always()
-        run: |
-          cd deployment/docker_compose
-          docker compose logs --no-color api_server > $GITHUB_WORKSPACE/api_server.log || true
-
-      - name: Dump all-container logs
-        if: always()
-        run: |
-          cd deployment/docker_compose
-          docker compose logs --no-color > $GITHUB_WORKSPACE/docker-compose.log || true
-
-      - name: Upload logs
-        if: always()
-        uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
-        with:
-          name: docker-all-logs-nightly-${{ inputs.provider }}-llm-provider
-          path: |
-            ${{ github.workspace }}/api_server.log
-            ${{ github.workspace }}/docker-compose.log
-
-      - name: Stop Docker containers
-        if: always()
-        run: |
-          cd deployment/docker_compose
-          docker compose down -v

AGENTS.md

@@ -548,7 +548,7 @@ class in the utils over directly calling the APIs with a library like `requests`
 calling the utilities directly (e.g. do NOT create admin users with
 `admin_user = UserManager.create(name="admin_user")`, instead use the `admin_user` fixture).
 
-A great example of this type of test is `backend/tests/integration/tests/streaming_endpoints/test_chat_stream.py`.
+A great example of this type of test is `backend/tests/integration/dev_apis/test_simple_chat_api.py`.
 
 To run them:
 
@@ -616,9 +616,3 @@ This is a minimal list - feel free to include more. Do NOT write code as part of
 Keep it high level. You can reference certain files or functions though.
 
 Before writing your plan, make sure to do research. Explore the relevant sections in the codebase.
-
-## Best Practices
-
-In addition to the other content in this file, best practices for contributing
-to the codebase can be found at `contributing_guides/best_practices.md`.
-Understand its contents and follow them.

backend/alembic/versions/07b98176f1de_code_interpreter_seed.py

@@ -1,29 +0,0 @@
-"""code interpreter seed
-
-Revision ID: 07b98176f1de
-Revises: 7cb492013621
-Create Date: 2026-02-23 15:55:07.606784
-
-"""
-
-from alembic import op
-import sqlalchemy as sa
-
-
-# revision identifiers, used by Alembic.
-revision = "07b98176f1de"
-down_revision = "7cb492013621"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # Seed the single instance of code_interpreter_server
-    # NOTE: There should only exist at most and at minimum 1 code_interpreter_server row
-    op.execute(
-        sa.text("INSERT INTO code_interpreter_server (server_enabled) VALUES (true)")
-    )
-
-
-def downgrade() -> None:
-    op.execute(sa.text("DELETE FROM code_interpreter_server"))

backend/alembic/versions/7616121f6e97_add_enterprise_fields_to_scim_user_mapping.py

@@ -1,48 +0,0 @@
-"""add enterprise and name fields to scim_user_mapping
-
-Revision ID: 7616121f6e97
-Revises: 07b98176f1de
-Create Date: 2026-02-23 12:00:00.000000
-
-"""
-
-from alembic import op
-import sqlalchemy as sa
-
-
-# revision identifiers, used by Alembic.
-revision = "7616121f6e97"
-down_revision = "07b98176f1de"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    op.add_column(
-        "scim_user_mapping",
-        sa.Column("department", sa.String(), nullable=True),
-    )
-    op.add_column(
-        "scim_user_mapping",
-        sa.Column("manager", sa.String(), nullable=True),
-    )
-    op.add_column(
-        "scim_user_mapping",
-        sa.Column("given_name", sa.String(), nullable=True),
-    )
-    op.add_column(
-        "scim_user_mapping",
-        sa.Column("family_name", sa.String(), nullable=True),
-    )
-    op.add_column(
-        "scim_user_mapping",
-        sa.Column("scim_emails_json", sa.Text(), nullable=True),
-    )
-
-
-def downgrade() -> None:
-    op.drop_column("scim_user_mapping", "scim_emails_json")
-    op.drop_column("scim_user_mapping", "family_name")
-    op.drop_column("scim_user_mapping", "given_name")
-    op.drop_column("scim_user_mapping", "manager")
-    op.drop_column("scim_user_mapping", "department")

backend/alembic/versions/8ffcc2bcfc11_add_needs_persona_sync_to_user_file.py

@@ -1,33 +0,0 @@
-"""add needs_persona_sync to user_file
-
-Revision ID: 8ffcc2bcfc11
-Revises: 7616121f6e97
-Create Date: 2026-02-23 10:48:48.343826
-
-"""
-
-from alembic import op
-import sqlalchemy as sa
-
-
-# revision identifiers, used by Alembic.
-revision = "8ffcc2bcfc11"
-down_revision = "7616121f6e97"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    op.add_column(
-        "user_file",
-        sa.Column(
-            "needs_persona_sync",
-            sa.Boolean(),
-            nullable=False,
-            server_default=sa.text("false"),
-        ),
-    )
-
-
-def downgrade() -> None:
-    op.drop_column("user_file", "needs_persona_sync")

backend/alembic/versions/c0c937d5c9e5_llm_provider_deprecate_fields.py

@@ -1,70 +0,0 @@
-"""llm provider deprecate fields
-
-Revision ID: c0c937d5c9e5
-Revises: 8ffcc2bcfc11
-Create Date: 2026-02-25 17:35:46.125102
-
-"""
-
-from alembic import op
-import sqlalchemy as sa
-
-
-# revision identifiers, used by Alembic.
-revision = "c0c937d5c9e5"
-down_revision = "8ffcc2bcfc11"
-branch_labels = None
-depends_on = None
-
-
-def upgrade() -> None:
-    # Make default_model_name nullable (was NOT NULL)
-    op.alter_column(
-        "llm_provider",
-        "default_model_name",
-        existing_type=sa.String(),
-        nullable=True,
-    )
-
-    # Drop unique constraint on is_default_provider (defaults now tracked via LLMModelFlow)
-    op.drop_constraint(
-        "llm_provider_is_default_provider_key",
-        "llm_provider",
-        type_="unique",
-    )
-
-    # Remove server_default from is_default_vision_provider (was server_default=false())
-    op.alter_column(
-        "llm_provider",
-        "is_default_vision_provider",
-        existing_type=sa.Boolean(),
-        server_default=None,
-    )
-
-
-def downgrade() -> None:
-    # Restore default_model_name to NOT NULL (set empty string for any NULLs first)
-    op.execute(
-        "UPDATE llm_provider SET default_model_name = '' WHERE default_model_name IS NULL"
-    )
-    op.alter_column(
-        "llm_provider",
-        "default_model_name",
-        existing_type=sa.String(),
-        nullable=False,
-    )
-
-    # Restore unique constraint on is_default_provider
-    op.create_unique_constraint(
-        "llm_provider_is_default_provider_key",
-        "llm_provider",
-        ["is_default_provider"],
-    )
-
-    # Restore server_default for is_default_vision_provider
-    op.alter_column(
-        "llm_provider",
-        "is_default_vision_provider",
-        existing_type=sa.Boolean(),
-        server_default=sa.false(),
-    )

backend/ee/onyx/db/scim.py

@@ -34,7 +34,6 @@ from sqlalchemy.dialects.postgresql import insert as pg_insert
 
 from ee.onyx.server.scim.filtering import ScimFilter
 from ee.onyx.server.scim.filtering import ScimFilterOperator
-from ee.onyx.server.scim.models import ScimMappingFields
 from onyx.db.dal import DAL
 from onyx.db.models import ScimGroupMapping
 from onyx.db.models import ScimToken
@@ -129,19 +128,12 @@ class ScimDAL(DAL):
         external_id: str,
         user_id: UUID,
         scim_username: str | None = None,
-        fields: ScimMappingFields | None = None,
     ) -> ScimUserMapping:
         """Create a mapping between a SCIM externalId and an Onyx user."""
-        f = fields or ScimMappingFields()
         mapping = ScimUserMapping(
             external_id=external_id,
             user_id=user_id,
             scim_username=scim_username,
-            department=f.department,
-            manager=f.manager,
-            given_name=f.given_name,
-            family_name=f.family_name,
-            scim_emails_json=f.scim_emails_json,
         )
         self._session.add(mapping)
         self._session.flush()
@@ -319,14 +311,8 @@ class ScimDAL(DAL):
         user_id: UUID,
         new_external_id: str | None,
         scim_username: str | None = None,
-        fields: ScimMappingFields | None = None,
     ) -> None:
-        """Create, update, or delete the external ID mapping for a user.
-
-        When *fields* is provided, all mapping fields are written
-        unconditionally — including ``None`` values — so that a caller can
-        clear a previously-set field (e.g. removing a department).
-        """
+        """Create, update, or delete the external ID mapping for a user."""
         mapping = self.get_user_mapping_by_user_id(user_id)
         if new_external_id:
             if mapping:
@@ -334,18 +320,11 @@ class ScimDAL(DAL):
                     mapping.external_id = new_external_id
                 if scim_username is not None:
                     mapping.scim_username = scim_username
-                if fields is not None:
-                    mapping.department = fields.department
-                    mapping.manager = fields.manager
-                    mapping.given_name = fields.given_name
-                    mapping.family_name = fields.family_name
-                    mapping.scim_emails_json = fields.scim_emails_json
             else:
                 self.create_user_mapping(
                     external_id=new_external_id,
                     user_id=user_id,
                     scim_username=scim_username,
-                    fields=fields,
                 )
         elif mapping:
             self.delete_user_mapping(mapping.id)

backend/ee/onyx/external_permissions/sharepoint/permission_utils.py

@@ -4,6 +4,7 @@ from collections import deque
 from collections.abc import Callable
 from collections.abc import Generator
 from typing import Any
+from urllib.parse import unquote
 from urllib.parse import urlparse
 
 import requests as _requests
@@ -597,12 +598,8 @@ def get_external_access_from_sharepoint(
         )
     elif site_page:
         site_url = site_page.get("webUrl")
-        # Keep percent-encoding intact so the path matches the encoding
-        # used by the Office365 library's SPResPath.create_relative(),
-        # which compares against urlparse(context.base_url).path.
-        # Decoding (e.g. %27 → ') causes a mismatch that duplicates
-        # the site prefix in the constructed URL.
-        server_relative_url = urlparse(site_url).path
+        # Prefer server-relative URL to avoid OData filters that break on apostrophes
+        server_relative_url = unquote(urlparse(site_url).path)
         file_obj = client_context.web.get_file_by_server_relative_url(
             server_relative_url
         )

backend/ee/onyx/server/scim/api.py

@@ -26,14 +26,14 @@ from sqlalchemy.orm import Session
 from ee.onyx.db.scim import ScimDAL
 from ee.onyx.server.scim.auth import verify_scim_token
 from ee.onyx.server.scim.filtering import parse_scim_filter
-from ee.onyx.server.scim.models import SCIM_LIST_RESPONSE_SCHEMA
 from ee.onyx.server.scim.models import ScimError
 from ee.onyx.server.scim.models import ScimGroupMember
 from ee.onyx.server.scim.models import ScimGroupResource
 from ee.onyx.server.scim.models import ScimListResponse
-from ee.onyx.server.scim.models import ScimMappingFields
 from ee.onyx.server.scim.models import ScimName
 from ee.onyx.server.scim.models import ScimPatchRequest
+from ee.onyx.server.scim.models import ScimResourceType
+from ee.onyx.server.scim.models import ScimSchemaDefinition
 from ee.onyx.server.scim.models import ScimServiceProviderConfig
 from ee.onyx.server.scim.models import ScimUserResource
 from ee.onyx.server.scim.patch import apply_group_patch
@@ -41,8 +41,6 @@ from ee.onyx.server.scim.patch import apply_user_patch
 from ee.onyx.server.scim.patch import ScimPatchError
 from ee.onyx.server.scim.providers.base import get_default_provider
 from ee.onyx.server.scim.providers.base import ScimProvider
-from ee.onyx.server.scim.providers.base import serialize_emails
-from ee.onyx.server.scim.schema_definitions import ENTERPRISE_USER_SCHEMA_DEF
 from ee.onyx.server.scim.schema_definitions import GROUP_RESOURCE_TYPE
 from ee.onyx.server.scim.schema_definitions import GROUP_SCHEMA_DEF
 from ee.onyx.server.scim.schema_definitions import SERVICE_PROVIDER_CONFIG
@@ -50,28 +48,15 @@ from ee.onyx.server.scim.schema_definitions import USER_RESOURCE_TYPE
 from ee.onyx.server.scim.schema_definitions import USER_SCHEMA_DEF
 from onyx.db.engine.sql_engine import get_session
 from onyx.db.models import ScimToken
-from onyx.db.models import ScimUserMapping
 from onyx.db.models import User
 from onyx.db.models import UserGroup
 from onyx.db.models import UserRole
-from onyx.utils.logger import setup_logger
 from onyx.utils.variable_functionality import fetch_ee_implementation_or_noop
 
-logger = setup_logger()
-
-
-class ScimJSONResponse(JSONResponse):
-    """JSONResponse with Content-Type: application/scim+json (RFC 7644 §3.1)."""
-
-    media_type = "application/scim+json"
-
-
 # NOTE: All URL paths in this router (/ServiceProviderConfig, /ResourceTypes,
 # /Schemas, /Users, /Groups) are mandated by the SCIM spec (RFC 7643/7644).
 # IdPs like Okta and Azure AD hardcode these exact paths, so they cannot be
 # changed to kebab-case.
-
-
 scim_router = APIRouter(prefix="/scim/v2", tags=["SCIM"])
 
 _pw_helper = PasswordHelper()
@@ -101,39 +86,15 @@ def get_service_provider_config() -> ScimServiceProviderConfig:
 
 
 @scim_router.get("/ResourceTypes")
-def get_resource_types() -> ScimJSONResponse:
-    """List available SCIM resource types (RFC 7643 §6).
-
-    Wrapped in a ListResponse envelope (RFC 7644 §3.4.2) because IdPs
-    like Entra ID expect a JSON object, not a bare array.
-    """
-    resources = [USER_RESOURCE_TYPE, GROUP_RESOURCE_TYPE]
-    return ScimJSONResponse(
-        content={
-            "schemas": [SCIM_LIST_RESPONSE_SCHEMA],
-            "totalResults": len(resources),
-            "Resources": [
-                r.model_dump(exclude_none=True, by_alias=True) for r in resources
-            ],
-        }
-    )
+def get_resource_types() -> list[ScimResourceType]:
+    """List available SCIM resource types (RFC 7643 §6)."""
+    return [USER_RESOURCE_TYPE, GROUP_RESOURCE_TYPE]
 
 
 @scim_router.get("/Schemas")
-def get_schemas() -> ScimJSONResponse:
-    """Return SCIM schema definitions (RFC 7643 §7).
-
-    Wrapped in a ListResponse envelope (RFC 7644 §3.4.2) because IdPs
-    like Entra ID expect a JSON object, not a bare array.
-    """
-    schemas = [USER_SCHEMA_DEF, GROUP_SCHEMA_DEF, ENTERPRISE_USER_SCHEMA_DEF]
-    return ScimJSONResponse(
-        content={
-            "schemas": [SCIM_LIST_RESPONSE_SCHEMA],
-            "totalResults": len(schemas),
-            "Resources": [s.model_dump(exclude_none=True) for s in schemas],
-        }
-    )
+def get_schemas() -> list[ScimSchemaDefinition]:
+    """Return SCIM schema definitions (RFC 7643 §7)."""
+    return [USER_SCHEMA_DEF, GROUP_SCHEMA_DEF]
 
 
 # ---------------------------------------------------------------------------
@@ -141,45 +102,15 @@ def get_schemas() -> ScimJSONResponse:
 # ---------------------------------------------------------------------------
 
 
-def _scim_error_response(status: int, detail: str) -> ScimJSONResponse:
+def _scim_error_response(status: int, detail: str) -> JSONResponse:
     """Build a SCIM-compliant error response (RFC 7644 §3.12)."""
-    logger.warning("SCIM error response: status=%s detail=%s", status, detail)
     body = ScimError(status=str(status), detail=detail)
-    return ScimJSONResponse(
+    return JSONResponse(
         status_code=status,
         content=body.model_dump(exclude_none=True),
     )
 
 
-def _parse_excluded_attributes(raw: str | None) -> set[str]:
-    """Parse the ``excludedAttributes`` query parameter (RFC 7644 §3.4.2.5).
-
-    Returns a set of lowercased attribute names to omit from responses.
-    """
-    if not raw:
-        return set()
-    return {attr.strip().lower() for attr in raw.split(",") if attr.strip()}
-
-
-def _apply_exclusions(
-    resource: ScimUserResource | ScimGroupResource,
-    excluded: set[str],
-) -> dict:
-    """Serialize a SCIM resource, omitting attributes the IdP excluded.
-
-    RFC 7644 §3.4.2.5 lets the IdP pass ``?excludedAttributes=groups,emails``
-    to reduce response payload size. We strip those fields after serialization
-    so the rest of the pipeline doesn't need to know about them.
-    """
-    data = resource.model_dump(exclude_none=True, by_alias=True)
-    for attr in excluded:
-        # Match case-insensitively against the camelCase field names
-        keys_to_remove = [k for k in data if k.lower() == attr]
-        for k in keys_to_remove:
-            del data[k]
-    return data
-
-
 def _check_seat_availability(dal: ScimDAL) -> str | None:
     """Return an error message if seat limit is reached, else None."""
     check_fn = fetch_ee_implementation_or_noop(
@@ -193,7 +124,7 @@ def _check_seat_availability(dal: ScimDAL) -> str | None:
     return None
 
 
-def _fetch_user_or_404(user_id: str, dal: ScimDAL) -> User | ScimJSONResponse:
+def _fetch_user_or_404(user_id: str, dal: ScimDAL) -> User | JSONResponse:
     """Parse *user_id* as UUID, look up the user, or return a 404 error."""
     try:
         uid = UUID(user_id)
@@ -213,95 +144,10 @@ def _scim_name_to_str(name: ScimName | None) -> str | None:
     """
     if not name:
         return None
-    # If the client explicitly provides ``formatted``, prefer it — the client
-    # knows what display string it wants. Otherwise build from components.
-    if name.formatted:
-        return name.formatted
+    # Build from givenName/familyName first — IdPs like Okta may send a stale
+    # ``formatted`` value while updating the individual name components.
     parts = " ".join(part for part in [name.givenName, name.familyName] if part)
-    return parts or None
-
-
-def _scim_resource_response(
-    resource: ScimUserResource | ScimGroupResource | ScimListResponse,
-    status_code: int = 200,
-) -> ScimJSONResponse:
-    """Serialize a SCIM resource as ``application/scim+json``."""
-    content = resource.model_dump(exclude_none=True, by_alias=True)
-    return ScimJSONResponse(
-        status_code=status_code,
-        content=content,
-    )
-
-
-def _build_list_response(
-    resources: list[ScimUserResource | ScimGroupResource],
-    total: int,
-    start_index: int,
-    count: int,
-    excluded: set[str] | None = None,
-) -> ScimListResponse | ScimJSONResponse:
-    """Build a SCIM list response, optionally applying attribute exclusions.
-
-    RFC 7644 §3.4.2.5 — IdPs may request certain attributes be omitted via
-    the ``excludedAttributes`` query parameter.
-    """
-    if excluded:
-        envelope = ScimListResponse(
-            totalResults=total,
-            startIndex=start_index,
-            itemsPerPage=count,
-        )
-        data = envelope.model_dump(exclude_none=True)
-        data["Resources"] = [_apply_exclusions(r, excluded) for r in resources]
-        return ScimJSONResponse(content=data)
-
-    return _scim_resource_response(
-        ScimListResponse(
-            totalResults=total,
-            startIndex=start_index,
-            itemsPerPage=count,
-            Resources=resources,
-        )
-    )
-
-
-def _extract_enterprise_fields(
-    resource: ScimUserResource,
-) -> tuple[str | None, str | None]:
-    """Extract department and manager from enterprise extension."""
-    ext = resource.enterprise_extension
-    if not ext:
-        return None, None
-    department = ext.department
-    manager = ext.manager.value if ext.manager else None
-    return department, manager
-
-
-def _mapping_to_fields(
-    mapping: ScimUserMapping | None,
-) -> ScimMappingFields | None:
-    """Extract round-trip fields from a SCIM user mapping."""
-    if not mapping:
-        return None
-    return ScimMappingFields(
-        department=mapping.department,
-        manager=mapping.manager,
-        given_name=mapping.given_name,
-        family_name=mapping.family_name,
-        scim_emails_json=mapping.scim_emails_json,
-    )
-
-
-def _fields_from_resource(resource: ScimUserResource) -> ScimMappingFields:
-    """Build mapping fields from an incoming SCIM user resource."""
-    department, manager = _extract_enterprise_fields(resource)
-    return ScimMappingFields(
-        department=department,
-        manager=manager,
-        given_name=resource.name.givenName if resource.name else None,
-        family_name=resource.name.familyName if resource.name else None,
-        scim_emails_json=serialize_emails(resource.emails),
-    )
+    return parts or name.formatted
 
 
 # ---------------------------------------------------------------------------
@@ -312,13 +158,12 @@ def _fields_from_resource(resource: ScimUserResource) -> ScimMappingFields:
 @scim_router.get("/Users", response_model=None)
 def list_users(
     filter: str | None = Query(None),
-    excludedAttributes: str | None = None,
     startIndex: int = Query(1, ge=1),
     count: int = Query(100, ge=0, le=500),
     _token: ScimToken = Depends(verify_scim_token),
     provider: ScimProvider = Depends(_get_provider),
     db_session: Session = Depends(get_session),
-) -> ScimListResponse | ScimJSONResponse:
+) -> ScimListResponse | JSONResponse:
     """List users with optional SCIM filter and pagination."""
     dal = ScimDAL(db_session)
     dal.update_token_last_used(_token.id)
@@ -340,54 +185,42 @@ def list_users(
             mapping.external_id if mapping else None,
             groups=user_groups_map.get(user.id, []),
             scim_username=mapping.scim_username if mapping else None,
-            fields=_mapping_to_fields(mapping),
         )
         for user, mapping in users_with_mappings
     ]
 
-    return _build_list_response(
-        resources,
-        total,
-        startIndex,
-        count,
-        excluded=_parse_excluded_attributes(excludedAttributes),
+    return ScimListResponse(
+        totalResults=total,
+        startIndex=startIndex,
+        itemsPerPage=count,
+        Resources=resources,
     )
 
 
 @scim_router.get("/Users/{user_id}", response_model=None)
 def get_user(
     user_id: str,
-    excludedAttributes: str | None = None,
     _token: ScimToken = Depends(verify_scim_token),
     provider: ScimProvider = Depends(_get_provider),
     db_session: Session = Depends(get_session),
-) -> ScimUserResource | ScimJSONResponse:
+) -> ScimUserResource | JSONResponse:
     """Get a single user by ID."""
     dal = ScimDAL(db_session)
     dal.update_token_last_used(_token.id)
 
     result = _fetch_user_or_404(user_id, dal)
-    if isinstance(result, ScimJSONResponse):
+    if isinstance(result, JSONResponse):
         return result
     user = result
 
     mapping = dal.get_user_mapping_by_user_id(user.id)
-
-    resource = provider.build_user_resource(
+    return provider.build_user_resource(
         user,
         mapping.external_id if mapping else None,
         groups=dal.get_user_groups(user.id),
         scim_username=mapping.scim_username if mapping else None,
-        fields=_mapping_to_fields(mapping),
     )
 
-    # RFC 7644 §3.4.2.5 — IdP may request certain attributes be omitted
-    excluded = _parse_excluded_attributes(excludedAttributes)
-    if excluded:
-        return ScimJSONResponse(content=_apply_exclusions(resource, excluded))
-
-    return _scim_resource_response(resource)
-
 
 @scim_router.post("/Users", status_code=201, response_model=None)
 def create_user(
@@ -395,7 +228,7 @@ def create_user(
     _token: ScimToken = Depends(verify_scim_token),
     provider: ScimProvider = Depends(_get_provider),
     db_session: Session = Depends(get_session),
-) -> ScimUserResource | ScimJSONResponse:
+) -> ScimUserResource | JSONResponse:
     """Create a new user from a SCIM provisioning request."""
     dal = ScimDAL(db_session)
     dal.update_token_last_used(_token.id)
@@ -437,25 +270,13 @@ def create_user(
     # Create SCIM mapping (externalId is validated above, always present)
     external_id = user_resource.externalId
     scim_username = user_resource.userName.strip()
-    fields = _fields_from_resource(user_resource)
     dal.create_user_mapping(
-        external_id=external_id,
-        user_id=user.id,
-        scim_username=scim_username,
-        fields=fields,
+        external_id=external_id, user_id=user.id, scim_username=scim_username
     )
 
     dal.commit()
 
-    return _scim_resource_response(
-        provider.build_user_resource(
-            user,
-            external_id,
-            scim_username=scim_username,
-            fields=fields,
-        ),
-        status_code=201,
-    )
+    return provider.build_user_resource(user, external_id, scim_username=scim_username)
 
 
 @scim_router.put("/Users/{user_id}", response_model=None)
@@ -465,13 +286,13 @@ def replace_user(
     _token: ScimToken = Depends(verify_scim_token),
     provider: ScimProvider = Depends(_get_provider),
     db_session: Session = Depends(get_session),
-) -> ScimUserResource | ScimJSONResponse:
+) -> ScimUserResource | JSONResponse:
     """Replace a user entirely (RFC 7644 §3.5.1)."""
     dal = ScimDAL(db_session)
     dal.update_token_last_used(_token.id)
 
     result = _fetch_user_or_404(user_id, dal)
-    if isinstance(result, ScimJSONResponse):
+    if isinstance(result, JSONResponse):
         return result
     user = result
 
@@ -492,24 +313,15 @@ def replace_user(
 
     new_external_id = user_resource.externalId
     scim_username = user_resource.userName.strip()
-    fields = _fields_from_resource(user_resource)
-    dal.sync_user_external_id(
-        user.id,
-        new_external_id,
-        scim_username=scim_username,
-        fields=fields,
-    )
+    dal.sync_user_external_id(user.id, new_external_id, scim_username=scim_username)
 
     dal.commit()
 
-    return _scim_resource_response(
-        provider.build_user_resource(
-            user,
-            new_external_id,
-            groups=dal.get_user_groups(user.id),
-            scim_username=scim_username,
-            fields=fields,
-        )
+    return provider.build_user_resource(
+        user,
+        new_external_id,
+        groups=dal.get_user_groups(user.id),
+        scim_username=scim_username,
     )
 
 
@@ -520,7 +332,7 @@ def patch_user(
     _token: ScimToken = Depends(verify_scim_token),
     provider: ScimProvider = Depends(_get_provider),
     db_session: Session = Depends(get_session),
-) -> ScimUserResource | ScimJSONResponse:
+) -> ScimUserResource | JSONResponse:
     """Partially update a user (RFC 7644 §3.5.2).
 
     This is the primary endpoint for user deprovisioning — Okta sends
@@ -530,25 +342,23 @@ def patch_user(
     dal.update_token_last_used(_token.id)
 
     result = _fetch_user_or_404(user_id, dal)
-    if isinstance(result, ScimJSONResponse):
+    if isinstance(result, JSONResponse):
         return result
     user = result
 
     mapping = dal.get_user_mapping_by_user_id(user.id)
     external_id = mapping.external_id if mapping else None
     current_scim_username = mapping.scim_username if mapping else None
-    current_fields = _mapping_to_fields(mapping)
 
     current = provider.build_user_resource(
         user,
         external_id,
         groups=dal.get_user_groups(user.id),
         scim_username=current_scim_username,
-        fields=current_fields,
     )
 
     try:
-        patched, ent_data = apply_user_patch(
+        patched = apply_user_patch(
             patch_request.Operations, current, provider.ignored_patch_paths
         )
     except ScimPatchError as e:
@@ -583,37 +393,17 @@ def patch_user(
         personal_name=personal_name,
     )
 
-    # Build updated fields by merging PATCH enterprise data with current values
-    cf = current_fields or ScimMappingFields()
-    fields = ScimMappingFields(
-        department=ent_data.get("department", cf.department),
-        manager=ent_data.get("manager", cf.manager),
-        given_name=patched.name.givenName if patched.name else cf.given_name,
-        family_name=patched.name.familyName if patched.name else cf.family_name,
-        scim_emails_json=(
-            serialize_emails(patched.emails)
-            if patched.emails is not None
-            else cf.scim_emails_json
-        ),
-    )
-
     dal.sync_user_external_id(
-        user.id,
-        patched.externalId,
-        scim_username=new_scim_username,
-        fields=fields,
+        user.id, patched.externalId, scim_username=new_scim_username
     )
 
     dal.commit()
 
-    return _scim_resource_response(
-        provider.build_user_resource(
-            user,
-            patched.externalId,
-            groups=dal.get_user_groups(user.id),
-            scim_username=new_scim_username,
-            fields=fields,
-        )
+    return provider.build_user_resource(
+        user,
+        patched.externalId,
+        groups=dal.get_user_groups(user.id),
+        scim_username=new_scim_username,
     )
 
 
@@ -622,29 +412,25 @@ def delete_user(
     user_id: str,
     _token: ScimToken = Depends(verify_scim_token),
     db_session: Session = Depends(get_session),
-) -> Response | ScimJSONResponse:
+) -> Response | JSONResponse:
     """Delete a user (RFC 7644 §3.6).
 
     Deactivates the user and removes the SCIM mapping. Note that Okta
     typically uses PATCH active=false instead of DELETE.
-    A second DELETE returns 404 per RFC 7644 §3.6.
     """
     dal = ScimDAL(db_session)
     dal.update_token_last_used(_token.id)
 
     result = _fetch_user_or_404(user_id, dal)
-    if isinstance(result, ScimJSONResponse):
+    if isinstance(result, JSONResponse):
         return result
     user = result
 
-    # If no SCIM mapping exists, the user was already deleted from
-    # SCIM's perspective — return 404 per RFC 7644 §3.6.
-    mapping = dal.get_user_mapping_by_user_id(user.id)
-    if not mapping:
-        return _scim_error_response(404, f"User {user_id} not found")
-
     dal.deactivate_user(user)
-    dal.delete_user_mapping(mapping.id)
+
+    mapping = dal.get_user_mapping_by_user_id(user.id)
+    if mapping:
+        dal.delete_user_mapping(mapping.id)
 
     dal.commit()
 
@@ -656,7 +442,7 @@ def delete_user(
 # ---------------------------------------------------------------------------
 
 
-def _fetch_group_or_404(group_id: str, dal: ScimDAL) -> UserGroup | ScimJSONResponse:
+def _fetch_group_or_404(group_id: str, dal: ScimDAL) -> UserGroup | JSONResponse:
     """Parse *group_id* as int, look up the group, or return a 404 error."""
     try:
         gid = int(group_id)
@@ -711,13 +497,12 @@ def _validate_and_parse_members(
 @scim_router.get("/Groups", response_model=None)
 def list_groups(
     filter: str | None = Query(None),
-    excludedAttributes: str | None = None,
     startIndex: int = Query(1, ge=1),
     count: int = Query(100, ge=0, le=500),
     _token: ScimToken = Depends(verify_scim_token),
     provider: ScimProvider = Depends(_get_provider),
     db_session: Session = Depends(get_session),
-) -> ScimListResponse | ScimJSONResponse:
+) -> ScimListResponse | JSONResponse:
     """List groups with optional SCIM filter and pagination."""
     dal = ScimDAL(db_session)
     dal.update_token_last_used(_token.id)
@@ -737,46 +522,37 @@ def list_groups(
         for group, ext_id in groups_with_ext_ids
     ]
 
-    return _build_list_response(
-        resources,
-        total,
-        startIndex,
-        count,
-        excluded=_parse_excluded_attributes(excludedAttributes),
+    return ScimListResponse(
+        totalResults=total,
+        startIndex=startIndex,
+        itemsPerPage=count,
+        Resources=resources,
     )
 
 
 @scim_router.get("/Groups/{group_id}", response_model=None)
 def get_group(
     group_id: str,
-    excludedAttributes: str | None = None,
     _token: ScimToken = Depends(verify_scim_token),
     provider: ScimProvider = Depends(_get_provider),
     db_session: Session = Depends(get_session),
-) -> ScimGroupResource | ScimJSONResponse:
+) -> ScimGroupResource | JSONResponse:
     """Get a single group by ID."""
     dal = ScimDAL(db_session)
     dal.update_token_last_used(_token.id)
 
     result = _fetch_group_or_404(group_id, dal)
-    if isinstance(result, ScimJSONResponse):
+    if isinstance(result, JSONResponse):
         return result
     group = result
 
     mapping = dal.get_group_mapping_by_group_id(group.id)
     members = dal.get_group_members(group.id)
 
-    resource = provider.build_group_resource(
+    return provider.build_group_resource(
         group, members, mapping.external_id if mapping else None
     )
 
-    # RFC 7644 §3.4.2.5 — IdP may request certain attributes be omitted
-    excluded = _parse_excluded_attributes(excludedAttributes)
-    if excluded:
-        return ScimJSONResponse(content=_apply_exclusions(resource, excluded))
-
-    return _scim_resource_response(resource)
-
 
 @scim_router.post("/Groups", status_code=201, response_model=None)
 def create_group(
@@ -784,7 +560,7 @@ def create_group(
     _token: ScimToken = Depends(verify_scim_token),
     provider: ScimProvider = Depends(_get_provider),
     db_session: Session = Depends(get_session),
-) -> ScimGroupResource | ScimJSONResponse:
+) -> ScimGroupResource | JSONResponse:
     """Create a new group from a SCIM provisioning request."""
     dal = ScimDAL(db_session)
     dal.update_token_last_used(_token.id)
@@ -820,10 +596,7 @@ def create_group(
     dal.commit()
 
     members = dal.get_group_members(db_group.id)
-    return _scim_resource_response(
-        provider.build_group_resource(db_group, members, external_id),
-        status_code=201,
-    )
+    return provider.build_group_resource(db_group, members, external_id)
 
 
 @scim_router.put("/Groups/{group_id}", response_model=None)
@@ -833,13 +606,13 @@ def replace_group(
     _token: ScimToken = Depends(verify_scim_token),
     provider: ScimProvider = Depends(_get_provider),
     db_session: Session = Depends(get_session),
-) -> ScimGroupResource | ScimJSONResponse:
+) -> ScimGroupResource | JSONResponse:
     """Replace a group entirely (RFC 7644 §3.5.1)."""
     dal = ScimDAL(db_session)
     dal.update_token_last_used(_token.id)
 
     result = _fetch_group_or_404(group_id, dal)
-    if isinstance(result, ScimJSONResponse):
+    if isinstance(result, JSONResponse):
         return result
     group = result
 
@@ -854,9 +627,7 @@ def replace_group(
     dal.commit()
 
     members = dal.get_group_members(group.id)
-    return _scim_resource_response(
-        provider.build_group_resource(group, members, group_resource.externalId)
-    )
+    return provider.build_group_resource(group, members, group_resource.externalId)
 
 
 @scim_router.patch("/Groups/{group_id}", response_model=None)
@@ -866,7 +637,7 @@ def patch_group(
     _token: ScimToken = Depends(verify_scim_token),
     provider: ScimProvider = Depends(_get_provider),
     db_session: Session = Depends(get_session),
-) -> ScimGroupResource | ScimJSONResponse:
+) -> ScimGroupResource | JSONResponse:
     """Partially update a group (RFC 7644 §3.5.2).
 
     Handles member add/remove operations from Okta and Azure AD.
@@ -875,7 +646,7 @@ def patch_group(
     dal.update_token_last_used(_token.id)
 
     result = _fetch_group_or_404(group_id, dal)
-    if isinstance(result, ScimJSONResponse):
+    if isinstance(result, JSONResponse):
         return result
     group = result
 
@@ -914,9 +685,7 @@ def patch_group(
     dal.commit()
 
     members = dal.get_group_members(group.id)
-    return _scim_resource_response(
-        provider.build_group_resource(group, members, patched.externalId)
-    )
+    return provider.build_group_resource(group, members, patched.externalId)
 
 
 @scim_router.delete("/Groups/{group_id}", status_code=204, response_model=None)
@@ -924,13 +693,13 @@ def delete_group(
     group_id: str,
     _token: ScimToken = Depends(verify_scim_token),
     db_session: Session = Depends(get_session),
-) -> Response | ScimJSONResponse:
+) -> Response | JSONResponse:
     """Delete a group (RFC 7644 §3.6)."""
     dal = ScimDAL(db_session)
     dal.update_token_last_used(_token.id)
 
     result = _fetch_group_or_404(group_id, dal)
-    if isinstance(result, ScimJSONResponse):
+    if isinstance(result, JSONResponse):
         return result
     group = result
 

backend/ee/onyx/server/scim/models.py

@@ -7,14 +7,12 @@ SCIM protocol schemas follow the wire format defined in:
 Admin API schemas are internal to Onyx and used for SCIM token management.
 """
 
-from dataclasses import dataclass
 from datetime import datetime
 from enum import Enum
 
 from pydantic import BaseModel
 from pydantic import ConfigDict
 from pydantic import Field
-from pydantic import field_validator
 
 
 # ---------------------------------------------------------------------------
@@ -33,9 +31,6 @@ SCIM_SERVICE_PROVIDER_CONFIG_SCHEMA = (
 )
 SCIM_RESOURCE_TYPE_SCHEMA = "urn:ietf:params:scim:schemas:core:2.0:ResourceType"
 SCIM_SCHEMA_SCHEMA = "urn:ietf:params:scim:schemas:core:2.0:Schema"
-SCIM_ENTERPRISE_USER_SCHEMA = (
-    "urn:ietf:params:scim:schemas:extension:enterprise:2.0:User"
-)
 
 
 # ---------------------------------------------------------------------------
@@ -75,36 +70,6 @@ class ScimUserGroupRef(BaseModel):
     display: str | None = None
 
 
-class ScimManagerRef(BaseModel):
-    """Manager sub-attribute for the enterprise extension (RFC 7643 §4.3)."""
-
-    value: str | None = None
-
-
-class ScimEnterpriseExtension(BaseModel):
-    """Enterprise User extension attributes (RFC 7643 §4.3)."""
-
-    department: str | None = None
-    manager: ScimManagerRef | None = None
-
-
-@dataclass
-class ScimMappingFields:
-    """Stored SCIM mapping fields that need to round-trip through the IdP.
-
-    Entra ID sends structured name components, email metadata, and enterprise
-    extension attributes that must be returned verbatim in subsequent GET
-    responses. These fields are persisted on ScimUserMapping and threaded
-    through the DAL, provider, and endpoint layers.
-    """
-
-    department: str | None = None
-    manager: str | None = None
-    given_name: str | None = None
-    family_name: str | None = None
-    scim_emails_json: str | None = None
-
-
 class ScimUserResource(BaseModel):
     """SCIM User resource representation (RFC 7643 §4.1).
 
@@ -113,8 +78,6 @@ class ScimUserResource(BaseModel):
     to match the SCIM wire format (not Python convention).
     """
 
-    model_config = ConfigDict(populate_by_name=True)
-
     schemas: list[str] = Field(default_factory=lambda: [SCIM_USER_SCHEMA])
     id: str | None = None  # Onyx's internal user ID, set on responses
     externalId: str | None = None  # IdP's identifier for this user
@@ -125,10 +88,6 @@ class ScimUserResource(BaseModel):
     active: bool = True
     groups: list[ScimUserGroupRef] = Field(default_factory=list)
     meta: ScimMeta | None = None
-    enterprise_extension: ScimEnterpriseExtension | None = Field(
-        default=None,
-        alias="urn:ietf:params:scim:schemas:extension:enterprise:2.0:User",
-    )
 
 
 class ScimGroupMember(BaseModel):
@@ -206,19 +165,6 @@ class ScimPatchOperation(BaseModel):
     path: str | None = None
     value: ScimPatchValue = None
 
-    @field_validator("op", mode="before")
-    @classmethod
-    def normalize_operation(cls, v: object) -> object:
-        """Normalize op to lowercase for case-insensitive matching.
-
-        Some IdPs (e.g. Entra ID) send capitalized ops like ``"Replace"``
-        instead of ``"replace"``. This is safe for all providers since the
-        enum values are lowercase. If a future provider requires other
-        pre-processing quirks, move patch deserialization into the provider
-        subclass instead of adding more special cases here.
-        """
-        return v.lower() if isinstance(v, str) else v
-
 
 class ScimPatchRequest(BaseModel):
     """PATCH request body (RFC 7644 §3.5.2).

backend/ee/onyx/server/scim/patch.py

@@ -14,13 +14,8 @@ responsible for persisting changes.
 
 from __future__ import annotations
 
-import logging
 import re
-from dataclasses import dataclass
-from dataclasses import field
-from typing import Any
 
-from ee.onyx.server.scim.models import SCIM_ENTERPRISE_USER_SCHEMA
 from ee.onyx.server.scim.models import ScimGroupMember
 from ee.onyx.server.scim.models import ScimGroupResource
 from ee.onyx.server.scim.models import ScimPatchOperation
@@ -29,55 +24,6 @@ from ee.onyx.server.scim.models import ScimPatchResourceValue
 from ee.onyx.server.scim.models import ScimPatchValue
 from ee.onyx.server.scim.models import ScimUserResource
 
-logger = logging.getLogger(__name__)
-
-# Lowercased enterprise extension URN for case-insensitive matching
-_ENTERPRISE_URN_LOWER = SCIM_ENTERPRISE_USER_SCHEMA.lower()
-
-# Pattern for email filter paths, e.g.:
-#   emails[primary eq true].value  (Okta)
-#   emails[type eq "work"].value   (Azure AD / Entra ID)
-_EMAIL_FILTER_RE = re.compile(
-    r"^emails\[.+\]\.value$",
-    re.IGNORECASE,
-)
-
-# Pattern for member removal path: members[value eq "user-id"]
-_MEMBER_FILTER_RE = re.compile(
-    r'^members\[value\s+eq\s+"([^"]+)"\]$',
-    re.IGNORECASE,
-)
-
-# ---------------------------------------------------------------------------
-# Dispatch tables for user PATCH paths
-#
-# Maps lowercased SCIM path → (camelCase key, target dict name).
-# "data" writes to the top-level resource dict, "name" writes to the
-# name sub-object dict. This replaces the elif chains for simple fields.
-# ---------------------------------------------------------------------------
-
-_USER_REPLACE_PATHS: dict[str, tuple[str, str]] = {
-    "active": ("active", "data"),
-    "username": ("userName", "data"),
-    "externalid": ("externalId", "data"),
-    "name.givenname": ("givenName", "name"),
-    "name.familyname": ("familyName", "name"),
-    "name.formatted": ("formatted", "name"),
-}
-
-_USER_REMOVE_PATHS: dict[str, tuple[str, str]] = {
-    "externalid": ("externalId", "data"),
-    "name.givenname": ("givenName", "name"),
-    "name.familyname": ("familyName", "name"),
-    "name.formatted": ("formatted", "name"),
-    "displayname": ("displayName", "data"),
-}
-
-_GROUP_REPLACE_PATHS: dict[str, tuple[str, str]] = {
-    "displayname": ("displayName", "data"),
-    "externalid": ("externalId", "data"),
-}
-
 
 class ScimPatchError(Exception):
     """Raised when a PATCH operation cannot be applied."""
@@ -88,25 +34,18 @@ class ScimPatchError(Exception):
         super().__init__(detail)
 
 
-@dataclass
-class _UserPatchCtx:
-    """Bundles the mutable state for user PATCH operations."""
-
-    data: dict[str, Any]
-    name_data: dict[str, Any]
-    ent_data: dict[str, str | None] = field(default_factory=dict)
-
-
-# ---------------------------------------------------------------------------
-# User PATCH
-# ---------------------------------------------------------------------------
+# Pattern for member removal path: members[value eq "user-id"]
+_MEMBER_FILTER_RE = re.compile(
+    r'^members\[value\s+eq\s+"([^"]+)"\]$',
+    re.IGNORECASE,
+)
 
 
 def apply_user_patch(
     operations: list[ScimPatchOperation],
     current: ScimUserResource,
     ignored_paths: frozenset[str] = frozenset(),
-) -> tuple[ScimUserResource, dict[str, str | None]]:
+) -> ScimUserResource:
     """Apply SCIM PATCH operations to a user resource.
 
     Args:
@@ -114,185 +53,79 @@ def apply_user_patch(
         current: The current user resource state.
         ignored_paths: SCIM attribute paths to silently skip (from provider).
 
-    Returns:
-        A tuple of (modified user resource, enterprise extension data dict).
-        The enterprise dict has keys ``"department"`` and ``"manager"``
-        with values set only when a PATCH operation touched them.
+    Returns a new ``ScimUserResource`` with the modifications applied.
+    The original object is not mutated.
 
     Raises:
         ScimPatchError: If an operation targets an unsupported path.
     """
     data = current.model_dump()
-    ctx = _UserPatchCtx(data=data, name_data=data.get("name") or {})
+    name_data = data.get("name") or {}
 
     for op in operations:
-        if op.op in (ScimPatchOperationType.REPLACE, ScimPatchOperationType.ADD):
-            _apply_user_replace(op, ctx, ignored_paths)
-        elif op.op == ScimPatchOperationType.REMOVE:
-            _apply_user_remove(op, ctx, ignored_paths)
+        if op.op == ScimPatchOperationType.REPLACE:
+            _apply_user_replace(op, data, name_data, ignored_paths)
+        elif op.op == ScimPatchOperationType.ADD:
+            _apply_user_replace(op, data, name_data, ignored_paths)
         else:
             raise ScimPatchError(
                 f"Unsupported operation '{op.op.value}' on User resource"
             )
 
-    ctx.data["name"] = ctx.name_data
-    return ScimUserResource.model_validate(ctx.data), ctx.ent_data
+    data["name"] = name_data
+    return ScimUserResource.model_validate(data)
 
 
 def _apply_user_replace(
     op: ScimPatchOperation,
-    ctx: _UserPatchCtx,
+    data: dict,
+    name_data: dict,
     ignored_paths: frozenset[str],
 ) -> None:
     """Apply a replace/add operation to user data."""
     path = (op.path or "").lower()
 
     if not path:
-        # No path — value is a resource dict of top-level attributes to set.
+        # No path — value is a resource dict of top-level attributes to set
         if isinstance(op.value, ScimPatchResourceValue):
             for key, val in op.value.model_dump(exclude_unset=True).items():
-                _set_user_field(key.lower(), val, ctx, ignored_paths, strict=False)
+                _set_user_field(key.lower(), val, data, name_data, ignored_paths)
         else:
             raise ScimPatchError("Replace without path requires a dict value")
         return
 
-    _set_user_field(path, op.value, ctx, ignored_paths)
-
-
-def _apply_user_remove(
-    op: ScimPatchOperation,
-    ctx: _UserPatchCtx,
-    ignored_paths: frozenset[str],
-) -> None:
-    """Apply a remove operation to user data — clears the target field."""
-    path = (op.path or "").lower()
-    if not path:
-        raise ScimPatchError("Remove operation requires a path")
-
-    if path in ignored_paths:
-        return
-
-    entry = _USER_REMOVE_PATHS.get(path)
-    if entry:
-        key, target = entry
-        target_dict = ctx.data if target == "data" else ctx.name_data
-        target_dict[key] = None
-        return
-
-    raise ScimPatchError(f"Unsupported remove path '{path}' for User PATCH")
+    _set_user_field(path, op.value, data, name_data, ignored_paths)
 
 
 def _set_user_field(
     path: str,
     value: ScimPatchValue,
-    ctx: _UserPatchCtx,
+    data: dict,
+    name_data: dict,
     ignored_paths: frozenset[str],
-    *,
-    strict: bool = True,
 ) -> None:
-    """Set a single field on user data by SCIM path.
-
-    Args:
-        strict: When ``False`` (path-less replace), unknown attributes are
-            silently skipped.  When ``True`` (explicit path), they raise.
-    """
+    """Set a single field on user data by SCIM path."""
     if path in ignored_paths:
         return
-
-    # Simple field writes handled by the dispatch table
-    entry = _USER_REPLACE_PATHS.get(path)
-    if entry:
-        key, target = entry
-        target_dict = ctx.data if target == "data" else ctx.name_data
-        target_dict[key] = value
-        return
-
-    # displayName sets both the top-level field and the name.formatted sub-field
-    if path == "displayname":
-        ctx.data["displayName"] = value
-        ctx.name_data["formatted"] = value
-    elif path == "name":
-        if isinstance(value, dict):
-            for k, v in value.items():
-                ctx.name_data[k] = v
-    elif path == "emails":
-        if isinstance(value, list):
-            ctx.data["emails"] = value
-    elif _EMAIL_FILTER_RE.match(path):
-        _update_primary_email(ctx.data, value)
-    elif path.startswith(_ENTERPRISE_URN_LOWER):
-        _set_enterprise_field(path, value, ctx.ent_data)
-    elif not strict:
-        return
+    elif path == "active":
+        data["active"] = value
+    elif path == "username":
+        data["userName"] = value
+    elif path == "externalid":
+        data["externalId"] = value
+    elif path == "name.givenname":
+        name_data["givenName"] = value
+    elif path == "name.familyname":
+        name_data["familyName"] = value
+    elif path == "name.formatted":
+        name_data["formatted"] = value
+    elif path == "displayname":
+        data["displayName"] = value
+        name_data["formatted"] = value
     else:
         raise ScimPatchError(f"Unsupported path '{path}' for User PATCH")
 
 
-def _update_primary_email(data: dict[str, Any], value: ScimPatchValue) -> None:
-    """Update the primary email entry via an email filter path."""
-    emails: list[dict] = data.get("emails") or []
-    for email_entry in emails:
-        if email_entry.get("primary"):
-            email_entry["value"] = value
-            break
-    else:
-        emails.append({"value": value, "type": "work", "primary": True})
-    data["emails"] = emails
-
-
-def _to_dict(value: ScimPatchValue) -> dict | None:
-    """Coerce a SCIM patch value to a plain dict if possible.
-
-    Pydantic may parse raw dicts as ``ScimPatchResourceValue`` (which uses
-    ``extra="allow"``), so we also dump those back to a dict.
-    """
-    if isinstance(value, dict):
-        return value
-    if isinstance(value, ScimPatchResourceValue):
-        return value.model_dump(exclude_unset=True)
-    return None
-
-
-def _set_enterprise_field(
-    path: str,
-    value: ScimPatchValue,
-    ent_data: dict[str, str | None],
-) -> None:
-    """Handle enterprise extension URN paths or value dicts."""
-    # Full URN as key with dict value (path-less PATCH)
-    # e.g. key="urn:...:user", value={"department": "Eng", "manager": {...}}
-    if path == _ENTERPRISE_URN_LOWER:
-        d = _to_dict(value)
-        if d is not None:
-            if "department" in d:
-                ent_data["department"] = d["department"]
-            if "manager" in d:
-                mgr = d["manager"]
-                if isinstance(mgr, dict):
-                    ent_data["manager"] = mgr.get("value")
-        return
-
-    # Dotted URN path, e.g. "urn:...:user:department"
-    suffix = path[len(_ENTERPRISE_URN_LOWER) :].lstrip(":").lower()
-    if suffix == "department":
-        ent_data["department"] = str(value) if value is not None else None
-    elif suffix == "manager":
-        d = _to_dict(value)
-        if d is not None:
-            ent_data["manager"] = d.get("value")
-        elif isinstance(value, str):
-            ent_data["manager"] = value
-    else:
-        # Unknown enterprise attributes are silently ignored rather than
-        # rejected — IdPs may send attributes we don't model yet.
-        logger.warning("Ignoring unknown enterprise extension attribute '%s'", suffix)
-
-
-# ---------------------------------------------------------------------------
-# Group PATCH
-# ---------------------------------------------------------------------------
-
-
 def apply_group_patch(
     operations: list[ScimPatchOperation],
     current: ScimGroupResource,
@@ -402,14 +235,12 @@ def _set_group_field(
     """Set a single field on group data by SCIM path."""
     if path in ignored_paths:
         return
-
-    entry = _GROUP_REPLACE_PATHS.get(path)
-    if entry:
-        key, _ = entry
-        data[key] = value
-        return
-
-    raise ScimPatchError(f"Unsupported path '{path}' for Group PATCH")
+    elif path == "displayname":
+        data["displayName"] = value
+    elif path == "externalid":
+        data["externalId"] = value
+    else:
+        raise ScimPatchError(f"Unsupported path '{path}' for Group PATCH")
 
 
 def _apply_group_add(

backend/ee/onyx/server/scim/providers/base.py

@@ -2,22 +2,13 @@
 
 from __future__ import annotations
 
-import json
-import logging
 from abc import ABC
 from abc import abstractmethod
 from uuid import UUID
 
-from pydantic import ValidationError
-
-from ee.onyx.server.scim.models import SCIM_ENTERPRISE_USER_SCHEMA
-from ee.onyx.server.scim.models import SCIM_USER_SCHEMA
 from ee.onyx.server.scim.models import ScimEmail
-from ee.onyx.server.scim.models import ScimEnterpriseExtension
 from ee.onyx.server.scim.models import ScimGroupMember
 from ee.onyx.server.scim.models import ScimGroupResource
-from ee.onyx.server.scim.models import ScimManagerRef
-from ee.onyx.server.scim.models import ScimMappingFields
 from ee.onyx.server.scim.models import ScimMeta
 from ee.onyx.server.scim.models import ScimName
 from ee.onyx.server.scim.models import ScimUserGroupRef
@@ -26,17 +17,6 @@ from onyx.db.models import User
 from onyx.db.models import UserGroup
 
 
-logger = logging.getLogger(__name__)
-
-COMMON_IGNORED_PATCH_PATHS: frozenset[str] = frozenset(
-    {
-        "id",
-        "schemas",
-        "meta",
-    }
-)
-
-
 class ScimProvider(ABC):
     """Base class for provider-specific SCIM behavior.
 
@@ -61,22 +41,12 @@ class ScimProvider(ABC):
         """
         ...
 
-    @property
-    def user_schemas(self) -> list[str]:
-        """Schema URIs to include in User resource responses.
-
-        Override in subclasses to advertise additional schemas (e.g. the
-        enterprise extension for Entra ID).
-        """
-        return [SCIM_USER_SCHEMA]
-
     def build_user_resource(
         self,
         user: User,
         external_id: str | None = None,
         groups: list[tuple[int, str]] | None = None,
         scim_username: str | None = None,
-        fields: ScimMappingFields | None = None,
     ) -> ScimUserResource:
         """Build a SCIM User response from an Onyx User.
 
@@ -88,48 +58,27 @@ class ScimProvider(ABC):
                 for newly-created users.
             scim_username: The original-case userName from the IdP. Falls
                 back to ``user.email`` (lowercase) when not available.
-            fields: Stored mapping fields that the IdP expects round-tripped.
         """
-        f = fields or ScimMappingFields()
         group_refs = [
             ScimUserGroupRef(value=str(gid), display=gname)
             for gid, gname in (groups or [])
         ]
 
+        # Use original-case userName if stored, otherwise fall back to the
+        # lowercased email from the User model.
         username = scim_username or user.email
 
-        # Build enterprise extension when at least one value is present.
-        # Dynamically add the enterprise URN to schemas per RFC 7643 §3.0.
-        enterprise_ext: ScimEnterpriseExtension | None = None
-        schemas = list(self.user_schemas)
-        if f.department is not None or f.manager is not None:
-            manager_ref = (
-                ScimManagerRef(value=f.manager) if f.manager is not None else None
-            )
-            enterprise_ext = ScimEnterpriseExtension(
-                department=f.department,
-                manager=manager_ref,
-            )
-            if SCIM_ENTERPRISE_USER_SCHEMA not in schemas:
-                schemas.append(SCIM_ENTERPRISE_USER_SCHEMA)
-
-        name = self.build_scim_name(user, f)
-        emails = _deserialize_emails(f.scim_emails_json, username)
-
-        resource = ScimUserResource(
-            schemas=schemas,
+        return ScimUserResource(
             id=str(user.id),
             externalId=external_id,
             userName=username,
-            name=name,
+            name=self._build_scim_name(user),
             displayName=user.personal_name,
-            emails=emails,
+            emails=[ScimEmail(value=username, type="work", primary=True)],
             active=user.is_active,
             groups=group_refs,
             meta=ScimMeta(resourceType="User"),
         )
-        resource.enterprise_extension = enterprise_ext
-        return resource
 
     def build_group_resource(
         self,
@@ -149,24 +98,9 @@ class ScimProvider(ABC):
             meta=ScimMeta(resourceType="Group"),
         )
 
-    def build_scim_name(
-        self,
-        user: User,
-        fields: ScimMappingFields,
-    ) -> ScimName | None:
-        """Build SCIM name components for the response.
-
-        Round-trips stored ``given_name``/``family_name`` when available (so
-        the IdP gets back what it sent). Falls back to splitting
-        ``personal_name`` for users provisioned before we stored components.
-        Providers may override for custom behavior.
-        """
-        if fields.given_name is not None or fields.family_name is not None:
-            return ScimName(
-                givenName=fields.given_name,
-                familyName=fields.family_name,
-                formatted=user.personal_name,
-            )
+    @staticmethod
+    def _build_scim_name(user: User) -> ScimName | None:
+        """Extract SCIM name components from a user's personal name."""
         if not user.personal_name:
             return None
         parts = user.personal_name.split(" ", 1)
@@ -177,27 +111,6 @@ class ScimProvider(ABC):
         )
 
 
-def _deserialize_emails(stored_json: str | None, username: str) -> list[ScimEmail]:
-    """Deserialize stored email entries or build a default work email."""
-    if stored_json:
-        try:
-            entries = json.loads(stored_json)
-            if isinstance(entries, list) and entries:
-                return [ScimEmail(**e) for e in entries]
-        except (json.JSONDecodeError, TypeError, ValidationError):
-            logger.warning(
-                "Corrupt scim_emails_json, falling back to default: %s", stored_json
-            )
-    return [ScimEmail(value=username, type="work", primary=True)]
-
-
-def serialize_emails(emails: list[ScimEmail]) -> str | None:
-    """Serialize SCIM email entries to JSON for storage."""
-    if not emails:
-        return None
-    return json.dumps([e.model_dump(exclude_none=True) for e in emails])
-
-
 def get_default_provider() -> ScimProvider:
     """Return the default SCIM provider.
 

backend/ee/onyx/server/scim/providers/entra.py

@@ -1,36 +0,0 @@
-"""Entra ID (Azure AD) SCIM provider."""
-
-from __future__ import annotations
-
-from ee.onyx.server.scim.models import SCIM_ENTERPRISE_USER_SCHEMA
-from ee.onyx.server.scim.models import SCIM_USER_SCHEMA
-from ee.onyx.server.scim.providers.base import COMMON_IGNORED_PATCH_PATHS
-from ee.onyx.server.scim.providers.base import ScimProvider
-
-_ENTRA_IGNORED_PATCH_PATHS = COMMON_IGNORED_PATCH_PATHS
-
-
-class EntraProvider(ScimProvider):
-    """Entra ID (Azure AD) SCIM provider.
-
-    Entra behavioral notes:
-      - Sends capitalized PATCH ops (``"Add"``, ``"Replace"``, ``"Remove"``)
-        — handled by ``ScimPatchOperation.normalize_op`` validator.
-      - Sends the enterprise extension URN as a key in path-less PATCH value
-        dicts — handled by ``_set_enterprise_field`` in ``patch.py`` to
-        store department/manager values.
-      - Expects the enterprise extension schema in ``schemas`` arrays and
-        ``/Schemas`` + ``/ResourceTypes`` discovery endpoints.
-    """
-
-    @property
-    def name(self) -> str:
-        return "entra"
-
-    @property
-    def ignored_patch_paths(self) -> frozenset[str]:
-        return _ENTRA_IGNORED_PATCH_PATHS
-
-    @property
-    def user_schemas(self) -> list[str]:
-        return [SCIM_USER_SCHEMA, SCIM_ENTERPRISE_USER_SCHEMA]

backend/ee/onyx/server/scim/providers/okta.py

@@ -2,7 +2,6 @@
 
 from __future__ import annotations
 
-from ee.onyx.server.scim.providers.base import COMMON_IGNORED_PATCH_PATHS
 from ee.onyx.server.scim.providers.base import ScimProvider
 
 
@@ -23,4 +22,4 @@ class OktaProvider(ScimProvider):
 
     @property
     def ignored_patch_paths(self) -> frozenset[str]:
-        return COMMON_IGNORED_PATCH_PATHS
+        return frozenset({"id", "schemas", "meta"})

backend/ee/onyx/server/scim/schema_definitions.py

@@ -4,7 +4,6 @@ Pre-built at import time — these never change at runtime. Separated from
 api.py to keep the endpoint module focused on request handling.
 """
 
-from ee.onyx.server.scim.models import SCIM_ENTERPRISE_USER_SCHEMA
 from ee.onyx.server.scim.models import SCIM_GROUP_SCHEMA
 from ee.onyx.server.scim.models import SCIM_USER_SCHEMA
 from ee.onyx.server.scim.models import ScimResourceType
@@ -21,9 +20,6 @@ USER_RESOURCE_TYPE = ScimResourceType.model_validate(
         "endpoint": "/scim/v2/Users",
         "description": "SCIM User resource",
         "schema": SCIM_USER_SCHEMA,
-        "schemaExtensions": [
-            {"schema": SCIM_ENTERPRISE_USER_SCHEMA, "required": False}
-        ],
     }
 )
 
@@ -108,31 +104,6 @@ USER_SCHEMA_DEF = ScimSchemaDefinition(
     ],
 )
 
-ENTERPRISE_USER_SCHEMA_DEF = ScimSchemaDefinition(
-    id=SCIM_ENTERPRISE_USER_SCHEMA,
-    name="EnterpriseUser",
-    description="Enterprise User extension (RFC 7643 §4.3)",
-    attributes=[
-        ScimSchemaAttribute(
-            name="department",
-            type="string",
-            description="Department.",
-        ),
-        ScimSchemaAttribute(
-            name="manager",
-            type="complex",
-            description="The user's manager.",
-            subAttributes=[
-                ScimSchemaAttribute(
-                    name="value",
-                    type="string",
-                    description="Manager user ID.",
-                ),
-            ],
-        ),
-    ],
-)
-
 GROUP_SCHEMA_DEF = ScimSchemaDefinition(
     id=SCIM_GROUP_SCHEMA,
     name="Group",

backend/onyx/auth/oauth_token_manager.py

@@ -58,27 +58,16 @@ class OAuthTokenManager:
         if not user_token.token_data:
             raise ValueError("No token data available for refresh")
 
-        if (
-            self.oauth_config.client_id is None
-            or self.oauth_config.client_secret is None
-        ):
-            raise ValueError(
-                "OAuth client_id and client_secret are required for token refresh"
-            )
-
         token_data = self._unwrap_token_data(user_token.token_data)
 
-        data: dict[str, str] = {
-            "grant_type": "refresh_token",
-            "refresh_token": token_data["refresh_token"],
-            "client_id": self._unwrap_sensitive_str(self.oauth_config.client_id),
-            "client_secret": self._unwrap_sensitive_str(
-                self.oauth_config.client_secret
-            ),
-        }
         response = requests.post(
             self.oauth_config.token_url,
-            data=data,
+            data={
+                "grant_type": "refresh_token",
+                "refresh_token": token_data["refresh_token"],
+                "client_id": self.oauth_config.client_id,
+                "client_secret": self.oauth_config.client_secret,
+            },
             headers={"Accept": "application/json"},
         )
         response.raise_for_status()
@@ -126,26 +115,15 @@ class OAuthTokenManager:
 
     def exchange_code_for_token(self, code: str, redirect_uri: str) -> dict[str, Any]:
         """Exchange authorization code for access token"""
-        if (
-            self.oauth_config.client_id is None
-            or self.oauth_config.client_secret is None
-        ):
-            raise ValueError(
-                "OAuth client_id and client_secret are required for code exchange"
-            )
-
-        data: dict[str, str] = {
-            "grant_type": "authorization_code",
-            "code": code,
-            "client_id": self._unwrap_sensitive_str(self.oauth_config.client_id),
-            "client_secret": self._unwrap_sensitive_str(
-                self.oauth_config.client_secret
-            ),
-            "redirect_uri": redirect_uri,
-        }
         response = requests.post(
             self.oauth_config.token_url,
-            data=data,
+            data={
+                "grant_type": "authorization_code",
+                "code": code,
+                "client_id": self.oauth_config.client_id,
+                "client_secret": self.oauth_config.client_secret,
+                "redirect_uri": redirect_uri,
+            },
             headers={"Accept": "application/json"},
         )
         response.raise_for_status()
@@ -163,13 +141,8 @@ class OAuthTokenManager:
         oauth_config: OAuthConfig, redirect_uri: str, state: str
     ) -> str:
         """Build OAuth authorization URL"""
-        if oauth_config.client_id is None:
-            raise ValueError("OAuth client_id is required to build authorization URL")
-
         params: dict[str, Any] = {
-            "client_id": OAuthTokenManager._unwrap_sensitive_str(
-                oauth_config.client_id
-            ),
+            "client_id": oauth_config.client_id,
             "redirect_uri": redirect_uri,
             "response_type": "code",
             "state": state,
@@ -188,12 +161,6 @@ class OAuthTokenManager:
 
         return f"{oauth_config.authorization_url}{separator}{urlencode(params)}"
 
-    @staticmethod
-    def _unwrap_sensitive_str(value: SensitiveValue[str] | str) -> str:
-        if isinstance(value, SensitiveValue):
-            return value.get_value(apply_mask=False)
-        return value
-
     @staticmethod
     def _unwrap_token_data(
         token_data: SensitiveValue[dict[str, Any]] | dict[str, Any],

backend/onyx/background/celery/tasks/opensearch_migration/tasks.py

@@ -48,7 +48,6 @@ from onyx.document_index.opensearch.opensearch_document_index import (
     OpenSearchDocumentIndex,
 )
 from onyx.document_index.vespa.vespa_document_index import VespaDocumentIndex
-from onyx.indexing.models import IndexingSetting
 from onyx.redis.redis_pool import get_redis_client
 from shared_configs.configs import MULTI_TENANT
 from shared_configs.contextvars import get_current_tenant_id
@@ -150,12 +149,8 @@ def migrate_chunks_from_vespa_to_opensearch_task(
             try_insert_opensearch_tenant_migration_record_with_commit(db_session)
             search_settings = get_current_search_settings(db_session)
             tenant_state = TenantState(tenant_id=tenant_id, multitenant=MULTI_TENANT)
-            indexing_setting = IndexingSetting.from_db_model(search_settings)
             opensearch_document_index = OpenSearchDocumentIndex(
-                tenant_state=tenant_state,
-                index_name=search_settings.index_name,
-                embedding_dim=indexing_setting.final_embedding_dim,
-                embedding_precision=indexing_setting.embedding_precision,
+                index_name=search_settings.index_name, tenant_state=tenant_state
             )
             vespa_document_index = VespaDocumentIndex(
                 index_name=search_settings.index_name,

backend/onyx/background/celery/tasks/opensearch_migration/transformer.py

@@ -22,7 +22,6 @@ from onyx.document_index.vespa_constants import HIDDEN
 from onyx.document_index.vespa_constants import IMAGE_FILE_NAME
 from onyx.document_index.vespa_constants import METADATA_LIST
 from onyx.document_index.vespa_constants import METADATA_SUFFIX
-from onyx.document_index.vespa_constants import PERSONAS
 from onyx.document_index.vespa_constants import PRIMARY_OWNERS
 from onyx.document_index.vespa_constants import SECONDARY_OWNERS
 from onyx.document_index.vespa_constants import SEMANTIC_IDENTIFIER
@@ -59,7 +58,6 @@ FIELDS_NEEDED_FOR_TRANSFORMATION: list[str] = [
     METADATA_SUFFIX,
     DOCUMENT_SETS,
     USER_PROJECT,
-    PERSONAS,
     PRIMARY_OWNERS,
     SECONDARY_OWNERS,
     ACCESS_CONTROL_LIST,
@@ -278,7 +276,6 @@ def transform_vespa_chunks_to_opensearch_chunks(
                 )
             )
             user_projects: list[int] | None = vespa_chunk.get(USER_PROJECT)
-            personas: list[int] | None = vespa_chunk.get(PERSONAS)
             primary_owners: list[str] | None = vespa_chunk.get(PRIMARY_OWNERS)
             secondary_owners: list[str] | None = vespa_chunk.get(SECONDARY_OWNERS)
 
@@ -328,7 +325,6 @@ def transform_vespa_chunks_to_opensearch_chunks(
                 metadata_suffix=metadata_suffix,
                 document_sets=document_sets,
                 user_projects=user_projects,
-                personas=personas,
                 primary_owners=primary_owners,
                 secondary_owners=secondary_owners,
                 tenant_id=tenant_state,

backend/onyx/background/celery/tasks/user_file_processing/tasks.py

@@ -5,14 +5,11 @@ from uuid import UUID
 
 import httpx
 import sqlalchemy as sa
-from celery import Celery
 from celery import shared_task
 from celery import Task
-from redis import Redis
 from redis.lock import Lock as RedisLock
 from retry import retry
 from sqlalchemy import select
-from sqlalchemy.orm import selectinload
 from sqlalchemy.orm import Session
 
 from onyx.background.celery.apps.app_base import task_logger
@@ -27,14 +24,12 @@ from onyx.configs.constants import CELERY_GENERIC_BEAT_LOCK_TIMEOUT
 from onyx.configs.constants import CELERY_USER_FILE_PROCESSING_LOCK_TIMEOUT
 from onyx.configs.constants import CELERY_USER_FILE_PROCESSING_TASK_EXPIRES
 from onyx.configs.constants import CELERY_USER_FILE_PROJECT_SYNC_LOCK_TIMEOUT
-from onyx.configs.constants import CELERY_USER_FILE_PROJECT_SYNC_TASK_EXPIRES
 from onyx.configs.constants import DocumentSource
 from onyx.configs.constants import OnyxCeleryPriority
 from onyx.configs.constants import OnyxCeleryQueues
 from onyx.configs.constants import OnyxCeleryTask
 from onyx.configs.constants import OnyxRedisLocks
 from onyx.configs.constants import USER_FILE_PROCESSING_MAX_QUEUE_DEPTH
-from onyx.configs.constants import USER_FILE_PROJECT_SYNC_MAX_QUEUE_DEPTH
 from onyx.connectors.file.connector import LocalFileConnector
 from onyx.connectors.models import Document
 from onyx.connectors.models import HierarchyNode
@@ -80,58 +75,10 @@ def _user_file_project_sync_lock_key(user_file_id: str | UUID) -> str:
     return f"{OnyxRedisLocks.USER_FILE_PROJECT_SYNC_LOCK_PREFIX}:{user_file_id}"
 
 
-def _user_file_project_sync_queued_key(user_file_id: str | UUID) -> str:
-    return f"{OnyxRedisLocks.USER_FILE_PROJECT_SYNC_QUEUED_PREFIX}:{user_file_id}"
-
-
 def _user_file_delete_lock_key(user_file_id: str | UUID) -> str:
     return f"{OnyxRedisLocks.USER_FILE_DELETE_LOCK_PREFIX}:{user_file_id}"
 
 
-def get_user_file_project_sync_queue_depth(celery_app: Celery) -> int:
-    redis_celery: Redis = celery_app.broker_connection().channel().client  # type: ignore
-    return celery_get_queue_length(
-        OnyxCeleryQueues.USER_FILE_PROJECT_SYNC, redis_celery
-    )
-
-
-def enqueue_user_file_project_sync_task(
-    *,
-    celery_app: Celery,
-    redis_client: Redis,
-    user_file_id: str | UUID,
-    tenant_id: str,
-    priority: OnyxCeleryPriority = OnyxCeleryPriority.HIGH,
-) -> bool:
-    """Enqueue a project-sync task if no matching queued task already exists."""
-    queued_key = _user_file_project_sync_queued_key(user_file_id)
-
-    # NX+EX gives us atomic dedupe and a self-healing TTL.
-    queued_guard_set = redis_client.set(
-        queued_key,
-        1,
-        nx=True,
-        ex=CELERY_USER_FILE_PROJECT_SYNC_TASK_EXPIRES,
-    )
-    if not queued_guard_set:
-        return False
-
-    try:
-        celery_app.send_task(
-            OnyxCeleryTask.PROCESS_SINGLE_USER_FILE_PROJECT_SYNC,
-            kwargs={"user_file_id": str(user_file_id), "tenant_id": tenant_id},
-            queue=OnyxCeleryQueues.USER_FILE_PROJECT_SYNC,
-            priority=priority,
-            expires=CELERY_USER_FILE_PROJECT_SYNC_TASK_EXPIRES,
-        )
-    except Exception:
-        # Roll back the queued guard if task publish fails.
-        redis_client.delete(queued_key)
-        raise
-
-    return True
-
-
 @retry(tries=3, delay=1, backoff=2, jitter=(0.0, 1.0))
 def _visit_chunks(
     *,
@@ -685,8 +632,8 @@ def process_single_user_file_delete(
     ignore_result=True,
 )
 def check_for_user_file_project_sync(self: Task, *, tenant_id: str) -> None:
-    """Scan for user files needing project sync and enqueue per-file tasks."""
-    task_logger.info("Starting")
+    """Scan for user files with PROJECT_SYNC status and enqueue per-file tasks."""
+    task_logger.info("check_for_user_file_project_sync - Starting")
 
     redis_client = get_redis_client(tenant_id=tenant_id)
     lock: RedisLock = redis_client.lock(
@@ -698,25 +645,13 @@ def check_for_user_file_project_sync(self: Task, *, tenant_id: str) -> None:
         return None
 
     enqueued = 0
-    skipped_guard = 0
     try:
-        queue_depth = get_user_file_project_sync_queue_depth(self.app)
-        if queue_depth > USER_FILE_PROJECT_SYNC_MAX_QUEUE_DEPTH:
-            task_logger.warning(
-                f"Queue depth {queue_depth} exceeds "
-                f"{USER_FILE_PROJECT_SYNC_MAX_QUEUE_DEPTH}, skipping enqueue for tenant={tenant_id}"
-            )
-            return None
-
         with get_session_with_current_tenant() as db_session:
             user_file_ids = (
                 db_session.execute(
                     select(UserFile.id).where(
                         sa.and_(
-                            sa.or_(
-                                UserFile.needs_project_sync.is_(True),
-                                UserFile.needs_persona_sync.is_(True),
-                            ),
+                            UserFile.needs_project_sync.is_(True),
                             UserFile.status == UserFileStatus.COMPLETED,
                         )
                     )
@@ -726,23 +661,19 @@ def check_for_user_file_project_sync(self: Task, *, tenant_id: str) -> None:
             )
 
             for user_file_id in user_file_ids:
-                if not enqueue_user_file_project_sync_task(
-                    celery_app=self.app,
-                    redis_client=redis_client,
-                    user_file_id=user_file_id,
-                    tenant_id=tenant_id,
+                self.app.send_task(
+                    OnyxCeleryTask.PROCESS_SINGLE_USER_FILE_PROJECT_SYNC,
+                    kwargs={"user_file_id": str(user_file_id), "tenant_id": tenant_id},
+                    queue=OnyxCeleryQueues.USER_FILE_PROJECT_SYNC,
                     priority=OnyxCeleryPriority.HIGH,
-                ):
-                    skipped_guard += 1
-                    continue
+                )
                 enqueued += 1
     finally:
         if lock.owned():
             lock.release()
 
     task_logger.info(
-        f"Enqueued {enqueued} "
-        f"Skipped guard {skipped_guard} tasks for tenant={tenant_id}"
+        f"check_for_user_file_project_sync - Enqueued {enqueued} tasks for tenant={tenant_id}"
     )
     return None
 
@@ -761,8 +692,6 @@ def process_single_user_file_project_sync(
     )
 
     redis_client = get_redis_client(tenant_id=tenant_id)
-    redis_client.delete(_user_file_project_sync_queued_key(user_file_id))
-
     file_lock: RedisLock = redis_client.lock(
         _user_file_project_sync_lock_key(user_file_id),
         timeout=CELERY_USER_FILE_PROJECT_SYNC_LOCK_TIMEOUT,
@@ -776,11 +705,7 @@ def process_single_user_file_project_sync(
 
     try:
         with get_session_with_current_tenant() as db_session:
-            user_file = db_session.execute(
-                select(UserFile)
-                .where(UserFile.id == _as_uuid(user_file_id))
-                .options(selectinload(UserFile.assistants))
-            ).scalar_one_or_none()
+            user_file = db_session.get(UserFile, _as_uuid(user_file_id))
             if not user_file:
                 task_logger.info(
                     f"process_single_user_file_project_sync - User file not found id={user_file_id}"
@@ -808,17 +733,13 @@ def process_single_user_file_project_sync(
                 ]
 
                 project_ids = [project.id for project in user_file.projects]
-                persona_ids = [p.id for p in user_file.assistants if not p.deleted]
                 for retry_document_index in retry_document_indices:
                     retry_document_index.update_single(
                         doc_id=str(user_file.id),
                         tenant_id=tenant_id,
                         chunk_count=user_file.chunk_count,
                         fields=None,
-                        user_fields=VespaDocumentUserFields(
-                            user_projects=project_ids,
-                            personas=persona_ids,
-                        ),
+                        user_fields=VespaDocumentUserFields(user_projects=project_ids),
                     )
 
             task_logger.info(
@@ -826,7 +747,6 @@ def process_single_user_file_project_sync(
             )
 
             user_file.needs_project_sync = False
-            user_file.needs_persona_sync = False
             user_file.last_project_sync_at = datetime.datetime.now(
                 datetime.timezone.utc
             )

backend/onyx/background/indexing/run_docfetching.py

@@ -58,8 +58,6 @@ from onyx.file_store.document_batch_storage import DocumentBatchStorage
 from onyx.file_store.document_batch_storage import get_document_batch_storage
 from onyx.indexing.indexing_heartbeat import IndexingHeartbeatInterface
 from onyx.indexing.indexing_pipeline import index_doc_batch_prepare
-from onyx.indexing.postgres_sanitization import sanitize_document_for_postgres
-from onyx.indexing.postgres_sanitization import sanitize_hierarchy_nodes_for_postgres
 from onyx.redis.redis_hierarchy import cache_hierarchy_nodes_batch
 from onyx.redis.redis_hierarchy import ensure_source_node_exists
 from onyx.redis.redis_hierarchy import get_node_id_from_raw_id
@@ -158,7 +156,36 @@ def strip_null_characters(doc_batch: list[Document]) -> list[Document]:
             logger.warning(
                 f"doc {doc.id} too large, Document size: {sys.getsizeof(doc)}"
             )
-        cleaned_batch.append(sanitize_document_for_postgres(doc))
+        cleaned_doc = doc.model_copy()
+
+        # Postgres cannot handle NUL characters in text fields
+        if "\x00" in cleaned_doc.id:
+            logger.warning(f"NUL characters found in document ID: {cleaned_doc.id}")
+            cleaned_doc.id = cleaned_doc.id.replace("\x00", "")
+
+        if cleaned_doc.title and "\x00" in cleaned_doc.title:
+            logger.warning(
+                f"NUL characters found in document title: {cleaned_doc.title}"
+            )
+            cleaned_doc.title = cleaned_doc.title.replace("\x00", "")
+
+        if "\x00" in cleaned_doc.semantic_identifier:
+            logger.warning(
+                f"NUL characters found in document semantic identifier: {cleaned_doc.semantic_identifier}"
+            )
+            cleaned_doc.semantic_identifier = cleaned_doc.semantic_identifier.replace(
+                "\x00", ""
+            )
+
+        for section in cleaned_doc.sections:
+            if section.link is not None:
+                section.link = section.link.replace("\x00", "")
+
+            # since text can be longer, just replace to avoid double scan
+            if isinstance(section, TextSection) and section.text is not None:
+                section.text = section.text.replace("\x00", "")
+
+        cleaned_batch.append(cleaned_doc)
 
     return cleaned_batch
 
@@ -575,13 +602,10 @@ def connector_document_extraction(
 
                 # Process hierarchy nodes batch - upsert to Postgres and cache in Redis
                 if hierarchy_node_batch:
-                    hierarchy_node_batch_cleaned = (
-                        sanitize_hierarchy_nodes_for_postgres(hierarchy_node_batch)
-                    )
                     with get_session_with_current_tenant() as db_session:
                         upserted_nodes = upsert_hierarchy_nodes_batch(
                             db_session=db_session,
-                            nodes=hierarchy_node_batch_cleaned,
+                            nodes=hierarchy_node_batch,
                             source=db_connector.source,
                             commit=True,
                             is_connector_public=is_connector_public,
@@ -600,7 +624,7 @@ def connector_document_extraction(
                         )
 
                     logger.debug(
-                        f"Persisted and cached {len(hierarchy_node_batch_cleaned)} hierarchy nodes "
+                        f"Persisted and cached {len(hierarchy_node_batch)} hierarchy nodes "
                         f"for attempt={index_attempt_id}"
                     )
 

backend/onyx/chat/llm_loop.py

@@ -30,7 +30,6 @@ from onyx.configs.constants import DocumentSource
 from onyx.configs.constants import MessageType
 from onyx.context.search.models import SearchDoc
 from onyx.context.search.models import SearchDocsResponse
-from onyx.db.engine.sql_engine import get_session_with_current_tenant
 from onyx.db.memory import add_memory
 from onyx.db.memory import update_memory_at_index
 from onyx.db.memory import UserMemoryContext
@@ -657,12 +656,7 @@ def run_llm_loop(
         fallback_extraction_attempted: bool = False
         citation_mapping: dict[int, str] = {}  # Maps citation_num -> document_id/URL
 
-        # Fetch this in a short-lived session so the long-running stream loop does
-        # not pin a connection just to keep read state alive.
-        with get_session_with_current_tenant() as prompt_db_session:
-            default_base_system_prompt: str = get_default_base_system_prompt(
-                prompt_db_session
-            )
+        default_base_system_prompt: str = get_default_base_system_prompt(db_session)
         system_prompt = None
         custom_agent_prompt_msg = None
 

backend/onyx/chat/process_message.py

@@ -856,11 +856,6 @@ def handle_stream_message_objects(
                 reserved_tokens=reserved_token_count,
             )
 
-        # Release any read transaction before entering the long-running LLM stream.
-        # Without this, the request-scoped session can keep a connection checked out
-        # for the full stream duration.
-        db_session.commit()
-
         # The stream generator can resume on a different worker thread after early yields.
         # Set this right before launching the LLM loop so run_in_background copies the right context.
         if new_msg_req.mock_llm_response is not None:

backend/onyx/configs/app_configs.py

@@ -210,10 +210,10 @@ AUTH_COOKIE_EXPIRE_TIME_SECONDS = int(
 REQUIRE_EMAIL_VERIFICATION = (
     os.environ.get("REQUIRE_EMAIL_VERIFICATION", "").lower() == "true"
 )
-SMTP_SERVER = os.environ.get("SMTP_SERVER") or ""
+SMTP_SERVER = os.environ.get("SMTP_SERVER") or "smtp.gmail.com"
 SMTP_PORT = int(os.environ.get("SMTP_PORT") or "587")
-SMTP_USER = os.environ.get("SMTP_USER") or ""
-SMTP_PASS = os.environ.get("SMTP_PASS") or ""
+SMTP_USER = os.environ.get("SMTP_USER", "your-email@gmail.com")
+SMTP_PASS = os.environ.get("SMTP_PASS", "your-gmail-password")
 EMAIL_FROM = os.environ.get("EMAIL_FROM") or SMTP_USER
 
 SENDGRID_API_KEY = os.environ.get("SENDGRID_API_KEY") or ""
@@ -294,12 +294,6 @@ ENABLE_OPENSEARCH_RETRIEVAL_FOR_ONYX = (
     ENABLE_OPENSEARCH_INDEXING_FOR_ONYX
     and os.environ.get("ENABLE_OPENSEARCH_RETRIEVAL_FOR_ONYX", "").lower() == "true"
 )
-# Whether we should check for and create an index if necessary every time we
-# instantiate an OpenSearchDocumentIndex on multitenant cloud. Defaults to True.
-VERIFY_CREATE_OPENSEARCH_INDEX_ON_INIT_MT = (
-    os.environ.get("VERIFY_CREATE_OPENSEARCH_INDEX_ON_INIT_MT", "true").lower()
-    == "true"
-)
 
 VESPA_HOST = os.environ.get("VESPA_HOST") or "localhost"
 # NOTE: this is used if and only if the vespa config server is accessible via a

backend/onyx/configs/constants.py

@@ -167,14 +167,6 @@ CELERY_USER_FILE_PROCESSING_TASK_EXPIRES = 60  # 1 minute (in seconds)
 # beat generator stops adding more.  Prevents unbounded queue growth when workers
 # fall behind.
 USER_FILE_PROCESSING_MAX_QUEUE_DEPTH = 500
-# How long a queued user-file-project-sync task remains valid.
-# Should be short enough to discard stale queue entries under load while still
-# allowing workers enough time to pick up new tasks.
-CELERY_USER_FILE_PROJECT_SYNC_TASK_EXPIRES = 60  # 1 minute (in seconds)
-
-# Max queue depth before user-file-project-sync producers stop enqueuing.
-# This applies backpressure when workers are falling behind.
-USER_FILE_PROJECT_SYNC_MAX_QUEUE_DEPTH = 500
 
 CELERY_USER_FILE_PROJECT_SYNC_LOCK_TIMEOUT = 5 * 60  # 5 minutes (in seconds)
 
@@ -467,7 +459,6 @@ class OnyxRedisLocks:
     USER_FILE_QUEUED_PREFIX = "da_lock:user_file_queued"
     USER_FILE_PROJECT_SYNC_BEAT_LOCK = "da_lock:check_user_file_project_sync_beat"
     USER_FILE_PROJECT_SYNC_LOCK_PREFIX = "da_lock:user_file_project_sync"
-    USER_FILE_PROJECT_SYNC_QUEUED_PREFIX = "da_lock:user_file_project_sync_queued"
     USER_FILE_DELETE_BEAT_LOCK = "da_lock:check_user_file_delete_beat"
     USER_FILE_DELETE_LOCK_PREFIX = "da_lock:user_file_delete"
 

backend/onyx/connectors/google_utils/google_utils.py

@@ -16,22 +16,6 @@ from onyx.utils.retry_wrapper import retry_builder
 
 logger = setup_logger()
 
-_RATE_LIMIT_REASONS = {"userRateLimitExceeded", "rateLimitExceeded"}
-
-
-def _is_rate_limit_error(error: HttpError) -> bool:
-    """Google sometimes returns rate-limit errors as 403 with reason
-    'userRateLimitExceeded' instead of 429. This helper detects both."""
-    if error.resp.status == 429:
-        return True
-    if error.resp.status != 403:
-        return False
-    error_details = getattr(error, "error_details", None) or []
-    for detail in error_details:
-        if isinstance(detail, dict) and detail.get("reason") in _RATE_LIMIT_REASONS:
-            return True
-    return "userRateLimitExceeded" in str(error) or "rateLimitExceeded" in str(error)
-
 
 # Google Drive APIs are quite flakey and may 500 for an
 # extended period of time. This is now addressed by checkpointing.
@@ -73,7 +57,7 @@ def _execute_with_retry(request: Any) -> Any:
         except HttpError as error:
             attempt += 1
 
-            if _is_rate_limit_error(error):
+            if error.resp.status == 429:
                 # Attempt to get 'Retry-After' from headers
                 retry_after = error.resp.get("Retry-After")
                 if retry_after:
@@ -156,16 +140,16 @@ def _execute_single_retrieval(
                 )
             logger.error(f"Error executing request: {e}")
             raise e
-        elif _is_rate_limit_error(e):
-            results = _execute_with_retry(
-                lambda: retrieval_function(**request_kwargs).execute()
-            )
         elif e.resp.status == 404 or e.resp.status == 403:
             if continue_on_404_or_403:
                 logger.debug(f"Error executing request: {e}")
                 results = {}
             else:
                 raise e
+        elif e.resp.status == 429:
+            results = _execute_with_retry(
+                lambda: retrieval_function(**request_kwargs).execute()
+            )
         else:
             logger.exception("Error executing request:")
             raise e

backend/onyx/connectors/microsoft_graph_env.py

@@ -1,96 +0,0 @@
-"""Inverse mapping from user-facing Microsoft host URLs to the SDK's AzureEnvironment.
-
-The office365 library's GraphClient requires an ``AzureEnvironment`` string
-(e.g. ``"Global"``, ``"GCC High"``) to route requests to the correct national
-cloud.  Our connectors instead expose free-text ``authority_host`` and
-``graph_api_host`` fields so the frontend doesn't need to know about SDK
-internals.
-
-This module bridges the gap: given the two host URLs the user configured, it
-resolves the matching ``AzureEnvironment`` value (and the implied SharePoint
-domain suffix) so callers can pass ``environment=…`` to ``GraphClient``.
-"""
-
-from office365.graph_client import AzureEnvironment  # type: ignore[import-untyped]
-from pydantic import BaseModel
-
-from onyx.connectors.exceptions import ConnectorValidationError
-
-
-class MicrosoftGraphEnvironment(BaseModel):
-    """One row of the inverse mapping."""
-
-    environment: str
-    graph_host: str
-    authority_host: str
-    sharepoint_domain_suffix: str
-
-
-_ENVIRONMENTS: list[MicrosoftGraphEnvironment] = [
-    MicrosoftGraphEnvironment(
-        environment=AzureEnvironment.Global,
-        graph_host="https://graph.microsoft.com",
-        authority_host="https://login.microsoftonline.com",
-        sharepoint_domain_suffix="sharepoint.com",
-    ),
-    MicrosoftGraphEnvironment(
-        environment=AzureEnvironment.USGovernmentHigh,
-        graph_host="https://graph.microsoft.us",
-        authority_host="https://login.microsoftonline.us",
-        sharepoint_domain_suffix="sharepoint.us",
-    ),
-    MicrosoftGraphEnvironment(
-        environment=AzureEnvironment.USGovernmentDoD,
-        graph_host="https://dod-graph.microsoft.us",
-        authority_host="https://login.microsoftonline.us",
-        sharepoint_domain_suffix="sharepoint.us",
-    ),
-    MicrosoftGraphEnvironment(
-        environment=AzureEnvironment.China,
-        graph_host="https://microsoftgraph.chinacloudapi.cn",
-        authority_host="https://login.chinacloudapi.cn",
-        sharepoint_domain_suffix="sharepoint.cn",
-    ),
-    MicrosoftGraphEnvironment(
-        environment=AzureEnvironment.Germany,
-        graph_host="https://graph.microsoft.de",
-        authority_host="https://login.microsoftonline.de",
-        sharepoint_domain_suffix="sharepoint.de",
-    ),
-]
-
-_GRAPH_HOST_INDEX: dict[str, MicrosoftGraphEnvironment] = {
-    env.graph_host: env for env in _ENVIRONMENTS
-}
-
-
-def resolve_microsoft_environment(
-    graph_api_host: str,
-    authority_host: str,
-) -> MicrosoftGraphEnvironment:
-    """Return the ``MicrosoftGraphEnvironment`` that matches the supplied hosts.
-
-    Raises ``ConnectorValidationError`` when the combination is unknown or
-    internally inconsistent (e.g. a GCC-High graph host paired with a
-    commercial authority host).
-    """
-    graph_api_host = graph_api_host.rstrip("/")
-    authority_host = authority_host.rstrip("/")
-
-    env = _GRAPH_HOST_INDEX.get(graph_api_host)
-    if env is None:
-        known = ", ".join(sorted(_GRAPH_HOST_INDEX))
-        raise ConnectorValidationError(
-            f"Unsupported Microsoft Graph API host '{graph_api_host}'. "
-            f"Recognised hosts: {known}"
-        )
-
-    if env.authority_host != authority_host:
-        raise ConnectorValidationError(
-            f"Authority host '{authority_host}' is inconsistent with "
-            f"graph API host '{graph_api_host}'. "
-            f"Expected authority host '{env.authority_host}' "
-            f"for the {env.environment} environment."
-        )
-
-    return env

backend/onyx/connectors/models.py

@@ -6,7 +6,6 @@ from typing import cast
 
 from pydantic import BaseModel
 from pydantic import Field
-from pydantic import field_validator
 from pydantic import model_validator
 
 from onyx.access.models import ExternalAccess
@@ -168,14 +167,6 @@ class DocumentBase(BaseModel):
     # list of strings.
     metadata: dict[str, str | list[str]]
 
-    @field_validator("metadata", mode="before")
-    @classmethod
-    def _coerce_metadata_values(cls, v: dict[str, Any]) -> dict[str, str | list[str]]:
-        return {
-            key: [str(item) for item in val] if isinstance(val, list) else str(val)
-            for key, val in v.items()
-        }
-
     # UTC time
     doc_updated_at: datetime | None = None
     chunk_count: int | None = None

backend/onyx/connectors/sharepoint/connector.py

@@ -47,7 +47,6 @@ from onyx.connectors.interfaces import GenerateSlimDocumentOutput
 from onyx.connectors.interfaces import IndexingHeartbeatInterface
 from onyx.connectors.interfaces import SecondsSinceUnixEpoch
 from onyx.connectors.interfaces import SlimConnectorWithPermSync
-from onyx.connectors.microsoft_graph_env import resolve_microsoft_environment
 from onyx.connectors.models import BasicExpertInfo
 from onyx.connectors.models import ConnectorCheckpoint
 from onyx.connectors.models import ConnectorFailure
@@ -147,9 +146,7 @@ class DriveItemData(BaseModel):
             self.id,
             ResourcePath("items", ResourcePath(self.drive_id, ResourcePath("drives"))),
         )
-        item = DriveItem(graph_client, path)
-        item.set_property("id", self.id)
-        return item
+        return DriveItem(graph_client, path)
 
 
 # The office365 library's ClientContext caches the access token from its
@@ -840,20 +837,10 @@ class SharepointConnector(
         self._cached_rest_ctx: ClientContext | None = None
         self._cached_rest_ctx_url: str | None = None
         self._cached_rest_ctx_created_at: float = 0.0
-
-        resolved_env = resolve_microsoft_environment(graph_api_host, authority_host)
-        self._azure_environment = resolved_env.environment
-        self.authority_host = resolved_env.authority_host
-        self.graph_api_host = resolved_env.graph_host
+        self.authority_host = authority_host.rstrip("/")
+        self.graph_api_host = graph_api_host.rstrip("/")
         self.graph_api_base = f"{self.graph_api_host}/v1.0"
-        self.sharepoint_domain_suffix = resolved_env.sharepoint_domain_suffix
-        if sharepoint_domain_suffix != resolved_env.sharepoint_domain_suffix:
-            logger.warning(
-                f"Configured sharepoint_domain_suffix '{sharepoint_domain_suffix}' "
-                f"differs from the expected suffix '{resolved_env.sharepoint_domain_suffix}' "
-                f"for the {resolved_env.environment} environment. "
-                f"Using '{resolved_env.sharepoint_domain_suffix}'."
-            )
+        self.sharepoint_domain_suffix = sharepoint_domain_suffix
 
     def validate_connector_settings(self) -> None:
         # Validate that at least one content type is enabled
@@ -1605,7 +1592,6 @@ class SharepointConnector(
             if certificate_data is None:
                 raise RuntimeError("Failed to load certificate")
 
-            logger.info(f"Creating MSAL app with authority url {authority_url}")
             self.msal_app = msal.ConfidentialClientApplication(
                 authority=authority_url,
                 client_id=sp_client_id,
@@ -1637,9 +1623,7 @@ class SharepointConnector(
                 raise ConnectorValidationError("Failed to acquire token for graph")
             return token
 
-        self._graph_client = GraphClient(
-            _acquire_token_for_graph, environment=self._azure_environment
-        )
+        self._graph_client = GraphClient(_acquire_token_for_graph)
         if auth_method == SharepointAuthMethod.CERTIFICATE.value:
             org = self.graph_client.organization.get().execute_query()
             if not org or len(org) == 0:

backend/onyx/connectors/slab/connector.py

@@ -11,7 +11,6 @@ from dateutil import parser
 
 from onyx.configs.app_configs import INDEX_BATCH_SIZE
 from onyx.configs.constants import DocumentSource
-from onyx.connectors.exceptions import ConnectorValidationError
 from onyx.connectors.interfaces import GenerateDocumentsOutput
 from onyx.connectors.interfaces import GenerateSlimDocumentOutput
 from onyx.connectors.interfaces import LoadConnector
@@ -259,21 +258,3 @@ class SlabConnector(LoadConnector, PollConnector, SlimConnectorWithPermSync):
                 slim_doc_batch = []
         if slim_doc_batch:
             yield slim_doc_batch
-
-    def validate_connector_settings(self) -> None:
-        """
-        Very basic validation, we could do more here
-        """
-        if not self.base_url.startswith("https://") and not self.base_url.startswith(
-            "http://"
-        ):
-            raise ConnectorValidationError(
-                "Base URL must start with https:// or http://"
-            )
-
-        try:
-            get_all_post_ids(self.slab_bot_token)
-        except ConnectorMissingCredentialError:
-            raise
-        except Exception as e:
-            raise ConnectorValidationError(f"Failed to fetch posts from Slab: {e}")

backend/onyx/connectors/teams/connector.py

@@ -23,7 +23,6 @@ from onyx.connectors.interfaces import CheckpointOutput
 from onyx.connectors.interfaces import GenerateSlimDocumentOutput
 from onyx.connectors.interfaces import SecondsSinceUnixEpoch
 from onyx.connectors.interfaces import SlimConnectorWithPermSync
-from onyx.connectors.microsoft_graph_env import resolve_microsoft_environment
 from onyx.connectors.models import ConnectorCheckpoint
 from onyx.connectors.models import ConnectorFailure
 from onyx.connectors.models import ConnectorMissingCredentialError
@@ -74,11 +73,8 @@ class TeamsConnector(
         self.msal_app: msal.ConfidentialClientApplication | None = None
         self.max_workers = max_workers
         self.requested_team_list: list[str] = teams
-
-        resolved_env = resolve_microsoft_environment(graph_api_host, authority_host)
-        self._azure_environment = resolved_env.environment
-        self.authority_host = resolved_env.authority_host
-        self.graph_api_host = resolved_env.graph_host
+        self.authority_host = authority_host.rstrip("/")
+        self.graph_api_host = graph_api_host.rstrip("/")
 
     # impls for BaseConnector
 
@@ -110,9 +106,7 @@ class TeamsConnector(
 
             return token
 
-        self.graph_client = GraphClient(
-            _acquire_token_func, environment=self._azure_environment
-        )
+        self.graph_client = GraphClient(_acquire_token_func)
         return None
 
     def validate_connector_settings(self) -> None:

backend/onyx/context/search/models.py

@@ -72,7 +72,6 @@ class BaseFilters(BaseModel):
 class UserFileFilters(BaseModel):
     user_file_ids: list[UUID] | None = None
     project_id: int | None = None
-    persona_id: int | None = None
 
 
 class AssistantKnowledgeFilters(BaseModel):

backend/onyx/context/search/pipeline.py

@@ -40,7 +40,6 @@ def _build_index_filters(
     user_provided_filters: BaseFilters | None,
     user: User,  # Used for ACLs, anonymous users only see public docs
     project_id: int | None,
-    persona_id: int | None,
     user_file_ids: list[UUID] | None,
     persona_document_sets: list[str] | None,
     persona_time_cutoff: datetime | None,
@@ -119,7 +118,6 @@ def _build_index_filters(
     final_filters = IndexFilters(
         user_file_ids=user_file_ids,
         project_id=project_id,
-        persona_id=persona_id,
         source_type=source_filter,
         document_set=document_set_filter,
         time_cutoff=time_filter,
@@ -267,8 +265,6 @@ def search_pipeline(
     llm: LLM | None = None,
     # If a project ID is provided, it will be exclusively scoped to that project
     project_id: int | None = None,
-    # If a persona_id is provided, search scopes to files attached to this persona
-    persona_id: int | None = None,
     # Pre-fetched data — when provided, avoids DB queries (no session needed)
     acl_filters: list[str] | None = None,
     embedding_model: EmbeddingModel | None = None,
@@ -303,7 +299,6 @@ def search_pipeline(
         user_provided_filters=chunk_search_request.user_selected_filters,
         user=user,
         project_id=project_id,
-        persona_id=persona_id,
         user_file_ids=user_uploaded_persona_files,
         persona_document_sets=persona_document_sets,
         persona_time_cutoff=persona_time_cutoff,

backend/onyx/db/code_interpreter.py

@@ -1,21 +0,0 @@
-from sqlalchemy import select
-from sqlalchemy.orm import Session
-
-from onyx.db.models import CodeInterpreterServer
-
-
-def fetch_code_interpreter_server(
-    db_session: Session,
-) -> CodeInterpreterServer:
-    server = db_session.scalars(select(CodeInterpreterServer)).one()
-    return server
-
-
-def update_code_interpreter_server_enabled(
-    db_session: Session,
-    enabled: bool,
-) -> CodeInterpreterServer:
-    server = db_session.scalars(select(CodeInterpreterServer)).one()
-    server.server_enabled = enabled
-    db_session.commit()
-    return server

backend/onyx/db/llm.py

@@ -619,7 +619,7 @@ def update_default_provider(provider_id: int, db_session: Session) -> None:
     _update_default_model(
         db_session,
         provider_id,
-        provider.default_model_name,  # type: ignore[arg-type]
+        provider.default_model_name,
         LLMModelFlowType.CHAT,
     )
 

backend/onyx/db/models.py

@@ -2822,17 +2822,13 @@ class LLMProvider(Base):
     custom_config: Mapped[dict[str, str] | None] = mapped_column(
         postgresql.JSONB(), nullable=True
     )
-
-    # Deprecated: use LLMModelFlow with CHAT flow type instead
-    default_model_name: Mapped[str | None] = mapped_column(String, nullable=True)
+    default_model_name: Mapped[str] = mapped_column(String)
 
     deployment_name: Mapped[str | None] = mapped_column(String, nullable=True)
 
-    # Deprecated: use LLMModelFlow.is_default with CHAT flow type instead
-    is_default_provider: Mapped[bool | None] = mapped_column(Boolean, nullable=True)
-    # Deprecated: use LLMModelFlow.is_default with VISION flow type instead
+    # should only be set for a single provider
+    is_default_provider: Mapped[bool | None] = mapped_column(Boolean, unique=True)
     is_default_vision_provider: Mapped[bool | None] = mapped_column(Boolean)
-    # Deprecated: use LLMModelFlow with VISION flow type instead
     default_vision_model: Mapped[str | None] = mapped_column(String, nullable=True)
     # EE only
     is_public: Mapped[bool] = mapped_column(Boolean, nullable=False, default=True)
@@ -2883,7 +2879,6 @@ class ModelConfiguration(Base):
     # - The end-user is configuring a model and chooses not to set a max-input-tokens limit.
     max_input_tokens: Mapped[int | None] = mapped_column(Integer, nullable=True)
 
-    # Deprecated: use LLMModelFlow with VISION flow type instead
     supports_image_input: Mapped[bool | None] = mapped_column(Boolean, nullable=True)
 
     # Human-readable display name for the model.
@@ -4275,9 +4270,6 @@ class UserFile(Base):
     needs_project_sync: Mapped[bool] = mapped_column(
         Boolean, nullable=False, default=False
     )
-    needs_persona_sync: Mapped[bool] = mapped_column(
-        Boolean, nullable=False, default=False
-    )
     last_project_sync_at: Mapped[datetime.datetime | None] = mapped_column(
         DateTime(timezone=True), nullable=True
     )
@@ -4948,11 +4940,6 @@ class ScimUserMapping(Base):
         ForeignKey("user.id", ondelete="CASCADE"), unique=True, nullable=False
     )
     scim_username: Mapped[str | None] = mapped_column(String, nullable=True)
-    department: Mapped[str | None] = mapped_column(String, nullable=True)
-    manager: Mapped[str | None] = mapped_column(String, nullable=True)
-    given_name: Mapped[str | None] = mapped_column(String, nullable=True)
-    family_name: Mapped[str | None] = mapped_column(String, nullable=True)
-    scim_emails_json: Mapped[str | None] = mapped_column(Text, nullable=True)
 
     created_at: Mapped[datetime.datetime] = mapped_column(
         DateTime(timezone=True), server_default=func.now(), nullable=False

backend/onyx/db/persona.py

@@ -765,9 +765,6 @@ def mark_persona_as_deleted(
 ) -> None:
     persona = get_persona_by_id(persona_id=persona_id, user=user, db_session=db_session)
     persona.deleted = True
-    affected_file_ids = [uf.id for uf in persona.user_files]
-    if affected_file_ids:
-        _mark_files_need_persona_sync(db_session, affected_file_ids)
     db_session.commit()
 
 
@@ -779,13 +776,11 @@ def mark_persona_as_not_deleted(
     persona = get_persona_by_id(
         persona_id=persona_id, user=user, db_session=db_session, include_deleted=True
     )
-    if not persona.deleted:
+    if persona.deleted:
+        persona.deleted = False
+        db_session.commit()
+    else:
         raise ValueError(f"Persona with ID {persona_id} is not deleted.")
-    persona.deleted = False
-    affected_file_ids = [uf.id for uf in persona.user_files]
-    if affected_file_ids:
-        _mark_files_need_persona_sync(db_session, affected_file_ids)
-    db_session.commit()
 
 
 def mark_delete_persona_by_name(
@@ -851,20 +846,6 @@ def update_personas_display_priority(
         db_session.commit()
 
 
-def _mark_files_need_persona_sync(
-    db_session: Session,
-    user_file_ids: list[UUID],
-) -> None:
-    """Flag the given UserFile rows so the background sync task picks them up
-    and updates their persona metadata in the vector DB."""
-    if not user_file_ids:
-        return
-    db_session.query(UserFile).filter(UserFile.id.in_(user_file_ids)).update(
-        {UserFile.needs_persona_sync: True},
-        synchronize_session=False,
-    )
-
-
 def upsert_persona(
     user: User | None,
     name: str,
@@ -1053,13 +1034,8 @@ def upsert_persona(
             existing_persona.tools = tools or []
 
         if user_file_ids is not None:
-            old_file_ids = {uf.id for uf in existing_persona.user_files}
-            new_file_ids = {uf.id for uf in (user_files or [])}
-            affected_file_ids = old_file_ids | new_file_ids
             existing_persona.user_files.clear()
             existing_persona.user_files = user_files or []
-            if affected_file_ids:
-                _mark_files_need_persona_sync(db_session, list(affected_file_ids))
 
         if hierarchy_node_ids is not None:
             existing_persona.hierarchy_nodes.clear()
@@ -1113,8 +1089,6 @@ def upsert_persona(
             attached_documents=attached_documents or [],
         )
         db_session.add(new_persona)
-        if user_files:
-            _mark_files_need_persona_sync(db_session, [uf.id for uf in user_files])
         persona = new_persona
     if commit:
         db_session.commit()

backend/onyx/db/seeding/chat_history_seeding.py

@@ -2,7 +2,6 @@ import random
 from datetime import datetime
 from datetime import timedelta
 from logging import getLogger
-from uuid import UUID
 
 from onyx.configs.constants import MessageType
 from onyx.db.chat import create_chat_session
@@ -14,26 +13,18 @@ from onyx.db.models import ChatSession
 logger = getLogger(__name__)
 
 
-def seed_chat_history(
-    num_sessions: int,
-    num_messages: int,
-    days: int,
-    user_id: UUID | None = None,
-    persona_id: int | None = None,
-) -> None:
+def seed_chat_history(num_sessions: int, num_messages: int, days: int) -> None:
     """Utility function to seed chat history for testing.
 
     num_sessions: the number of sessions to seed
     num_messages: the number of messages to seed per sessions
     days: the number of days looking backwards from the current time over which to randomize
     the times.
-    user_id: optional user to associate with sessions
-    persona_id: optional persona/assistant to associate with sessions
     """
     with get_session_with_current_tenant() as db_session:
         logger.info(f"Seeding {num_sessions} sessions.")
         for y in range(0, num_sessions):
-            create_chat_session(db_session, f"pytest_session_{y}", user_id, persona_id)
+            create_chat_session(db_session, f"pytest_session_{y}", None, None)
 
         # randomize all session times
         logger.info(f"Seeding {num_messages} messages per session.")

backend/onyx/db/user_file.py

@@ -3,7 +3,6 @@ from uuid import UUID
 
 from sqlalchemy import func
 from sqlalchemy import select
-from sqlalchemy.orm import selectinload
 from sqlalchemy.orm import Session
 
 from onyx.db.models import UserFile
@@ -65,23 +64,6 @@ def fetch_user_project_ids_for_user_files(
     }
 
 
-def fetch_persona_ids_for_user_files(
-    user_file_ids: list[str],
-    db_session: Session,
-) -> dict[str, list[int]]:
-    """Fetch persona (assistant) ids for specified user files."""
-    stmt = (
-        select(UserFile)
-        .where(UserFile.id.in_(user_file_ids))
-        .options(selectinload(UserFile.assistants))
-    )
-    results = db_session.execute(stmt).scalars().all()
-    return {
-        str(user_file.id): [persona.id for persona in user_file.assistants]
-        for user_file in results
-    }
-
-
 def update_last_accessed_at_for_user_files(
     user_file_ids: list[UUID],
     db_session: Session,

backend/onyx/document_index/factory.py

@@ -11,7 +11,6 @@ from onyx.document_index.opensearch.opensearch_document_index import (
     OpenSearchOldDocumentIndex,
 )
 from onyx.document_index.vespa.index import VespaIndex
-from onyx.indexing.models import IndexingSetting
 from shared_configs.configs import MULTI_TENANT
 
 
@@ -50,11 +49,8 @@ def get_default_document_index(
 
     opensearch_retrieval_enabled = get_opensearch_retrieval_state(db_session)
     if opensearch_retrieval_enabled:
-        indexing_setting = IndexingSetting.from_db_model(search_settings)
         return OpenSearchOldDocumentIndex(
             index_name=search_settings.index_name,
-            embedding_dim=indexing_setting.final_embedding_dim,
-            embedding_precision=indexing_setting.embedding_precision,
             secondary_index_name=secondary_index_name,
             large_chunks_enabled=search_settings.large_chunks_enabled,
             secondary_large_chunks_enabled=secondary_large_chunks_enabled,
@@ -122,11 +118,8 @@ def get_all_document_indices(
     )
     opensearch_document_index: OpenSearchOldDocumentIndex | None = None
     if ENABLE_OPENSEARCH_INDEXING_FOR_ONYX:
-        indexing_setting = IndexingSetting.from_db_model(search_settings)
         opensearch_document_index = OpenSearchOldDocumentIndex(
             index_name=search_settings.index_name,
-            embedding_dim=indexing_setting.final_embedding_dim,
-            embedding_precision=indexing_setting.embedding_precision,
             secondary_index_name=None,
             large_chunks_enabled=False,
             secondary_large_chunks_enabled=None,

backend/onyx/document_index/interfaces.py

@@ -121,7 +121,6 @@ class VespaDocumentUserFields:
     """
 
     user_projects: list[int] | None = None
-    personas: list[int] | None = None
 
 
 @dataclass

backend/onyx/document_index/interfaces_new.py

@@ -148,7 +148,6 @@ class MetadataUpdateRequest(BaseModel):
     hidden: bool | None = None
     secondary_index_updated: bool | None = None
     project_ids: set[int] | None = None
-    persona_ids: set[int] | None = None
 
 
 class IndexRetrievalFilters(BaseModel):

backend/onyx/document_index/opensearch/client.py

@@ -1,7 +1,5 @@
 import logging
 import time
-from contextlib import AbstractContextManager
-from contextlib import nullcontext
 from typing import Any
 from typing import Generic
 from typing import TypeVar
@@ -85,26 +83,22 @@ def get_new_body_without_vectors(body: dict[str, Any]) -> dict[str, Any]:
     return new_body
 
 
-class OpenSearchClient(AbstractContextManager):
-    """Client for interacting with OpenSearch for cluster-level operations.
+class OpenSearchClient:
+    """Client for interacting with OpenSearch.
 
-    Args:
-        host: The host of the OpenSearch cluster.
-        port: The port of the OpenSearch cluster.
-        auth: The authentication credentials for the OpenSearch cluster. A tuple
-            of (username, password).
-        use_ssl: Whether to use SSL for the OpenSearch cluster. Defaults to
-            True.
-        verify_certs: Whether to verify the SSL certificates for the OpenSearch
-            cluster. Defaults to False.
-        ssl_show_warn: Whether to show warnings for SSL certificates. Defaults
-            to False.
-        timeout: The timeout for the OpenSearch cluster. Defaults to
-            DEFAULT_OPENSEARCH_CLIENT_TIMEOUT_S.
+    OpenSearch's Python module has pretty bad typing support so this client
+    attempts to protect the rest of the codebase from this. As a consequence,
+    most methods here return the minimum data needed for the rest of Onyx, and
+    tend to rely on Exceptions to handle errors.
+
+    TODO(andrei): This class currently assumes the structure of the database
+    schema when it returns a DocumentChunk. Make the class, or at least the
+    search method, templated on the structure the caller can expect.
     """
 
     def __init__(
         self,
+        index_name: str,
         host: str = OPENSEARCH_HOST,
         port: int = OPENSEARCH_REST_API_PORT,
         auth: tuple[str, str] = (OPENSEARCH_ADMIN_USERNAME, OPENSEARCH_ADMIN_PASSWORD),
@@ -113,8 +107,9 @@ class OpenSearchClient(AbstractContextManager):
         ssl_show_warn: bool = False,
         timeout: int = DEFAULT_OPENSEARCH_CLIENT_TIMEOUT_S,
     ):
+        self._index_name = index_name
         logger.debug(
-            f"Creating OpenSearch client with host {host}, port {port} and timeout {timeout} seconds."
+            f"Creating OpenSearch client for index {index_name} with host {host} and port {port} and timeout {timeout} seconds."
         )
         self._client = OpenSearch(
             hosts=[{"host": host, "port": port}],
@@ -130,142 +125,6 @@ class OpenSearchClient(AbstractContextManager):
             # your request body that is less than this value.
             timeout=timeout,
         )
-
-    def __exit__(self, *_: Any) -> None:
-        self.close()
-
-    def __del__(self) -> None:
-        try:
-            self.close()
-        except Exception:
-            pass
-
-    @log_function_time(print_only=True, debug_only=True, include_args=True)
-    def create_search_pipeline(
-        self,
-        pipeline_id: str,
-        pipeline_body: dict[str, Any],
-    ) -> None:
-        """Creates a search pipeline.
-
-        See the OpenSearch documentation for more information on the search
-        pipeline body.
-        https://docs.opensearch.org/latest/search-plugins/search-pipelines/index/
-
-        Args:
-            pipeline_id: The ID of the search pipeline to create.
-            pipeline_body: The body of the search pipeline to create.
-
-        Raises:
-            Exception: There was an error creating the search pipeline.
-        """
-        result = self._client.search_pipeline.put(id=pipeline_id, body=pipeline_body)
-        if not result.get("acknowledged", False):
-            raise RuntimeError(f"Failed to create search pipeline {pipeline_id}.")
-
-    @log_function_time(print_only=True, debug_only=True, include_args=True)
-    def delete_search_pipeline(self, pipeline_id: str) -> None:
-        """Deletes a search pipeline.
-
-        Args:
-            pipeline_id: The ID of the search pipeline to delete.
-
-        Raises:
-            Exception: There was an error deleting the search pipeline.
-        """
-        result = self._client.search_pipeline.delete(id=pipeline_id)
-        if not result.get("acknowledged", False):
-            raise RuntimeError(f"Failed to delete search pipeline {pipeline_id}.")
-
-    @log_function_time(print_only=True, debug_only=True, include_args=True)
-    def put_cluster_settings(self, settings: dict[str, Any]) -> bool:
-        """Puts cluster settings.
-
-        Args:
-            settings: The settings to put.
-
-        Raises:
-            Exception: There was an error putting the cluster settings.
-
-        Returns:
-            True if the settings were put successfully, False otherwise.
-        """
-        response = self._client.cluster.put_settings(body=settings)
-        if response.get("acknowledged", False):
-            logger.info("Successfully put cluster settings.")
-            return True
-        else:
-            logger.error(f"Failed to put cluster settings: {response}.")
-            return False
-
-    @log_function_time(print_only=True, debug_only=True)
-    def ping(self) -> bool:
-        """Pings the OpenSearch cluster.
-
-        Returns:
-            True if OpenSearch could be reached, False if it could not.
-        """
-        return self._client.ping()
-
-    @log_function_time(print_only=True, debug_only=True)
-    def close(self) -> None:
-        """Closes the client.
-
-        Raises:
-            Exception: There was an error closing the client.
-        """
-        self._client.close()
-
-
-class OpenSearchIndexClient(OpenSearchClient):
-    """Client for interacting with OpenSearch for index-level operations.
-
-    OpenSearch's Python module has pretty bad typing support so this client
-    attempts to protect the rest of the codebase from this. As a consequence,
-    most methods here return the minimum data needed for the rest of Onyx, and
-    tend to rely on Exceptions to handle errors.
-
-    TODO(andrei): This class currently assumes the structure of the database
-    schema when it returns a DocumentChunk. Make the class, or at least the
-    search method, templated on the structure the caller can expect.
-
-    Args:
-        index_name: The name of the index to interact with.
-        host: The host of the OpenSearch cluster.
-        port: The port of the OpenSearch cluster.
-        auth: The authentication credentials for the OpenSearch cluster. A tuple
-            of (username, password).
-        use_ssl: Whether to use SSL for the OpenSearch cluster. Defaults to
-            True.
-        verify_certs: Whether to verify the SSL certificates for the OpenSearch
-            cluster. Defaults to False.
-        ssl_show_warn: Whether to show warnings for SSL certificates. Defaults
-            to False.
-        timeout: The timeout for the OpenSearch cluster. Defaults to
-            DEFAULT_OPENSEARCH_CLIENT_TIMEOUT_S.
-    """
-
-    def __init__(
-        self,
-        index_name: str,
-        host: str = OPENSEARCH_HOST,
-        port: int = OPENSEARCH_REST_API_PORT,
-        auth: tuple[str, str] = (OPENSEARCH_ADMIN_USERNAME, OPENSEARCH_ADMIN_PASSWORD),
-        use_ssl: bool = True,
-        verify_certs: bool = False,
-        ssl_show_warn: bool = False,
-        timeout: int = DEFAULT_OPENSEARCH_CLIENT_TIMEOUT_S,
-    ):
-        super().__init__(
-            host=host,
-            port=port,
-            auth=auth,
-            use_ssl=use_ssl,
-            verify_certs=verify_certs,
-            ssl_show_warn=ssl_show_warn,
-            timeout=timeout,
-        )
-        self._index_name = index_name
         logger.debug(
             f"OpenSearch client created successfully for index {self._index_name}."
         )
@@ -333,38 +192,6 @@ class OpenSearchIndexClient(OpenSearchClient):
         """
         return self._client.indices.exists(index=self._index_name)
 
-    @log_function_time(print_only=True, debug_only=True, include_args=True)
-    def put_mapping(self, mappings: dict[str, Any]) -> None:
-        """Updates the index mapping in an idempotent manner.
-
-        - Existing fields with the same definition: No-op (succeeds silently).
-        - New fields: Added to the index.
-        - Existing fields with different types: Raises exception (requires
-          reindex).
-
-        See the OpenSearch documentation for more information:
-        https://docs.opensearch.org/latest/api-reference/index-apis/put-mapping/
-
-        Args:
-            mappings: The complete mapping definition to apply. This will be
-                merged with existing mappings in the index.
-
-        Raises:
-            Exception: There was an error updating the mappings, such as
-                attempting to change the type of an existing field.
-        """
-        logger.debug(
-            f"Putting mappings for index {self._index_name} with mappings {mappings}."
-        )
-        response = self._client.indices.put_mapping(
-            index=self._index_name, body=mappings
-        )
-        if not response.get("acknowledged", False):
-            raise RuntimeError(
-                f"Failed to put the mapping update for index {self._index_name}."
-            )
-        logger.debug(f"Successfully put mappings for index {self._index_name}.")
-
     @log_function_time(print_only=True, debug_only=True, include_args=True)
     def validate_index(self, expected_mappings: dict[str, Any]) -> bool:
         """Validates the index.
@@ -783,6 +610,43 @@ class OpenSearchIndexClient(OpenSearchClient):
         )
         return DocumentChunk.model_validate(document_chunk_source)
 
+    @log_function_time(print_only=True, debug_only=True, include_args=True)
+    def create_search_pipeline(
+        self,
+        pipeline_id: str,
+        pipeline_body: dict[str, Any],
+    ) -> None:
+        """Creates a search pipeline.
+
+        See the OpenSearch documentation for more information on the search
+        pipeline body.
+        https://docs.opensearch.org/latest/search-plugins/search-pipelines/index/
+
+        Args:
+            pipeline_id: The ID of the search pipeline to create.
+            pipeline_body: The body of the search pipeline to create.
+
+        Raises:
+            Exception: There was an error creating the search pipeline.
+        """
+        result = self._client.search_pipeline.put(id=pipeline_id, body=pipeline_body)
+        if not result.get("acknowledged", False):
+            raise RuntimeError(f"Failed to create search pipeline {pipeline_id}.")
+
+    @log_function_time(print_only=True, debug_only=True, include_args=True)
+    def delete_search_pipeline(self, pipeline_id: str) -> None:
+        """Deletes a search pipeline.
+
+        Args:
+            pipeline_id: The ID of the search pipeline to delete.
+
+        Raises:
+            Exception: There was an error deleting the search pipeline.
+        """
+        result = self._client.search_pipeline.delete(id=pipeline_id)
+        if not result.get("acknowledged", False):
+            raise RuntimeError(f"Failed to delete search pipeline {pipeline_id}.")
+
     @log_function_time(print_only=True, debug_only=True)
     def search(
         self, body: dict[str, Any], search_pipeline_id: str | None
@@ -943,6 +807,48 @@ class OpenSearchIndexClient(OpenSearchClient):
         """
         self._client.indices.refresh(index=self._index_name)
 
+    @log_function_time(print_only=True, debug_only=True, include_args=True)
+    def put_cluster_settings(self, settings: dict[str, Any]) -> bool:
+        """Puts cluster settings.
+
+        Args:
+            settings: The settings to put.
+
+        Raises:
+            Exception: There was an error putting the cluster settings.
+
+        Returns:
+            True if the settings were put successfully, False otherwise.
+        """
+        response = self._client.cluster.put_settings(body=settings)
+        if response.get("acknowledged", False):
+            logger.info("Successfully put cluster settings.")
+            return True
+        else:
+            logger.error(f"Failed to put cluster settings: {response}.")
+            return False
+
+    @log_function_time(print_only=True, debug_only=True)
+    def ping(self) -> bool:
+        """Pings the OpenSearch cluster.
+
+        Returns:
+            True if OpenSearch could be reached, False if it could not.
+        """
+        return self._client.ping()
+
+    @log_function_time(print_only=True, debug_only=True)
+    def close(self) -> None:
+        """Closes the client.
+
+        TODO(andrei): Can we have some way to auto close when the client no
+        longer has any references?
+
+        Raises:
+            Exception: There was an error closing the client.
+        """
+        self._client.close()
+
     def _get_hits_and_profile_from_search_result(
         self, result: dict[str, Any]
     ) -> tuple[list[Any], int | None, bool | None, dict[str, Any], dict[str, Any]]:
@@ -1039,7 +945,14 @@ def wait_for_opensearch_with_timeout(
     Returns:
         True if OpenSearch is ready, False otherwise.
     """
-    with nullcontext(client) if client else OpenSearchClient() as client:
+    made_client = False
+    try:
+        if client is None:
+            # NOTE: index_name does not matter because we are only using this object
+            # to ping.
+            # TODO(andrei): Make this better.
+            client = OpenSearchClient(index_name="")
+            made_client = True
         time_start = time.monotonic()
         while True:
             if client.ping():
@@ -1056,3 +969,7 @@ def wait_for_opensearch_with_timeout(
                 f"[OpenSearch] Readiness probe ongoing. elapsed={time_elapsed:.1f} timeout={wait_limit_s:.1f}"
             )
             time.sleep(wait_interval_s)
+    finally:
+        if made_client:
+            assert client is not None
+            client.close()

backend/onyx/document_index/opensearch/opensearch_document_index.py

@@ -7,7 +7,6 @@ from opensearchpy import NotFoundError
 
 from onyx.access.models import DocumentAccess
 from onyx.configs.app_configs import USING_AWS_MANAGED_OPENSEARCH
-from onyx.configs.app_configs import VERIFY_CREATE_OPENSEARCH_INDEX_ON_INIT_MT
 from onyx.configs.chat_configs import NUM_RETURNED_HITS
 from onyx.configs.chat_configs import TITLE_CONTENT_RATIO
 from onyx.configs.constants import PUBLIC_DOC_PAT
@@ -41,7 +40,6 @@ from onyx.document_index.interfaces_new import IndexingMetadata
 from onyx.document_index.interfaces_new import MetadataUpdateRequest
 from onyx.document_index.interfaces_new import TenantState
 from onyx.document_index.opensearch.client import OpenSearchClient
-from onyx.document_index.opensearch.client import OpenSearchIndexClient
 from onyx.document_index.opensearch.client import SearchHit
 from onyx.document_index.opensearch.cluster_settings import OPENSEARCH_CLUSTER_SETTINGS
 from onyx.document_index.opensearch.schema import ACCESS_CONTROL_LIST_FIELD_NAME
@@ -52,7 +50,6 @@ from onyx.document_index.opensearch.schema import DocumentSchema
 from onyx.document_index.opensearch.schema import get_opensearch_doc_chunk_id
 from onyx.document_index.opensearch.schema import GLOBAL_BOOST_FIELD_NAME
 from onyx.document_index.opensearch.schema import HIDDEN_FIELD_NAME
-from onyx.document_index.opensearch.schema import PERSONAS_FIELD_NAME
 from onyx.document_index.opensearch.schema import USER_PROJECTS_FIELD_NAME
 from onyx.document_index.opensearch.search import DocumentQuery
 from onyx.document_index.opensearch.search import (
@@ -95,25 +92,6 @@ def generate_opensearch_filtered_access_control_list(
     return list(access_control_list)
 
 
-def set_cluster_state(client: OpenSearchClient) -> None:
-    if not client.put_cluster_settings(settings=OPENSEARCH_CLUSTER_SETTINGS):
-        logger.error(
-            "Failed to put cluster settings. If the settings have never been set before, "
-            "this may cause unexpected index creation when indexing documents into an "
-            "index that does not exist, or may cause expected logs to not appear. If this "
-            "is not the first time running Onyx against this instance of OpenSearch, these "
-            "settings have likely already been set. Not taking any further action..."
-        )
-    client.create_search_pipeline(
-        pipeline_id=MIN_MAX_NORMALIZATION_PIPELINE_NAME,
-        pipeline_body=MIN_MAX_NORMALIZATION_PIPELINE_CONFIG,
-    )
-    client.create_search_pipeline(
-        pipeline_id=ZSCORE_NORMALIZATION_PIPELINE_NAME,
-        pipeline_body=ZSCORE_NORMALIZATION_PIPELINE_CONFIG,
-    )
-
-
 def _convert_retrieved_opensearch_chunk_to_inference_chunk_uncleaned(
     chunk: DocumentChunk,
     score: float | None,
@@ -237,7 +215,6 @@ def _convert_onyx_chunk_to_opensearch_document(
         # OpenSearch and it will not store any data at all for this field, which
         # is different from supplying an empty list.
         user_projects=chunk.user_project or None,
-        personas=chunk.personas or None,
         primary_owners=get_experts_stores_representations(
             chunk.source_document.primary_owners
         ),
@@ -269,8 +246,6 @@ class OpenSearchOldDocumentIndex(OldDocumentIndex):
     def __init__(
         self,
         index_name: str,
-        embedding_dim: int,
-        embedding_precision: EmbeddingPrecision,
         secondary_index_name: str | None,
         large_chunks_enabled: bool,  # noqa: ARG002
         secondary_large_chunks_enabled: bool | None,  # noqa: ARG002
@@ -281,6 +256,10 @@ class OpenSearchOldDocumentIndex(OldDocumentIndex):
             index_name=index_name,
             secondary_index_name=secondary_index_name,
         )
+        if multitenant:
+            raise ValueError(
+                "Bug: OpenSearch is not yet ready for multitenant environments but something tried to use it."
+            )
         if multitenant != MULTI_TENANT:
             raise ValueError(
                 "Bug: Multitenant mismatch when initializing an OpenSearchDocumentIndex. "
@@ -288,10 +267,8 @@ class OpenSearchOldDocumentIndex(OldDocumentIndex):
             )
         tenant_id = get_current_tenant_id()
         self._real_index = OpenSearchDocumentIndex(
-            tenant_state=TenantState(tenant_id=tenant_id, multitenant=multitenant),
             index_name=index_name,
-            embedding_dim=embedding_dim,
-            embedding_precision=embedding_precision,
+            tenant_state=TenantState(tenant_id=tenant_id, multitenant=multitenant),
         )
 
     @staticmethod
@@ -300,8 +277,9 @@ class OpenSearchOldDocumentIndex(OldDocumentIndex):
         embedding_dims: list[int],
         embedding_precisions: list[EmbeddingPrecision],
     ) -> None:
+        # TODO(andrei): Implement.
         raise NotImplementedError(
-            "Bug: Multitenant index registration is not supported for OpenSearch."
+            "Multitenant index registration is not yet implemented for OpenSearch."
         )
 
     def ensure_indices_exist(
@@ -384,11 +362,6 @@ class OpenSearchOldDocumentIndex(OldDocumentIndex):
                 if user_fields and user_fields.user_projects
                 else None
             ),
-            persona_ids=(
-                set(user_fields.personas)
-                if user_fields and user_fields.personas
-                else None
-            ),
         )
 
         try:
@@ -491,37 +464,19 @@ class OpenSearchDocumentIndex(DocumentIndex):
     for an OpenSearch search engine instance. It handles the complete lifecycle
     of document chunks within a specific OpenSearch index/schema.
 
-    Each kind of embedding used should correspond to a different instance of
-    this class, and therefore a different index in OpenSearch.
-
-    If in a multitenant environment and
-    VERIFY_CREATE_OPENSEARCH_INDEX_ON_INIT_MT, will verify and create the index
-    if necessary on initialization. This is because there is no logic which runs
-    on cluster restart which scans through all search settings over all tenants
-    and creates the relevant indices.
-
-    Args:
-        tenant_state: The tenant state of the caller.
-        index_name: The name of the index to interact with.
-        embedding_dim: The dimensionality of the embeddings used for the index.
-        embedding_precision: The precision of the embeddings used for the index.
+    Although not yet used in this way in the codebase, each kind of embedding
+    used should correspond to a different instance of this class, and therefore
+    a different index in OpenSearch.
     """
 
     def __init__(
         self,
-        tenant_state: TenantState,
         index_name: str,
-        embedding_dim: int,
-        embedding_precision: EmbeddingPrecision,
+        tenant_state: TenantState,
     ) -> None:
         self._index_name: str = index_name
         self._tenant_state: TenantState = tenant_state
-        self._client = OpenSearchIndexClient(index_name=self._index_name)
-
-        if self._tenant_state.multitenant and VERIFY_CREATE_OPENSEARCH_INDEX_ON_INIT_MT:
-            self.verify_and_create_index_if_necessary(
-                embedding_dim=embedding_dim, embedding_precision=embedding_precision
-            )
+        self._os_client = OpenSearchClient(index_name=self._index_name)
 
     def verify_and_create_index_if_necessary(
         self,
@@ -530,15 +485,10 @@ class OpenSearchDocumentIndex(DocumentIndex):
     ) -> None:
         """Verifies and creates the index if necessary.
 
-        Also puts the desired cluster settings if not in a multitenant
-        environment.
+        Also puts the desired cluster settings.
 
-        Also puts the desired search pipeline state if not in a multitenant
-        environment, creating the pipelines if they do not exist and updating
-        them otherwise.
-
-        In a multitenant environment, the above steps happen explicitly on
-        setup.
+        Also puts the desired search pipeline state, creating the pipelines if
+        they do not exist and updating them otherwise.
 
         Args:
             embedding_dim: Vector dimensionality for the vector similarity part
@@ -551,38 +501,47 @@ class OpenSearchDocumentIndex(DocumentIndex):
                 search pipelines.
         """
         logger.debug(
-            f"[OpenSearchDocumentIndex] Verifying and creating index {self._index_name} if "
-            f"necessary, with embedding dimension {embedding_dim}."
+            f"[OpenSearchDocumentIndex] Verifying and creating index {self._index_name} if necessary, "
+            f"with embedding dimension {embedding_dim}."
         )
-
-        if not self._tenant_state.multitenant:
-            set_cluster_state(self._client)
-
         expected_mappings = DocumentSchema.get_document_schema(
             embedding_dim, self._tenant_state.multitenant
         )
-
-        if not self._client.index_exists():
+        if not self._os_client.put_cluster_settings(
+            settings=OPENSEARCH_CLUSTER_SETTINGS
+        ):
+            logger.error(
+                f"Failed to put cluster settings for index {self._index_name}. If the settings have never been set before this "
+                "may cause unexpected index creation when indexing documents into an index that does not exist, or may cause "
+                "expected logs to not appear. If this is not the first time running Onyx against this instance of OpenSearch, "
+                "these settings have likely already been set. Not taking any further action..."
+            )
+        if not self._os_client.index_exists():
             if USING_AWS_MANAGED_OPENSEARCH:
                 index_settings = (
                     DocumentSchema.get_index_settings_for_aws_managed_opensearch()
                 )
             else:
                 index_settings = DocumentSchema.get_index_settings()
-            self._client.create_index(
+            self._os_client.create_index(
                 mappings=expected_mappings,
                 settings=index_settings,
             )
-        else:
-            # Ensure schema is up to date by applying the current mappings.
-            try:
-                self._client.put_mapping(expected_mappings)
-            except Exception as e:
-                logger.error(
-                    f"Failed to update mappings for index {self._index_name}. This likely means a "
-                    f"field type was changed which requires reindexing. Error: {e}"
-                )
-                raise
+        if not self._os_client.validate_index(
+            expected_mappings=expected_mappings,
+        ):
+            raise RuntimeError(
+                f"The index {self._index_name} is not valid. The expected mappings do not match the actual mappings."
+            )
+
+        self._os_client.create_search_pipeline(
+            pipeline_id=MIN_MAX_NORMALIZATION_PIPELINE_NAME,
+            pipeline_body=MIN_MAX_NORMALIZATION_PIPELINE_CONFIG,
+        )
+        self._os_client.create_search_pipeline(
+            pipeline_id=ZSCORE_NORMALIZATION_PIPELINE_NAME,
+            pipeline_body=ZSCORE_NORMALIZATION_PIPELINE_CONFIG,
+        )
 
     def index(
         self,
@@ -654,7 +613,7 @@ class OpenSearchDocumentIndex(DocumentIndex):
             )
             # Now index. This will raise if a chunk of the same ID exists, which
             # we do not expect because we should have deleted all chunks.
-            self._client.bulk_index_documents(
+            self._os_client.bulk_index_documents(
                 documents=chunk_batch,
                 tenant_state=self._tenant_state,
             )
@@ -694,7 +653,7 @@ class OpenSearchDocumentIndex(DocumentIndex):
             tenant_state=self._tenant_state,
         )
 
-        return self._client.delete_by_query(query_body)
+        return self._os_client.delete_by_query(query_body)
 
     def update(
         self,
@@ -750,10 +709,6 @@ class OpenSearchDocumentIndex(DocumentIndex):
                 properties_to_update[USER_PROJECTS_FIELD_NAME] = list(
                     update_request.project_ids
                 )
-            if update_request.persona_ids is not None:
-                properties_to_update[PERSONAS_FIELD_NAME] = list(
-                    update_request.persona_ids
-                )
 
             if not properties_to_update:
                 if len(update_request.document_ids) > 1:
@@ -794,7 +749,7 @@ class OpenSearchDocumentIndex(DocumentIndex):
                         document_id=doc_id,
                         chunk_index=chunk_index,
                     )
-                    self._client.update_document(
+                    self._os_client.update_document(
                         document_chunk_id=document_chunk_id,
                         properties_to_update=properties_to_update,
                     )
@@ -833,7 +788,7 @@ class OpenSearchDocumentIndex(DocumentIndex):
                 min_chunk_index=chunk_request.min_chunk_ind,
                 max_chunk_index=chunk_request.max_chunk_ind,
             )
-            search_hits = self._client.search(
+            search_hits = self._os_client.search(
                 body=query_body,
                 search_pipeline_id=None,
             )
@@ -883,7 +838,7 @@ class OpenSearchDocumentIndex(DocumentIndex):
         # NOTE: Using z-score normalization here because it's better for hybrid search from a theoretical standpoint.
         # Empirically on a small dataset of up to 10K docs, it's not very different. Likely more impactful at scale.
         # https://opensearch.org/blog/introducing-the-z-score-normalization-technique-for-hybrid-search/
-        search_hits: list[SearchHit[DocumentChunk]] = self._client.search(
+        search_hits: list[SearchHit[DocumentChunk]] = self._os_client.search(
             body=query_body,
             search_pipeline_id=ZSCORE_NORMALIZATION_PIPELINE_NAME,
         )
@@ -915,7 +870,7 @@ class OpenSearchDocumentIndex(DocumentIndex):
             index_filters=filters,
             num_to_retrieve=num_to_retrieve,
         )
-        search_hits: list[SearchHit[DocumentChunk]] = self._client.search(
+        search_hits: list[SearchHit[DocumentChunk]] = self._os_client.search(
             body=query_body,
             search_pipeline_id=None,
         )
@@ -943,6 +898,6 @@ class OpenSearchDocumentIndex(DocumentIndex):
         # Do not raise if the document already exists, just update. This is
         # because the document may already have been indexed during the
         # OpenSearch transition period.
-        self._client.bulk_index_documents(
+        self._os_client.bulk_index_documents(
             documents=chunks, tenant_state=self._tenant_state, update_if_exists=True
         )

backend/onyx/document_index/opensearch/schema.py

@@ -41,7 +41,6 @@ IMAGE_FILE_ID_FIELD_NAME = "image_file_id"
 SOURCE_LINKS_FIELD_NAME = "source_links"
 DOCUMENT_SETS_FIELD_NAME = "document_sets"
 USER_PROJECTS_FIELD_NAME = "user_projects"
-PERSONAS_FIELD_NAME = "personas"
 DOCUMENT_ID_FIELD_NAME = "document_id"
 CHUNK_INDEX_FIELD_NAME = "chunk_index"
 MAX_CHUNK_SIZE_FIELD_NAME = "max_chunk_size"
@@ -157,7 +156,6 @@ class DocumentChunk(BaseModel):
 
     document_sets: list[str] | None = None
     user_projects: list[int] | None = None
-    personas: list[int] | None = None
     primary_owners: list[str] | None = None
     secondary_owners: list[str] | None = None
 
@@ -487,7 +485,6 @@ class DocumentSchema:
                 # Product-specific fields.
                 DOCUMENT_SETS_FIELD_NAME: {"type": "keyword"},
                 USER_PROJECTS_FIELD_NAME: {"type": "integer"},
-                PERSONAS_FIELD_NAME: {"type": "integer"},
                 PRIMARY_OWNERS_FIELD_NAME: {"type": "keyword"},
                 SECONDARY_OWNERS_FIELD_NAME: {"type": "keyword"},
                 # OpenSearch metadata fields.

backend/onyx/document_index/opensearch/search.py

@@ -28,7 +28,6 @@ from onyx.document_index.opensearch.schema import HIDDEN_FIELD_NAME
 from onyx.document_index.opensearch.schema import LAST_UPDATED_FIELD_NAME
 from onyx.document_index.opensearch.schema import MAX_CHUNK_SIZE_FIELD_NAME
 from onyx.document_index.opensearch.schema import METADATA_LIST_FIELD_NAME
-from onyx.document_index.opensearch.schema import PERSONAS_FIELD_NAME
 from onyx.document_index.opensearch.schema import PUBLIC_FIELD_NAME
 from onyx.document_index.opensearch.schema import set_or_convert_timezone_to_utc
 from onyx.document_index.opensearch.schema import SOURCE_TYPE_FIELD_NAME
@@ -145,7 +144,6 @@ class DocumentQuery:
             document_sets=index_filters.document_set or [],
             user_file_ids=index_filters.user_file_ids or [],
             project_id=index_filters.project_id,
-            persona_id=index_filters.persona_id,
             time_cutoff=index_filters.time_cutoff,
             min_chunk_index=min_chunk_index,
             max_chunk_index=max_chunk_index,
@@ -204,7 +202,6 @@ class DocumentQuery:
             document_sets=[],
             user_file_ids=[],
             project_id=None,
-            persona_id=None,
             time_cutoff=None,
             min_chunk_index=None,
             max_chunk_index=None,
@@ -270,7 +267,6 @@ class DocumentQuery:
             document_sets=index_filters.document_set or [],
             user_file_ids=index_filters.user_file_ids or [],
             project_id=index_filters.project_id,
-            persona_id=index_filters.persona_id,
             time_cutoff=index_filters.time_cutoff,
             min_chunk_index=None,
             max_chunk_index=None,
@@ -338,7 +334,6 @@ class DocumentQuery:
             document_sets=index_filters.document_set or [],
             user_file_ids=index_filters.user_file_ids or [],
             project_id=index_filters.project_id,
-            persona_id=index_filters.persona_id,
             time_cutoff=index_filters.time_cutoff,
             min_chunk_index=None,
             max_chunk_index=None,
@@ -501,7 +496,6 @@ class DocumentQuery:
         document_sets: list[str],
         user_file_ids: list[UUID],
         project_id: int | None,
-        persona_id: int | None,
         time_cutoff: datetime | None,
         min_chunk_index: int | None,
         max_chunk_index: int | None,
@@ -536,8 +530,6 @@ class DocumentQuery:
                 retrieved.
             project_id: If not None, only documents with this project ID in user
                 projects will be retrieved.
-            persona_id: If not None, only documents whose personas array
-                contains this persona ID will be retrieved.
             time_cutoff: Time cutoff for the documents to retrieve. If not None,
                 Documents which were last updated before this date will not be
                 returned. For documents which do not have a value for their last
@@ -635,9 +627,6 @@ class DocumentQuery:
             )
             return user_project_filter
 
-        def _get_persona_filter(persona_id: int) -> dict[str, Any]:
-            return {"term": {PERSONAS_FIELD_NAME: {"value": persona_id}}}
-
         def _get_time_cutoff_filter(time_cutoff: datetime) -> dict[str, Any]:
             # Convert to UTC if not already so the cutoff is comparable to the
             # document data.
@@ -791,9 +780,6 @@ class DocumentQuery:
             # document's user projects list.
             filter_clauses.append(_get_user_project_filter(project_id))
 
-        if persona_id is not None:
-            filter_clauses.append(_get_persona_filter(persona_id))
-
         if time_cutoff is not None:
             # If a time cutoff is provided, the caller will only retrieve
             # documents where the document was last updated at or after the time

backend/onyx/document_index/vespa/app_config/schemas/danswer_chunk.sd.jinja

@@ -181,11 +181,6 @@ schema {{ schema_name }} {
             rank: filter
             attribute: fast-search
         }
-        field personas type array<int> {
-            indexing: summary | attribute
-            rank: filter
-            attribute: fast-search
-        }
     }
 
     # If using different tokenization settings, the fieldset has to be removed, and the field must

backend/onyx/document_index/vespa/index.py

@@ -689,9 +689,6 @@ class VespaIndex(DocumentIndex):
         project_ids: set[int] | None = None
         if user_fields is not None and user_fields.user_projects is not None:
             project_ids = set(user_fields.user_projects)
-        persona_ids: set[int] | None = None
-        if user_fields is not None and user_fields.personas is not None:
-            persona_ids = set(user_fields.personas)
         update_request = MetadataUpdateRequest(
             document_ids=[doc_id],
             doc_id_to_chunk_cnt={
@@ -702,7 +699,6 @@ class VespaIndex(DocumentIndex):
             boost=fields.boost if fields is not None else None,
             hidden=fields.hidden if fields is not None else None,
             project_ids=project_ids,
-            persona_ids=persona_ids,
         )
 
         vespa_document_index.update([update_request])

backend/onyx/document_index/vespa/indexing_utils.py

@@ -46,7 +46,6 @@ from onyx.document_index.vespa_constants import METADATA
 from onyx.document_index.vespa_constants import METADATA_LIST
 from onyx.document_index.vespa_constants import METADATA_SUFFIX
 from onyx.document_index.vespa_constants import NUM_THREADS
-from onyx.document_index.vespa_constants import PERSONAS
 from onyx.document_index.vespa_constants import PRIMARY_OWNERS
 from onyx.document_index.vespa_constants import SECONDARY_OWNERS
 from onyx.document_index.vespa_constants import SECTION_CONTINUATION
@@ -219,7 +218,6 @@ def _index_vespa_chunk(
         # still called `image_file_name` in Vespa for backwards compatibility
         IMAGE_FILE_NAME: chunk.image_file_id,
         USER_PROJECT: chunk.user_project if chunk.user_project is not None else [],
-        PERSONAS: chunk.personas if chunk.personas is not None else [],
         BOOST: chunk.boost,
         AGGREGATED_CHUNK_BOOST_FACTOR: chunk.aggregated_chunk_boost_factor,
     }

backend/onyx/document_index/vespa/shared_utils/vespa_request_builders.py

@@ -12,7 +12,6 @@ from onyx.document_index.vespa_constants import DOCUMENT_ID
 from onyx.document_index.vespa_constants import DOCUMENT_SETS
 from onyx.document_index.vespa_constants import HIDDEN
 from onyx.document_index.vespa_constants import METADATA_LIST
-from onyx.document_index.vespa_constants import PERSONAS
 from onyx.document_index.vespa_constants import SOURCE_TYPE
 from onyx.document_index.vespa_constants import TENANT_ID
 from onyx.document_index.vespa_constants import USER_PROJECT
@@ -150,18 +149,6 @@ def build_vespa_filters(
         # Vespa YQL 'contains' expects a string literal; quote the integer
         return f'({USER_PROJECT} contains "{pid}") and '
 
-    def _build_persona_filter(
-        persona_id: int | None,
-    ) -> str:
-        if persona_id is None:
-            return ""
-        try:
-            pid = int(persona_id)
-        except Exception:
-            logger.warning(f"Invalid persona ID: {persona_id}")
-            return ""
-        return f'({PERSONAS} contains "{pid}") and '
-
     # Start building the filter string
     filter_str = f"!({HIDDEN}=true) and " if not include_hidden else ""
 
@@ -205,9 +192,6 @@ def build_vespa_filters(
     # User project filter (array<int> attribute membership)
     filter_str += _build_user_project_filter(filters.project_id)
 
-    # Persona filter (array<int> attribute membership)
-    filter_str += _build_persona_filter(filters.persona_id)
-
     # Time filter
     filter_str += _build_time_filter(filters.time_cutoff)
 

backend/onyx/document_index/vespa/vespa_document_index.py

@@ -183,10 +183,6 @@ def _update_single_chunk(
         model_config = {"frozen": True}
         assign: list[int]
 
-    class _Personas(BaseModel):
-        model_config = {"frozen": True}
-        assign: list[int]
-
     class _VespaPutFields(BaseModel):
         model_config = {"frozen": True}
         # The names of these fields are based the Vespa schema. Changes to the
@@ -197,7 +193,6 @@ def _update_single_chunk(
         access_control_list: _AccessControl | None = None
         hidden: _Hidden | None = None
         user_project: _UserProjects | None = None
-        personas: _Personas | None = None
 
     class _VespaPutRequest(BaseModel):
         model_config = {"frozen": True}
@@ -232,11 +227,6 @@ def _update_single_chunk(
         if update_request.project_ids is not None
         else None
     )
-    personas_update: _Personas | None = (
-        _Personas(assign=list(update_request.persona_ids))
-        if update_request.persona_ids is not None
-        else None
-    )
 
     vespa_put_fields = _VespaPutFields(
         boost=boost_update,
@@ -244,7 +234,6 @@ def _update_single_chunk(
         access_control_list=access_update,
         hidden=hidden_update,
         user_project=user_projects_update,
-        personas=personas_update,
     )
 
     vespa_put_request = _VespaPutRequest(

backend/onyx/document_index/vespa_constants.py

@@ -58,7 +58,6 @@ DOCUMENT_SETS = "document_sets"
 USER_FILE = "user_file"
 USER_FOLDER = "user_folder"
 USER_PROJECT = "user_project"
-PERSONAS = "personas"
 LARGE_CHUNK_REFERENCE_IDS = "large_chunk_reference_ids"
 METADATA = "metadata"
 METADATA_LIST = "metadata_list"

backend/onyx/image_gen/providers/azure_img_gen.py

@@ -12,9 +12,6 @@ if TYPE_CHECKING:
 
 
 class AzureImageGenerationProvider(ImageGenerationProvider):
-    _GPT_IMAGE_MODEL_PREFIX = "gpt-image-"
-    _DALL_E_2_MODEL_NAME = "dall-e-2"
-
     def __init__(
         self,
         api_key: str,
@@ -56,25 +53,6 @@ class AzureImageGenerationProvider(ImageGenerationProvider):
             deployment_name=credentials.deployment_name,
         )
 
-    @property
-    def supports_reference_images(self) -> bool:
-        return True
-
-    @property
-    def max_reference_images(self) -> int:
-        # Azure GPT image models support up to 16 input images for edits.
-        return 16
-
-    def _normalize_model_name(self, model: str) -> str:
-        return model.rsplit("/", 1)[-1]
-
-    def _model_supports_image_edits(self, model: str) -> bool:
-        normalized_model = self._normalize_model_name(model)
-        return (
-            normalized_model.startswith(self._GPT_IMAGE_MODEL_PREFIX)
-            or normalized_model == self._DALL_E_2_MODEL_NAME
-        )
-
     def generate_image(
         self,
         prompt: str,
@@ -82,44 +60,14 @@ class AzureImageGenerationProvider(ImageGenerationProvider):
         size: str,
         n: int,
         quality: str | None = None,
-        reference_images: list[ReferenceImage] | None = None,
+        reference_images: list[ReferenceImage] | None = None,  # noqa: ARG002
         **kwargs: Any,
     ) -> ImageGenerationResponse:
+        from litellm import image_generation
+
         deployment = self._deployment_name or model
         model_name = f"azure/{deployment}"
 
-        if reference_images:
-            if not self._model_supports_image_edits(model):
-                raise ValueError(
-                    f"Model '{model}' does not support image edits with reference images."
-                )
-
-            normalized_model = self._normalize_model_name(model)
-            if (
-                normalized_model == self._DALL_E_2_MODEL_NAME
-                and len(reference_images) > 1
-            ):
-                raise ValueError(
-                    "Model 'dall-e-2' only supports a single reference image for edits."
-                )
-
-            from litellm import image_edit
-
-            return image_edit(
-                image=[image.data for image in reference_images],
-                prompt=prompt,
-                model=model_name,
-                api_key=self._api_key,
-                api_base=self._api_base,
-                api_version=self._api_version,
-                size=size,
-                n=n,
-                quality=quality,
-                **kwargs,
-            )
-
-        from litellm import image_generation
-
         return image_generation(
             prompt=prompt,
             model=model_name,

backend/onyx/image_gen/providers/openai_img_gen.py

@@ -12,9 +12,6 @@ if TYPE_CHECKING:
 
 
 class OpenAIImageGenerationProvider(ImageGenerationProvider):
-    _GPT_IMAGE_MODEL_PREFIX = "gpt-image-"
-    _DALL_E_2_MODEL_NAME = "dall-e-2"
-
     def __init__(
         self,
         api_key: str,
@@ -42,25 +39,6 @@ class OpenAIImageGenerationProvider(ImageGenerationProvider):
             api_base=credentials.api_base,
         )
 
-    @property
-    def supports_reference_images(self) -> bool:
-        return True
-
-    @property
-    def max_reference_images(self) -> int:
-        # GPT image models support up to 16 input images for edits.
-        return 16
-
-    def _normalize_model_name(self, model: str) -> str:
-        return model.rsplit("/", 1)[-1]
-
-    def _model_supports_image_edits(self, model: str) -> bool:
-        normalized_model = self._normalize_model_name(model)
-        return (
-            normalized_model.startswith(self._GPT_IMAGE_MODEL_PREFIX)
-            or normalized_model == self._DALL_E_2_MODEL_NAME
-        )
-
     def generate_image(
         self,
         prompt: str,
@@ -68,38 +46,9 @@ class OpenAIImageGenerationProvider(ImageGenerationProvider):
         size: str,
         n: int,
         quality: str | None = None,
-        reference_images: list[ReferenceImage] | None = None,
+        reference_images: list[ReferenceImage] | None = None,  # noqa: ARG002
         **kwargs: Any,
     ) -> ImageGenerationResponse:
-        if reference_images:
-            if not self._model_supports_image_edits(model):
-                raise ValueError(
-                    f"Model '{model}' does not support image edits with reference images."
-                )
-
-            normalized_model = self._normalize_model_name(model)
-            if (
-                normalized_model == self._DALL_E_2_MODEL_NAME
-                and len(reference_images) > 1
-            ):
-                raise ValueError(
-                    "Model 'dall-e-2' only supports a single reference image for edits."
-                )
-
-            from litellm import image_edit
-
-            return image_edit(
-                image=[image.data for image in reference_images],
-                prompt=prompt,
-                model=model,
-                api_key=self._api_key,
-                api_base=self._api_base,
-                size=size,
-                n=n,
-                quality=quality,
-                **kwargs,
-            )
-
         from litellm import image_generation
 
         return image_generation(

backend/onyx/indexing/adapters/document_indexing_adapter.py

@@ -146,7 +146,6 @@ class DocumentIndexingBatchAdapter:
                     doc_id_to_document_set.get(chunk.source_document.id, [])
                 ),
                 user_project=[],
-                personas=[],
                 boost=(
                     context.id_to_boost_map[chunk.source_document.id]
                     if chunk.source_document.id in context.id_to_boost_map

backend/onyx/indexing/adapters/user_file_indexing_adapter.py

@@ -20,7 +20,6 @@ from onyx.db.models import Persona
 from onyx.db.models import UserFile
 from onyx.db.notification import create_notification
 from onyx.db.user_file import fetch_chunk_counts_for_user_files
-from onyx.db.user_file import fetch_persona_ids_for_user_files
 from onyx.db.user_file import fetch_user_project_ids_for_user_files
 from onyx.file_store.utils import store_user_file_plaintext
 from onyx.indexing.indexing_pipeline import DocumentBatchPrepareContext
@@ -120,10 +119,6 @@ class UserFileIndexingAdapter:
             user_file_ids=updatable_ids,
             db_session=self.db_session,
         )
-        user_file_id_to_persona_ids = fetch_persona_ids_for_user_files(
-            user_file_ids=updatable_ids,
-            db_session=self.db_session,
-        )
         user_file_id_to_access: dict[str, DocumentAccess] = get_access_for_user_files(
             user_file_ids=updatable_ids,
             db_session=self.db_session,
@@ -187,7 +182,7 @@ class UserFileIndexingAdapter:
                 user_project=user_file_id_to_project_ids.get(
                     chunk.source_document.id, []
                 ),
-                personas=user_file_id_to_persona_ids.get(chunk.source_document.id, []),
+                # we are going to index userfiles only once, so we just set the boost to the default
                 boost=DEFAULT_BOOST,
                 tenant_id=tenant_id,
                 aggregated_chunk_boost_factor=chunk_content_scores[chunk_num],

backend/onyx/indexing/indexing_pipeline.py

@@ -49,7 +49,6 @@ from onyx.indexing.embedder import IndexingEmbedder
 from onyx.indexing.models import DocAwareChunk
 from onyx.indexing.models import IndexingBatchAdapter
 from onyx.indexing.models import UpdatableChunkData
-from onyx.indexing.postgres_sanitization import sanitize_documents_for_postgres
 from onyx.indexing.vector_db_insertion import write_chunks_to_vector_db_with_backoff
 from onyx.llm.factory import get_default_llm_with_vision
 from onyx.llm.factory import get_llm_for_contextual_rag
@@ -229,8 +228,6 @@ def index_doc_batch_prepare(
 ) -> DocumentBatchPrepareContext | None:
     """Sets up the documents in the relational DB (source of truth) for permissions, metadata, etc.
     This preceeds indexing it into the actual document index."""
-    documents = sanitize_documents_for_postgres(documents)
-
     # Create a trimmed list of docs that don't have a newer updated at
     # Shortcuts the time-consuming flow on connector index retries
     document_ids: list[str] = [document.id for document in documents]

backend/onyx/indexing/models.py

@@ -112,7 +112,6 @@ class DocMetadataAwareIndexChunk(IndexChunk):
     access: "DocumentAccess"
     document_sets: set[str]
     user_project: list[int]
-    personas: list[int]
     boost: int
     aggregated_chunk_boost_factor: float
     # Full ancestor path from root hierarchy node to document's parent.
@@ -127,7 +126,6 @@ class DocMetadataAwareIndexChunk(IndexChunk):
         access: "DocumentAccess",
         document_sets: set[str],
         user_project: list[int],
-        personas: list[int],
         boost: int,
         aggregated_chunk_boost_factor: float,
         tenant_id: str,
@@ -139,7 +137,6 @@ class DocMetadataAwareIndexChunk(IndexChunk):
             access=access,
             document_sets=document_sets,
             user_project=user_project,
-            personas=personas,
             boost=boost,
             aggregated_chunk_boost_factor=aggregated_chunk_boost_factor,
             tenant_id=tenant_id,

backend/onyx/indexing/postgres_sanitization.py

@@ -1,150 +0,0 @@
-from typing import Any
-
-from onyx.access.models import ExternalAccess
-from onyx.connectors.models import BasicExpertInfo
-from onyx.connectors.models import Document
-from onyx.connectors.models import HierarchyNode
-
-
-def _sanitize_string(value: str) -> str:
-    return value.replace("\x00", "")
-
-
-def _sanitize_json_like(value: Any) -> Any:
-    if isinstance(value, str):
-        return _sanitize_string(value)
-
-    if isinstance(value, list):
-        return [_sanitize_json_like(item) for item in value]
-
-    if isinstance(value, tuple):
-        return tuple(_sanitize_json_like(item) for item in value)
-
-    if isinstance(value, dict):
-        sanitized: dict[Any, Any] = {}
-        for key, nested_value in value.items():
-            cleaned_key = _sanitize_string(key) if isinstance(key, str) else key
-            sanitized[cleaned_key] = _sanitize_json_like(nested_value)
-        return sanitized
-
-    return value
-
-
-def _sanitize_expert_info(expert: BasicExpertInfo) -> BasicExpertInfo:
-    return expert.model_copy(
-        update={
-            "display_name": (
-                _sanitize_string(expert.display_name)
-                if expert.display_name is not None
-                else None
-            ),
-            "first_name": (
-                _sanitize_string(expert.first_name)
-                if expert.first_name is not None
-                else None
-            ),
-            "middle_initial": (
-                _sanitize_string(expert.middle_initial)
-                if expert.middle_initial is not None
-                else None
-            ),
-            "last_name": (
-                _sanitize_string(expert.last_name)
-                if expert.last_name is not None
-                else None
-            ),
-            "email": (
-                _sanitize_string(expert.email) if expert.email is not None else None
-            ),
-        }
-    )
-
-
-def _sanitize_external_access(external_access: ExternalAccess) -> ExternalAccess:
-    return ExternalAccess(
-        external_user_emails={
-            _sanitize_string(email) for email in external_access.external_user_emails
-        },
-        external_user_group_ids={
-            _sanitize_string(group_id)
-            for group_id in external_access.external_user_group_ids
-        },
-        is_public=external_access.is_public,
-    )
-
-
-def sanitize_document_for_postgres(document: Document) -> Document:
-    cleaned_doc = document.model_copy(deep=True)
-
-    cleaned_doc.id = _sanitize_string(cleaned_doc.id)
-    cleaned_doc.semantic_identifier = _sanitize_string(cleaned_doc.semantic_identifier)
-    if cleaned_doc.title is not None:
-        cleaned_doc.title = _sanitize_string(cleaned_doc.title)
-    if cleaned_doc.parent_hierarchy_raw_node_id is not None:
-        cleaned_doc.parent_hierarchy_raw_node_id = _sanitize_string(
-            cleaned_doc.parent_hierarchy_raw_node_id
-        )
-
-    cleaned_doc.metadata = {
-        _sanitize_string(key): (
-            [_sanitize_string(item) for item in value]
-            if isinstance(value, list)
-            else _sanitize_string(value)
-        )
-        for key, value in cleaned_doc.metadata.items()
-    }
-
-    if cleaned_doc.doc_metadata is not None:
-        cleaned_doc.doc_metadata = _sanitize_json_like(cleaned_doc.doc_metadata)
-
-    if cleaned_doc.primary_owners is not None:
-        cleaned_doc.primary_owners = [
-            _sanitize_expert_info(expert) for expert in cleaned_doc.primary_owners
-        ]
-    if cleaned_doc.secondary_owners is not None:
-        cleaned_doc.secondary_owners = [
-            _sanitize_expert_info(expert) for expert in cleaned_doc.secondary_owners
-        ]
-
-    if cleaned_doc.external_access is not None:
-        cleaned_doc.external_access = _sanitize_external_access(
-            cleaned_doc.external_access
-        )
-
-    for section in cleaned_doc.sections:
-        if section.link is not None:
-            section.link = _sanitize_string(section.link)
-        if section.text is not None:
-            section.text = _sanitize_string(section.text)
-        if section.image_file_id is not None:
-            section.image_file_id = _sanitize_string(section.image_file_id)
-
-    return cleaned_doc
-
-
-def sanitize_documents_for_postgres(documents: list[Document]) -> list[Document]:
-    return [sanitize_document_for_postgres(document) for document in documents]
-
-
-def sanitize_hierarchy_node_for_postgres(node: HierarchyNode) -> HierarchyNode:
-    cleaned_node = node.model_copy(deep=True)
-
-    cleaned_node.raw_node_id = _sanitize_string(cleaned_node.raw_node_id)
-    cleaned_node.display_name = _sanitize_string(cleaned_node.display_name)
-    if cleaned_node.raw_parent_id is not None:
-        cleaned_node.raw_parent_id = _sanitize_string(cleaned_node.raw_parent_id)
-    if cleaned_node.link is not None:
-        cleaned_node.link = _sanitize_string(cleaned_node.link)
-
-    if cleaned_node.external_access is not None:
-        cleaned_node.external_access = _sanitize_external_access(
-            cleaned_node.external_access
-        )
-
-    return cleaned_node
-
-
-def sanitize_hierarchy_nodes_for_postgres(
-    nodes: list[HierarchyNode],
-) -> list[HierarchyNode]:
-    return [sanitize_hierarchy_node_for_postgres(node) for node in nodes]

backend/onyx/main.py

@@ -97,9 +97,6 @@ from onyx.server.features.web_search.api import router as web_search_router
 from onyx.server.federated.api import router as federated_router
 from onyx.server.kg.api import admin_router as kg_admin_router
 from onyx.server.manage.administrative import router as admin_router
-from onyx.server.manage.code_interpreter.api import (
-    admin_router as code_interpreter_admin_router,
-)
 from onyx.server.manage.discord_bot.api import router as discord_bot_router
 from onyx.server.manage.embedding.api import admin_router as embedding_admin_router
 from onyx.server.manage.embedding.api import basic_router as embedding_router
@@ -424,9 +421,6 @@ def get_application(lifespan_override: Lifespan | None = None) -> FastAPI:
     include_router_with_global_prefix_prepended(application, llm_admin_router)
     include_router_with_global_prefix_prepended(application, kg_admin_router)
     include_router_with_global_prefix_prepended(application, llm_router)
-    include_router_with_global_prefix_prepended(
-        application, code_interpreter_admin_router
-    )
     include_router_with_global_prefix_prepended(
         application, image_generation_admin_router
     )

backend/onyx/onyxbot/slack/blocks.py

@@ -592,8 +592,11 @@ def build_slack_response_blocks(
         )
 
     citations_blocks = []
+    document_blocks = []
     if answer.citation_info:
         citations_blocks = _build_citations_blocks(answer)
+    else:
+        document_blocks = _priority_ordered_documents_blocks(answer)
 
     citations_divider = [DividerBlock()] if citations_blocks else []
     buttons_divider = [DividerBlock()] if web_follow_up_block or follow_up_block else []
@@ -605,6 +608,7 @@ def build_slack_response_blocks(
         + ai_feedback_block
         + citations_divider
         + citations_blocks
+        + document_blocks
         + buttons_divider
         + web_follow_up_block
         + follow_up_block

backend/onyx/onyxbot/slack/formatting.py

@@ -1,149 +1,20 @@
-import re
-from collections.abc import Callable
 from typing import Any
 
 from mistune import create_markdown
 from mistune import HTMLRenderer
 
-# Tags that should be replaced with a newline (line-break and block-level elements)
-_HTML_NEWLINE_TAG_PATTERN = re.compile(
-    r"<br\s*/?>|</(?:p|div|li|h[1-6]|tr|blockquote|section|article)>",
-    re.IGNORECASE,
-)
-
-# Strips HTML tags but excludes autolinks like <https://...> and <mailto:...>
-_HTML_TAG_PATTERN = re.compile(
-    r"<(?!https?://|mailto:)/?[a-zA-Z][^>]*>",
-)
-
-# Matches fenced code blocks (``` ... ```) so we can skip sanitization inside them
-_FENCED_CODE_BLOCK_PATTERN = re.compile(r"```[\s\S]*?```")
-
-# Matches the start of any markdown link: [text]( or [[n]](
-# The inner group handles nested brackets for citation links like [[1]](.
-_MARKDOWN_LINK_PATTERN = re.compile(r"\[(?:[^\[\]]|\[[^\]]*\])*\]\(")
-
-# Matches Slack-style links <url|text> that LLMs sometimes output directly.
-# Mistune doesn't recognise this syntax, so text() would escape the angle
-# brackets and Slack would render them as literal text instead of links.
-_SLACK_LINK_PATTERN = re.compile(r"<(https?://[^|>]+)\|([^>]+)>")
-
-
-def _sanitize_html(text: str) -> str:
-    """Strip HTML tags from a text fragment.
-
-    Block-level closing tags and <br> are converted to newlines.
-    All other HTML tags are removed. Autolinks (<https://...>) are preserved.
-    """
-    text = _HTML_NEWLINE_TAG_PATTERN.sub("\n", text)
-    text = _HTML_TAG_PATTERN.sub("", text)
-    return text
-
-
-def _transform_outside_code_blocks(
-    message: str, transform: Callable[[str], str]
-) -> str:
-    """Apply *transform* only to text outside fenced code blocks."""
-    parts = _FENCED_CODE_BLOCK_PATTERN.split(message)
-    code_blocks = _FENCED_CODE_BLOCK_PATTERN.findall(message)
-
-    result: list[str] = []
-    for i, part in enumerate(parts):
-        result.append(transform(part))
-        if i < len(code_blocks):
-            result.append(code_blocks[i])
-
-    return "".join(result)
-
-
-def _extract_link_destination(message: str, start_idx: int) -> tuple[str, int | None]:
-    """Extract markdown link destination, allowing nested parentheses in the URL."""
-    depth = 0
-    i = start_idx
-
-    while i < len(message):
-        curr = message[i]
-        if curr == "\\":
-            i += 2
-            continue
-
-        if curr == "(":
-            depth += 1
-        elif curr == ")":
-            if depth == 0:
-                return message[start_idx:i], i
-            depth -= 1
-        i += 1
-
-    return message[start_idx:], None
-
-
-def _normalize_link_destinations(message: str) -> str:
-    """Wrap markdown link URLs in angle brackets so the parser handles special chars safely.
-
-    Markdown link syntax [text](url) breaks when the URL contains unescaped
-    parentheses, spaces, or other special characters. Wrapping the URL in angle
-    brackets — [text](<url>) — tells the parser to treat everything inside as
-    a literal URL. This applies to all links, not just citations.
-    """
-    if "](" not in message:
-        return message
-
-    normalized_parts: list[str] = []
-    cursor = 0
-
-    while match := _MARKDOWN_LINK_PATTERN.search(message, cursor):
-        normalized_parts.append(message[cursor : match.end()])
-        destination_start = match.end()
-        destination, end_idx = _extract_link_destination(message, destination_start)
-        if end_idx is None:
-            normalized_parts.append(message[destination_start:])
-            return "".join(normalized_parts)
-
-        already_wrapped = destination.startswith("<") and destination.endswith(">")
-        if destination and not already_wrapped:
-            destination = f"<{destination}>"
-
-        normalized_parts.append(destination)
-        normalized_parts.append(")")
-        cursor = end_idx + 1
-
-    normalized_parts.append(message[cursor:])
-    return "".join(normalized_parts)
-
-
-def _convert_slack_links_to_markdown(message: str) -> str:
-    """Convert Slack-style <url|text> links to standard markdown [text](url).
-
-    LLMs sometimes emit Slack mrkdwn link syntax directly. Mistune doesn't
-    recognise it, so the angle brackets would be escaped by text() and Slack
-    would render the link as literal text instead of a clickable link.
-    """
-    return _transform_outside_code_blocks(
-        message, lambda text: _SLACK_LINK_PATTERN.sub(r"[\2](\1)", text)
-    )
-
 
 def format_slack_message(message: str | None) -> str:
     if message is None:
         return ""
-    message = _transform_outside_code_blocks(message, _sanitize_html)
-    message = _convert_slack_links_to_markdown(message)
-    normalized_message = _normalize_link_destinations(message)
     md = create_markdown(renderer=SlackRenderer(), plugins=["strikethrough"])
-    result = md(normalized_message)
+    result = md(message)
     # With HTMLRenderer, result is always str (not AST list)
     assert isinstance(result, str)
-    return result.rstrip("\n")
+    return result
 
 
 class SlackRenderer(HTMLRenderer):
-    """Renders markdown as Slack mrkdwn format instead of HTML.
-
-    Overrides all HTMLRenderer methods that produce HTML tags to ensure
-    no raw HTML ever appears in Slack messages.
-    """
-
     SPECIALS: dict[str, str] = {"&": "&amp;", "<": "&lt;", ">": "&gt;"}
 
     def escape_special(self, text: str) -> str:
@@ -152,7 +23,7 @@ class SlackRenderer(HTMLRenderer):
         return text
 
     def heading(self, text: str, level: int, **attrs: Any) -> str:  # noqa: ARG002
-        return f"*{text}*\n\n"
+        return f"*{text}*\n"
 
     def emphasis(self, text: str) -> str:
         return f"_{text}_"
@@ -171,7 +42,7 @@ class SlackRenderer(HTMLRenderer):
                 count += 1
                 prefix = f"{count}. " if ordered else "• "
                 lines[i] = f"{prefix}{line[4:]}"
-        return "\n".join(lines) + "\n"
+        return "\n".join(lines)
 
     def list_item(self, text: str) -> str:
         return f"li: {text}\n"
@@ -193,30 +64,7 @@ class SlackRenderer(HTMLRenderer):
         return f"`{text}`"
 
     def block_code(self, code: str, info: str | None = None) -> str:  # noqa: ARG002
-        return f"```\n{code.rstrip(chr(10))}\n```\n\n"
-
-    def linebreak(self) -> str:
-        return "\n"
-
-    def thematic_break(self) -> str:
-        return "---\n\n"
-
-    def block_quote(self, text: str) -> str:
-        lines = text.strip().split("\n")
-        quoted = "\n".join(f">{line}" for line in lines)
-        return quoted + "\n\n"
-
-    def block_html(self, html: str) -> str:
-        return _sanitize_html(html) + "\n\n"
-
-    def block_error(self, text: str) -> str:
-        return f"```\n{text}\n```\n\n"
-
-    def text(self, text: str) -> str:
-        # Only escape the three entities Slack recognizes: & < >
-        # HTMLRenderer.text() also escapes " to &quot; which Slack renders
-        # as literal &quot; text since Slack doesn't recognize that entity.
-        return self.escape_special(text)
+        return f"```\n{code}\n```\n"
 
     def paragraph(self, text: str) -> str:
-        return f"{text}\n\n"
+        return f"{text}\n"

backend/onyx/server/features/build/api/sessions_api.py

@@ -762,43 +762,6 @@ def download_webapp(
     )
 
 
-@router.get("/{session_id}/download-directory/{path:path}")
-def download_directory(
-    session_id: UUID,
-    path: str,
-    user: User = Depends(current_user),
-    db_session: Session = Depends(get_session),
-) -> Response:
-    """
-    Download a directory as a zip file.
-
-    Returns the specified directory as a zip archive.
-    """
-    user_id: UUID = user.id
-    session_manager = SessionManager(db_session)
-
-    try:
-        result = session_manager.download_directory(session_id, user_id, path)
-    except ValueError as e:
-        error_message = str(e)
-        if "path traversal" in error_message.lower():
-            raise HTTPException(status_code=403, detail="Access denied")
-        raise HTTPException(status_code=400, detail=error_message)
-
-    if result is None:
-        raise HTTPException(status_code=404, detail="Directory not found")
-
-    zip_bytes, filename = result
-
-    return Response(
-        content=zip_bytes,
-        media_type="application/zip",
-        headers={
-            "Content-Disposition": f'attachment; filename="{filename}"',
-        },
-    )
-
-
 @router.post("/{session_id}/upload", response_model=UploadResponse)
 def upload_file_endpoint(
     session_id: UUID,

backend/onyx/server/features/build/db/user_library.py

@@ -107,23 +107,27 @@ def get_or_create_craft_connector(db_session: Session, user: User) -> tuple[int,
     )
 
     for cc_pair in cc_pairs:
-        if (
-            cc_pair.connector.source == DocumentSource.CRAFT_FILE
-            and cc_pair.creator_id == user.id
-        ):
+        if cc_pair.connector.source == DocumentSource.CRAFT_FILE:
             return cc_pair.connector.id, cc_pair.credential.id
 
-    # No cc_pair for this user — find or create the shared CRAFT_FILE connector
+    # Check for orphaned connector (created but cc_pair creation failed previously)
     existing_connectors = fetch_connectors(
         db_session, sources=[DocumentSource.CRAFT_FILE]
     )
-    connector_id: int | None = None
+    orphaned_connector = None
     for conn in existing_connectors:
-        if conn.name == USER_LIBRARY_CONNECTOR_NAME:
-            connector_id = conn.id
+        if conn.name != USER_LIBRARY_CONNECTOR_NAME:
+            continue
+        if not conn.credentials:
+            orphaned_connector = conn
             break
 
-    if connector_id is None:
+    if orphaned_connector:
+        connector_id = orphaned_connector.id
+        logger.info(
+            f"Found orphaned User Library connector {connector_id}, completing setup"
+        )
+    else:
         connector_data = ConnectorBase(
             name=USER_LIBRARY_CONNECTOR_NAME,
             source=DocumentSource.CRAFT_FILE,

backend/onyx/server/features/build/sandbox/kubernetes/docker/generate_agents_md.py

@@ -1,19 +1,15 @@
 #!/usr/bin/env python3
 """Generate AGENTS.md by scanning the files directory and populating the template.
 
-This script runs during session setup, AFTER files have been synced from S3
-and the files symlink has been created. It reads an existing AGENTS.md (which
-contains the {{KNOWLEDGE_SOURCES_SECTION}} placeholder), replaces the
-placeholder by scanning the knowledge source directory, and writes it back.
+This script runs at container startup, AFTER the init container has synced files
+from S3. It scans the /workspace/files directory to discover what knowledge sources
+are available and generates appropriate documentation.
 
-Usage:
-    python3 generate_agents_md.py <agents_md_path> <files_path>
-
-Arguments:
-    agents_md_path: Path to the AGENTS.md file to update in place
-    files_path: Path to the files directory to scan for knowledge sources
+Environment variables:
+- AGENT_INSTRUCTIONS: The template content with placeholders to replace
 """
 
+import os
 import sys
 from pathlib import Path
 
@@ -193,39 +189,49 @@ def build_knowledge_sources_section(files_path: Path) -> str:
 def main() -> None:
     """Main entry point for container startup script.
 
-    Reads an existing AGENTS.md, replaces the {{KNOWLEDGE_SOURCES_SECTION}}
-    placeholder by scanning the files directory, and writes it back.
-
-    Usage:
-        python3 generate_agents_md.py <agents_md_path> <files_path>
+    Is called by the container startup script to scan /workspace/files and populate
+    the knowledge sources section.
     """
-    if len(sys.argv) != 3:
-        print(
-            f"Usage: {sys.argv[0]} <agents_md_path> <files_path>",
-            file=sys.stderr,
-        )
-        sys.exit(1)
+    # Read template from environment variable
+    template = os.environ.get("AGENT_INSTRUCTIONS", "")
+    if not template:
+        print("Warning: No AGENT_INSTRUCTIONS template provided", file=sys.stderr)
+        template = "# Agent Instructions\n\nNo instructions provided."
 
-    agents_md_path = Path(sys.argv[1])
-    files_path = Path(sys.argv[2])
+    # Scan files directory - check /workspace/files first, then /workspace/demo_data
+    files_path = Path("/workspace/files")
+    demo_data_path = Path("/workspace/demo_data")
 
-    if not agents_md_path.exists():
-        print(f"Error: {agents_md_path} not found", file=sys.stderr)
-        sys.exit(1)
+    # Use demo_data if files doesn't exist or is empty
+    if not files_path.exists() or not any(files_path.iterdir()):
+        if demo_data_path.exists():
+            files_path = demo_data_path
 
-    template = agents_md_path.read_text()
+    knowledge_sources_section = build_knowledge_sources_section(files_path)
 
-    # Resolve symlinks (handles both direct symlinks and dirs containing symlinks)
-    resolved_files_path = files_path.resolve()
-
-    knowledge_sources_section = build_knowledge_sources_section(resolved_files_path)
-
-    # Replace placeholder and write back
-    content = template.replace(
+    # Replace placeholders
+    content = template
+    content = content.replace(
         "{{KNOWLEDGE_SOURCES_SECTION}}", knowledge_sources_section
     )
-    agents_md_path.write_text(content)
-    print(f"Populated knowledge sources in {agents_md_path}")
+
+    # Write AGENTS.md
+    output_path = Path("/workspace/AGENTS.md")
+    output_path.write_text(content)
+
+    # Log result
+    source_count = 0
+    if files_path.exists():
+        source_count = len(
+            [
+                d
+                for d in files_path.iterdir()
+                if d.is_dir() and not d.name.startswith(".")
+            ]
+        )
+    print(
+        f"Generated AGENTS.md with {source_count} knowledge sources from {files_path}"
+    )
 
 
 if __name__ == "__main__":

backend/onyx/server/features/build/sandbox/kubernetes/kubernetes_sandbox_manager.py

@@ -1352,9 +1352,6 @@ fi
 echo "Writing AGENTS.md"
 printf '%s' '{agent_instructions_escaped}' > {session_path}/AGENTS.md
 
-# Populate knowledge sources by scanning the files directory
-python3 /usr/local/bin/generate_agents_md.py {session_path}/AGENTS.md {session_path}/files || true
-
 # Write opencode config
 echo "Writing opencode.json"
 printf '%s' '{opencode_json_escaped}' > {session_path}/opencode.json
@@ -1783,9 +1780,6 @@ ln -sf {symlink_target} {session_path}/files
 echo "Writing AGENTS.md"
 printf '%s' '{agent_instructions_escaped}' > {session_path}/AGENTS.md
 
-# Populate knowledge sources by scanning the files directory
-python3 /usr/local/bin/generate_agents_md.py {session_path}/AGENTS.md {session_path}/files || true
-
 # Write opencode config
 echo "Writing opencode.json"
 printf '%s' '{opencode_json_escaped}' > {session_path}/opencode.json

backend/onyx/server/features/build/session/manager.py

@@ -68,7 +68,6 @@ from onyx.server.features.build.db.sandbox import create_sandbox__no_commit
 from onyx.server.features.build.db.sandbox import get_running_sandbox_count_by_tenant
 from onyx.server.features.build.db.sandbox import get_sandbox_by_session_id
 from onyx.server.features.build.db.sandbox import get_sandbox_by_user_id
-from onyx.server.features.build.db.sandbox import get_snapshots_for_session
 from onyx.server.features.build.db.sandbox import update_sandbox_heartbeat
 from onyx.server.features.build.db.sandbox import update_sandbox_status__no_commit
 from onyx.server.features.build.sandbox import get_sandbox_manager
@@ -647,30 +646,16 @@ class SessionManager:
 
             if sandbox and sandbox.status.is_active():
                 # Quick health check to verify sandbox is actually responsive
-                # AND verify the session workspace still exists on disk
-                # (it may have been wiped if the sandbox was re-provisioned)
-                is_healthy = self._sandbox_manager.health_check(sandbox.id, timeout=5.0)
-                workspace_exists = (
-                    is_healthy
-                    and self._sandbox_manager.session_workspace_exists(
-                        sandbox.id, existing.id
-                    )
-                )
-                if is_healthy and workspace_exists:
+                if self._sandbox_manager.health_check(sandbox.id, timeout=5.0):
                     logger.info(
                         f"Returning existing empty session {existing.id} for user {user_id}"
                     )
                     return existing
-                elif not is_healthy:
+                else:
                     logger.warning(
                         f"Empty session {existing.id} has unhealthy sandbox {sandbox.id}. "
                         f"Deleting and creating fresh session."
                     )
-                else:
-                    logger.warning(
-                        f"Empty session {existing.id} workspace missing in sandbox "
-                        f"{sandbox.id}. Deleting and creating fresh session."
-                    )
             else:
                 logger.warning(
                     f"Empty session {existing.id} has no active sandbox "
@@ -1050,23 +1035,6 @@ class SessionManager:
                 # workspace cleanup fails (e.g., if pod is already terminated)
                 logger.warning(f"Failed to cleanup session workspace {session_id}: {e}")
 
-        # Delete snapshot files from S3 before removing DB records
-        snapshots = get_snapshots_for_session(self._db_session, session_id)
-        if snapshots:
-            from onyx.file_store.file_store import get_default_file_store
-            from onyx.server.features.build.sandbox.manager.snapshot_manager import (
-                SnapshotManager,
-            )
-
-            snapshot_manager = SnapshotManager(get_default_file_store())
-            for snapshot in snapshots:
-                try:
-                    snapshot_manager.delete_snapshot(snapshot.storage_path)
-                except Exception as e:
-                    logger.warning(
-                        f"Failed to delete snapshot file {snapshot.storage_path}: {e}"
-                    )
-
         # Delete session (uses flush, caller commits)
         return delete_build_session__no_commit(session_id, user_id, self._db_session)
 
@@ -1935,94 +1903,6 @@ class SessionManager:
 
         return zip_buffer.getvalue(), filename
 
-    def download_directory(
-        self,
-        session_id: UUID,
-        user_id: UUID,
-        path: str,
-    ) -> tuple[bytes, str] | None:
-        """
-        Create a zip file of an arbitrary directory in the session workspace.
-
-        Args:
-            session_id: The session UUID
-            user_id: The user ID to verify ownership
-            path: Relative path to the directory (within session workspace)
-
-        Returns:
-            Tuple of (zip_bytes, filename) or None if session not found
-
-        Raises:
-            ValueError: If path traversal attempted or path is not a directory
-        """
-        # Verify session ownership
-        session = get_build_session(session_id, user_id, self._db_session)
-        if session is None:
-            return None
-
-        sandbox = get_sandbox_by_user_id(self._db_session, user_id)
-        if sandbox is None:
-            return None
-
-        # Check if directory exists
-        try:
-            self._sandbox_manager.list_directory(
-                sandbox_id=sandbox.id,
-                session_id=session_id,
-                path=path,
-            )
-        except ValueError:
-            return None
-
-        # Recursively collect all files
-        def collect_files(dir_path: str) -> list[tuple[str, str]]:
-            """Collect all files recursively, returning (full_path, arcname) tuples."""
-            files: list[tuple[str, str]] = []
-            try:
-                entries = self._sandbox_manager.list_directory(
-                    sandbox_id=sandbox.id,
-                    session_id=session_id,
-                    path=dir_path,
-                )
-                for entry in entries:
-                    if entry.is_directory:
-                        files.extend(collect_files(entry.path))
-                    else:
-                        # arcname is relative to the target directory
-                        prefix_len = len(path) + 1  # +1 for trailing slash
-                        arcname = entry.path[prefix_len:]
-                        files.append((entry.path, arcname))
-            except ValueError:
-                pass
-            return files
-
-        file_list = collect_files(path)
-
-        # Create zip file in memory
-        zip_buffer = io.BytesIO()
-        with zipfile.ZipFile(zip_buffer, "w", zipfile.ZIP_DEFLATED) as zip_file:
-            for full_path, arcname in file_list:
-                try:
-                    content = self._sandbox_manager.read_file(
-                        sandbox_id=sandbox.id,
-                        session_id=session_id,
-                        path=full_path,
-                    )
-                    zip_file.writestr(arcname, content)
-                except ValueError:
-                    pass
-
-        zip_buffer.seek(0)
-
-        # Use the directory name for the zip filename
-        dir_name = Path(path).name
-        safe_name = "".join(
-            c if c.isalnum() or c in ("-", "_", ".") else "_" for c in dir_name
-        )
-        filename = f"{safe_name}.zip"
-
-        return zip_buffer.getvalue(), filename
-
     # =========================================================================
     # File System Operations
     # =========================================================================
@@ -2057,18 +1937,11 @@ class SessionManager:
             return None
 
         # Use sandbox manager to list directory (works for both local and K8s)
-        # If the directory doesn't exist (e.g., session workspace not yet loaded),
-        # return an empty listing rather than erroring out.
-        try:
-            raw_entries = self._sandbox_manager.list_directory(
-                sandbox_id=sandbox.id,
-                session_id=session_id,
-                path=path,
-            )
-        except ValueError as e:
-            if "path traversal" in str(e).lower():
-                raise
-            return DirectoryListing(path=path, entries=[])
+        raw_entries = self._sandbox_manager.list_directory(
+            sandbox_id=sandbox.id,
+            session_id=session_id,
+            path=path,
+        )
 
         # Filter hidden files and directories
         entries: list[FileSystemEntry] = [

backend/onyx/server/features/projects/api.py

@@ -12,18 +12,11 @@ from pydantic import BaseModel
 from sqlalchemy.orm import Session
 
 from onyx.auth.users import current_user
-from onyx.background.celery.tasks.user_file_processing.tasks import (
-    enqueue_user_file_project_sync_task,
-)
-from onyx.background.celery.tasks.user_file_processing.tasks import (
-    get_user_file_project_sync_queue_depth,
-)
 from onyx.background.celery.versioned_apps.client import app as client_app
 from onyx.configs.constants import OnyxCeleryPriority
 from onyx.configs.constants import OnyxCeleryQueues
 from onyx.configs.constants import OnyxCeleryTask
 from onyx.configs.constants import PUBLIC_API_TAGS
-from onyx.configs.constants import USER_FILE_PROJECT_SYNC_MAX_QUEUE_DEPTH
 from onyx.db.engine.sql_engine import get_session
 from onyx.db.enums import UserFileStatus
 from onyx.db.models import ChatSession
@@ -34,7 +27,6 @@ from onyx.db.models import UserProject
 from onyx.db.persona import get_personas_by_ids
 from onyx.db.projects import get_project_token_count
 from onyx.db.projects import upload_files_to_user_files_with_indexing
-from onyx.redis.redis_pool import get_redis_client
 from onyx.server.features.projects.models import CategorizedFilesSnapshot
 from onyx.server.features.projects.models import ChatSessionRequest
 from onyx.server.features.projects.models import TokenCountResponse
@@ -55,33 +47,6 @@ class UserFileDeleteResult(BaseModel):
     assistant_names: list[str] = []
 
 
-def _trigger_user_file_project_sync(user_file_id: UUID, tenant_id: str) -> None:
-    queue_depth = get_user_file_project_sync_queue_depth(client_app)
-    if queue_depth > USER_FILE_PROJECT_SYNC_MAX_QUEUE_DEPTH:
-        logger.warning(
-            f"Skipping immediate project sync for user_file_id={user_file_id} due to "
-            f"queue depth {queue_depth}>{USER_FILE_PROJECT_SYNC_MAX_QUEUE_DEPTH}. "
-            "It will be picked up by beat later."
-        )
-        return
-
-    redis_client = get_redis_client(tenant_id=tenant_id)
-    enqueued = enqueue_user_file_project_sync_task(
-        celery_app=client_app,
-        redis_client=redis_client,
-        user_file_id=user_file_id,
-        tenant_id=tenant_id,
-        priority=OnyxCeleryPriority.HIGHEST,
-    )
-    if not enqueued:
-        logger.info(
-            f"Skipped duplicate project sync enqueue for user_file_id={user_file_id}"
-        )
-        return
-
-    logger.info(f"Triggered project sync for user_file_id={user_file_id}")
-
-
 @router.get("", tags=PUBLIC_API_TAGS)
 def get_projects(
     user: User = Depends(current_user),
@@ -224,7 +189,15 @@ def unlink_user_file_from_project(
         db_session.commit()
 
     tenant_id = get_current_tenant_id()
-    _trigger_user_file_project_sync(user_file.id, tenant_id)
+    task = client_app.send_task(
+        OnyxCeleryTask.PROCESS_SINGLE_USER_FILE_PROJECT_SYNC,
+        kwargs={"user_file_id": user_file.id, "tenant_id": tenant_id},
+        queue=OnyxCeleryQueues.USER_FILE_PROJECT_SYNC,
+        priority=OnyxCeleryPriority.HIGHEST,
+    )
+    logger.info(
+        f"Triggered project sync for user_file_id={user_file.id} with task_id={task.id}"
+    )
 
     return Response(status_code=204)
 
@@ -268,7 +241,15 @@ def link_user_file_to_project(
         db_session.commit()
 
     tenant_id = get_current_tenant_id()
-    _trigger_user_file_project_sync(user_file.id, tenant_id)
+    task = client_app.send_task(
+        OnyxCeleryTask.PROCESS_SINGLE_USER_FILE_PROJECT_SYNC,
+        kwargs={"user_file_id": user_file.id, "tenant_id": tenant_id},
+        queue=OnyxCeleryQueues.USER_FILE_PROJECT_SYNC,
+        priority=OnyxCeleryPriority.HIGHEST,
+    )
+    logger.info(
+        f"Triggered project sync for user_file_id={user_file.id} with task_id={task.id}"
+    )
 
     return UserFileSnapshot.from_model(user_file)
 

backend/onyx/server/manage/code_interpreter/api.py

@@ -1,47 +0,0 @@
-from fastapi import APIRouter
-from fastapi import Depends
-from sqlalchemy.orm import Session
-
-from onyx.auth.users import current_admin_user
-from onyx.db.code_interpreter import fetch_code_interpreter_server
-from onyx.db.code_interpreter import update_code_interpreter_server_enabled
-from onyx.db.engine.sql_engine import get_session
-from onyx.db.models import User
-from onyx.server.manage.code_interpreter.models import CodeInterpreterServer
-from onyx.server.manage.code_interpreter.models import CodeInterpreterServerHealth
-from onyx.tools.tool_implementations.python.code_interpreter_client import (
-    CodeInterpreterClient,
-)
-
-admin_router = APIRouter(prefix="/admin/code-interpreter")
-
-
-@admin_router.get("/health")
-def get_code_interpreter_health(
-    _: User = Depends(current_admin_user),
-) -> CodeInterpreterServerHealth:
-    try:
-        client = CodeInterpreterClient()
-        return CodeInterpreterServerHealth(healthy=client.health())
-    except ValueError:
-        return CodeInterpreterServerHealth(healthy=False)
-
-
-@admin_router.get("")
-def get_code_interpreter(
-    _: User = Depends(current_admin_user), db_session: Session = Depends(get_session)
-) -> CodeInterpreterServer:
-    ci_server = fetch_code_interpreter_server(db_session)
-    return CodeInterpreterServer(enabled=ci_server.server_enabled)
-
-
-@admin_router.put("")
-def update_code_interpreter(
-    update: CodeInterpreterServer,
-    _: User = Depends(current_admin_user),
-    db_session: Session = Depends(get_session),
-) -> None:
-    update_code_interpreter_server_enabled(
-        db_session=db_session,
-        enabled=update.enabled,
-    )

backend/onyx/server/manage/code_interpreter/models.py

@@ -1,9 +0,0 @@
-from pydantic import BaseModel
-
-
-class CodeInterpreterServer(BaseModel):
-    enabled: bool
-
-
-class CodeInterpreterServerHealth(BaseModel):
-    healthy: bool

backend/onyx/server/manage/llm/models.py

@@ -105,9 +105,7 @@ class LLMProviderDescriptor(BaseModel):
         is_default_provider = bool(default_model_name)
         is_default_vision_provider = default_vision_model is not None
 
-        default_model_name = (
-            default_model_name or llm_provider_model.default_model_name or ""
-        )
+        default_model_name = default_model_name or llm_provider_model.default_model_name
 
         return cls(
             name=llm_provider_model.name,
@@ -186,9 +184,7 @@ class LLMProviderView(LLMProvider):
         is_default_provider = bool(default_model_name)
         is_default_vision_provider = default_vision_model is not None
 
-        default_model_name = (
-            default_model_name or llm_provider_model.default_model_name or ""
-        )
+        default_model_name = default_model_name or llm_provider_model.default_model_name
 
         return cls(
             id=llm_provider_model.id,

backend/onyx/server/manage/models.py

@@ -35,18 +35,6 @@ if TYPE_CHECKING:
     pass
 
 
-class EmailInviteStatus(str, Enum):
-    SENT = "SENT"
-    NOT_CONFIGURED = "NOT_CONFIGURED"
-    SEND_FAILED = "SEND_FAILED"
-    DISABLED = "DISABLED"
-
-
-class BulkInviteResponse(BaseModel):
-    invited_count: int
-    email_invite_status: EmailInviteStatus
-
-
 class VersionResponse(BaseModel):
     backend_version: str
 

backend/onyx/server/manage/users.py

@@ -36,7 +36,6 @@ from onyx.configs.app_configs import AUTH_BACKEND
 from onyx.configs.app_configs import AUTH_TYPE
 from onyx.configs.app_configs import AuthBackend
 from onyx.configs.app_configs import DEV_MODE
-from onyx.configs.app_configs import EMAIL_CONFIGURED
 from onyx.configs.app_configs import ENABLE_EMAIL_INVITES
 from onyx.configs.app_configs import NUM_FREE_TRIAL_USER_INVITES
 from onyx.configs.app_configs import REDIS_AUTH_KEY_PREFIX
@@ -79,10 +78,8 @@ from onyx.server.documents.models import PaginatedReturn
 from onyx.server.features.projects.models import UserFileSnapshot
 from onyx.server.manage.models import AllUsersResponse
 from onyx.server.manage.models import AutoScrollRequest
-from onyx.server.manage.models import BulkInviteResponse
 from onyx.server.manage.models import ChatBackgroundRequest
 from onyx.server.manage.models import DefaultAppModeRequest
-from onyx.server.manage.models import EmailInviteStatus
 from onyx.server.manage.models import MemoryItem
 from onyx.server.manage.models import PersonalizationUpdateRequest
 from onyx.server.manage.models import TenantInfo
@@ -371,7 +368,7 @@ def bulk_invite_users(
     emails: list[str] = Body(..., embed=True),
     current_user: User = Depends(current_admin_user),
     db_session: Session = Depends(get_session),
-) -> BulkInviteResponse:
+) -> int:
     """emails are string validated. If any email fails validation, no emails are
     invited and an exception is raised."""
     tenant_id = get_current_tenant_id()
@@ -430,41 +427,34 @@ def bulk_invite_users(
     number_of_invited_users = write_invited_users(all_emails)
 
     # send out email invitations only to new users (not already invited or existing)
-    if not ENABLE_EMAIL_INVITES:
-        email_invite_status = EmailInviteStatus.DISABLED
-    elif not EMAIL_CONFIGURED:
-        email_invite_status = EmailInviteStatus.NOT_CONFIGURED
-    else:
+    if ENABLE_EMAIL_INVITES:
         try:
             for email in emails_needing_seats:
                 send_user_email_invite(email, current_user, AUTH_TYPE)
-            email_invite_status = EmailInviteStatus.SENT
         except Exception as e:
             logger.error(f"Error sending email invite to invited users: {e}")
-            email_invite_status = EmailInviteStatus.SEND_FAILED
 
-    if MULTI_TENANT and not DEV_MODE:
-        # for billing purposes, write to the control plane about the number of new users
-        try:
-            logger.info("Registering tenant users")
-            fetch_ee_implementation_or_noop(
-                "onyx.server.tenants.billing", "register_tenant_users", None
-            )(tenant_id, get_live_users_count(db_session))
-        except Exception as e:
-            logger.error(f"Failed to register tenant users: {str(e)}")
-            logger.info(
-                "Reverting changes: removing users from tenant and resetting invited users"
-            )
-            write_invited_users(initial_invited_users)  # Reset to original state
-            fetch_ee_implementation_or_noop(
-                "onyx.server.tenants.user_mapping", "remove_users_from_tenant", None
-            )(new_invited_emails, tenant_id)
-            raise e
+    if not MULTI_TENANT or DEV_MODE:
+        return number_of_invited_users
 
-    return BulkInviteResponse(
-        invited_count=number_of_invited_users,
-        email_invite_status=email_invite_status,
-    )
+    # for billing purposes, write to the control plane about the number of new users
+    try:
+        logger.info("Registering tenant users")
+        fetch_ee_implementation_or_noop(
+            "onyx.server.tenants.billing", "register_tenant_users", None
+        )(tenant_id, get_live_users_count(db_session))
+
+        return number_of_invited_users
+    except Exception as e:
+        logger.error(f"Failed to register tenant users: {str(e)}")
+        logger.info(
+            "Reverting changes: removing users from tenant and resetting invited users"
+        )
+        write_invited_users(initial_invited_users)  # Reset to original state
+        fetch_ee_implementation_or_noop(
+            "onyx.server.tenants.user_mapping", "remove_users_from_tenant", None
+        )(new_invited_emails, tenant_id)
+        raise e
 
 
 @router.patch("/manage/admin/remove-invited-user", tags=PUBLIC_API_TAGS)

backend/onyx/server/query_and_chat/chat_backend.py

@@ -587,7 +587,6 @@ def handle_send_chat_message(
                     request.headers
                 ),
                 mcp_headers=chat_message_req.mcp_headers,
-                additional_context=chat_message_req.additional_context,
                 external_state_container=state_container,
             )
             result = gather_stream_full(packets, state_container)
@@ -610,7 +609,6 @@ def handle_send_chat_message(
                         request.headers
                     ),
                     mcp_headers=chat_message_req.mcp_headers,
-                    additional_context=chat_message_req.additional_context,
                     external_state_container=state_container,
                 ):
                     yield get_json_line(obj.model_dump())

backend/onyx/server/query_and_chat/models.py

@@ -125,11 +125,6 @@ class SendMessageRequest(BaseModel):
     # - No CitationInfo packets are emitted during streaming
     include_citations: bool = True
 
-    # Additional context injected into the LLM call but NOT stored in the DB
-    # (not shown in chat history). Used e.g. by the Chrome extension to pass
-    # the current tab URL when "Read this tab" is enabled.
-    additional_context: str | None = None
-
     @model_validator(mode="after")
     def check_chat_session_id_or_info(self) -> "SendMessageRequest":
         # If neither is provided, default to creating a new chat session using the

backend/onyx/setup.py

@@ -4,7 +4,6 @@ from sqlalchemy.orm import Session
 
 from onyx.configs.app_configs import DISABLE_INDEX_UPDATE_ON_SWAP
 from onyx.configs.app_configs import DISABLE_VECTOR_DB
-from onyx.configs.app_configs import ENABLE_OPENSEARCH_INDEXING_FOR_ONYX
 from onyx.configs.app_configs import INTEGRATION_TESTS_MODE
 from onyx.configs.app_configs import MANAGED_VESPA
 from onyx.configs.app_configs import VESPA_NUM_ATTEMPTS_ON_STARTUP
@@ -33,9 +32,6 @@ from onyx.db.search_settings import update_current_search_settings
 from onyx.db.swap_index import check_and_perform_index_swap
 from onyx.document_index.factory import get_all_document_indices
 from onyx.document_index.interfaces import DocumentIndex
-from onyx.document_index.opensearch.client import OpenSearchClient
-from onyx.document_index.opensearch.client import wait_for_opensearch_with_timeout
-from onyx.document_index.opensearch.opensearch_document_index import set_cluster_state
 from onyx.document_index.vespa.index import VespaIndex
 from onyx.indexing.models import IndexingSetting
 from onyx.key_value_store.factory import get_kv_store
@@ -315,14 +311,7 @@ def setup_multitenant_onyx() -> None:
         logger.notice("DISABLE_VECTOR_DB is set — skipping multitenant Vespa setup.")
         return
 
-    if ENABLE_OPENSEARCH_INDEXING_FOR_ONYX:
-        opensearch_client = OpenSearchClient()
-        if not wait_for_opensearch_with_timeout(client=opensearch_client):
-            raise RuntimeError("Failed to connect to OpenSearch.")
-        set_cluster_state(opensearch_client)
-
     # For Managed Vespa, the schema is sent over via the Vespa Console manually.
-    # NOTE: Pretty sure this code is never hit in any production environment.
     if not MANAGED_VESPA:
         setup_vespa_multitenant(SUPPORTED_EMBEDDING_MODELS)
 

backend/onyx/tools/tool_constructor.py

@@ -54,7 +54,6 @@ logger = setup_logger()
 class SearchToolConfig(BaseModel):
     user_selected_filters: BaseFilters | None = None
     project_id: int | None = None
-    persona_id: int | None = None
     bypass_acl: bool = False
     additional_context: str | None = None
     slack_context: SlackContext | None = None
@@ -181,7 +180,6 @@ def construct_tools(
                     document_index=document_index,
                     user_selected_filters=search_tool_config.user_selected_filters,
                     project_id=search_tool_config.project_id,
-                    persona_id=search_tool_config.persona_id,
                     bypass_acl=search_tool_config.bypass_acl,
                     slack_context=search_tool_config.slack_context,
                     enable_slack_search=search_tool_config.enable_slack_search,
@@ -429,7 +427,6 @@ def construct_tools(
             document_index=document_index,
             user_selected_filters=search_tool_config.user_selected_filters,
             project_id=search_tool_config.project_id,
-            persona_id=search_tool_config.persona_id,
             bypass_acl=search_tool_config.bypass_acl,
             slack_context=search_tool_config.slack_context,
             enable_slack_search=search_tool_config.enable_slack_search,

backend/onyx/tools/tool_implementations/python/code_interpreter_client.py

@@ -1,8 +1,5 @@
-import json
-from collections.abc import Generator
 from typing import Literal
 from typing import TypedDict
-from typing import Union
 
 import requests
 from pydantic import BaseModel
@@ -39,39 +36,6 @@ class ExecuteResponse(BaseModel):
     files: list[WorkspaceFile]
 
 
-class StreamOutputEvent(BaseModel):
-    """SSE 'output' event: a chunk of stdout or stderr"""
-
-    stream: Literal["stdout", "stderr"]
-    data: str
-
-
-class StreamResultEvent(BaseModel):
-    """SSE 'result' event: final execution result"""
-
-    exit_code: int | None
-    timed_out: bool
-    duration_ms: int
-    files: list[WorkspaceFile]
-
-
-class StreamErrorEvent(BaseModel):
-    """SSE 'error' event: execution-level error"""
-
-    message: str
-
-
-StreamEvent = Union[StreamOutputEvent, StreamResultEvent, StreamErrorEvent]
-
-_SSE_EVENT_MAP: dict[
-    str, type[StreamOutputEvent | StreamResultEvent | StreamErrorEvent]
-] = {
-    "output": StreamOutputEvent,
-    "result": StreamResultEvent,
-    "error": StreamErrorEvent,
-}
-
-
 class CodeInterpreterClient:
     """Client for Code Interpreter service"""
 
@@ -81,34 +45,6 @@ class CodeInterpreterClient:
         self.base_url = base_url.rstrip("/")
         self.session = requests.Session()
 
-    def _build_payload(
-        self,
-        code: str,
-        stdin: str | None,
-        timeout_ms: int,
-        files: list[FileInput] | None,
-    ) -> dict:
-        payload: dict = {
-            "code": code,
-            "timeout_ms": timeout_ms,
-        }
-        if stdin is not None:
-            payload["stdin"] = stdin
-        if files:
-            payload["files"] = files
-        return payload
-
-    def health(self) -> bool:
-        """Check if the Code Interpreter service is healthy"""
-        url = f"{self.base_url}/health"
-        try:
-            response = self.session.get(url, timeout=5)
-            response.raise_for_status()
-            return response.json().get("status") == "ok"
-        except Exception as e:
-            logger.warning(f"Exception caught when checking health, e={e}")
-            return False
-
     def execute(
         self,
         code: str,
@@ -116,110 +52,25 @@ class CodeInterpreterClient:
         timeout_ms: int = 30000,
         files: list[FileInput] | None = None,
     ) -> ExecuteResponse:
-        """Execute Python code (batch)"""
+        """Execute Python code"""
         url = f"{self.base_url}/v1/execute"
-        payload = self._build_payload(code, stdin, timeout_ms, files)
+
+        payload = {
+            "code": code,
+            "timeout_ms": timeout_ms,
+        }
+
+        if stdin is not None:
+            payload["stdin"] = stdin
+
+        if files:
+            payload["files"] = files
 
         response = self.session.post(url, json=payload, timeout=timeout_ms / 1000 + 10)
         response.raise_for_status()
 
         return ExecuteResponse(**response.json())
 
-    def execute_streaming(
-        self,
-        code: str,
-        stdin: str | None = None,
-        timeout_ms: int = 30000,
-        files: list[FileInput] | None = None,
-    ) -> Generator[StreamEvent, None, None]:
-        """Execute Python code with streaming SSE output.
-
-        Yields StreamEvent objects (StreamOutputEvent, StreamResultEvent,
-        StreamErrorEvent) as execution progresses. Falls back to batch
-        execution if the streaming endpoint is not available (older
-        code-interpreter versions).
-        """
-        url = f"{self.base_url}/v1/execute/stream"
-        payload = self._build_payload(code, stdin, timeout_ms, files)
-
-        response = self.session.post(
-            url,
-            json=payload,
-            stream=True,
-            timeout=timeout_ms / 1000 + 10,
-        )
-
-        if response.status_code == 404:
-            logger.info(
-                "Streaming endpoint not available, " "falling back to batch execution"
-            )
-            response.close()
-            yield from self._batch_as_stream(code, stdin, timeout_ms, files)
-            return
-
-        response.raise_for_status()
-        yield from self._parse_sse(response)
-
-    def _parse_sse(
-        self, response: requests.Response
-    ) -> Generator[StreamEvent, None, None]:
-        """Parse SSE streaming response into StreamEvent objects.
-
-        Expected format per event:
-            event: <type>
-            data: <json>
-            <blank line>
-        """
-        event_type: str | None = None
-        data_lines: list[str] = []
-
-        for line in response.iter_lines(decode_unicode=True):
-            if line is None:
-                continue
-
-            if line == "":
-                # Blank line marks end of an SSE event
-                if event_type is not None and data_lines:
-                    data = "\n".join(data_lines)
-                    model_cls = _SSE_EVENT_MAP.get(event_type)
-                    if model_cls is not None:
-                        yield model_cls(**json.loads(data))
-                    else:
-                        logger.warning(f"Unknown SSE event type: {event_type}")
-                event_type = None
-                data_lines = []
-            elif line.startswith("event:"):
-                event_type = line[len("event:") :].strip()
-            elif line.startswith("data:"):
-                data_lines.append(line[len("data:") :].strip())
-
-        if event_type is not None or data_lines:
-            logger.warning(
-                f"SSE stream ended with incomplete event: "
-                f"event_type={event_type}, data_lines={data_lines}"
-            )
-
-    def _batch_as_stream(
-        self,
-        code: str,
-        stdin: str | None,
-        timeout_ms: int,
-        files: list[FileInput] | None,
-    ) -> Generator[StreamEvent, None, None]:
-        """Execute via batch endpoint and yield results as stream events."""
-        result = self.execute(code, stdin, timeout_ms, files)
-
-        if result.stdout:
-            yield StreamOutputEvent(stream="stdout", data=result.stdout)
-        if result.stderr:
-            yield StreamOutputEvent(stream="stderr", data=result.stderr)
-        yield StreamResultEvent(
-            exit_code=result.exit_code,
-            timed_out=result.timed_out,
-            duration_ms=result.duration_ms,
-            files=result.files,
-        )
-
     def upload_file(self, file_content: bytes, filename: str) -> str:
         """Upload file to Code Interpreter and return file_id"""
         url = f"{self.base_url}/v1/files"

backend/onyx/tools/tool_implementations/python/python_tool.py

@@ -12,7 +12,6 @@ from onyx.configs.app_configs import CODE_INTERPRETER_BASE_URL
 from onyx.configs.app_configs import CODE_INTERPRETER_DEFAULT_TIMEOUT_MS
 from onyx.configs.app_configs import CODE_INTERPRETER_MAX_OUTPUT_LENGTH
 from onyx.configs.constants import FileOrigin
-from onyx.db.code_interpreter import fetch_code_interpreter_server
 from onyx.file_store.utils import build_full_frontend_file_url
 from onyx.file_store.utils import get_default_file_store
 from onyx.server.query_and_chat.placement import Placement
@@ -29,15 +28,6 @@ from onyx.tools.tool_implementations.python.code_interpreter_client import (
     CodeInterpreterClient,
 )
 from onyx.tools.tool_implementations.python.code_interpreter_client import FileInput
-from onyx.tools.tool_implementations.python.code_interpreter_client import (
-    StreamErrorEvent,
-)
-from onyx.tools.tool_implementations.python.code_interpreter_client import (
-    StreamOutputEvent,
-)
-from onyx.tools.tool_implementations.python.code_interpreter_client import (
-    StreamResultEvent,
-)
 from onyx.utils.logger import setup_logger
 
 
@@ -104,10 +94,8 @@ class PythonTool(Tool[PythonToolOverrideKwargs]):
     @override
     @classmethod
     def is_available(cls, db_session: Session) -> bool:
-        if not CODE_INTERPRETER_BASE_URL:
-            return False
-        server = fetch_code_interpreter_server(db_session)
-        return server.server_enabled
+        is_available = bool(CODE_INTERPRETER_BASE_URL)
+        return is_available
 
     def tool_definition(self) -> dict:
         return {
@@ -193,50 +181,19 @@ class PythonTool(Tool[PythonToolOverrideKwargs]):
         try:
             logger.debug(f"Executing code: {code}")
 
-            # Execute code with streaming (falls back to batch if unavailable)
-            stdout_parts: list[str] = []
-            stderr_parts: list[str] = []
-            result_event: StreamResultEvent | None = None
-
-            for event in client.execute_streaming(
+            # Execute code with timeout
+            response = client.execute(
                 code=code,
                 timeout_ms=CODE_INTERPRETER_DEFAULT_TIMEOUT_MS,
                 files=files_to_stage or None,
-            ):
-                if isinstance(event, StreamOutputEvent):
-                    if event.stream == "stdout":
-                        stdout_parts.append(event.data)
-                    else:
-                        stderr_parts.append(event.data)
-                    # Emit incremental delta to frontend
-                    self.emitter.emit(
-                        Packet(
-                            placement=placement,
-                            obj=PythonToolDelta(
-                                stdout=event.data if event.stream == "stdout" else "",
-                                stderr=event.data if event.stream == "stderr" else "",
-                            ),
-                        )
-                    )
-                elif isinstance(event, StreamResultEvent):
-                    result_event = event
-                elif isinstance(event, StreamErrorEvent):
-                    raise RuntimeError(f"Code interpreter error: {event.message}")
-
-            if result_event is None:
-                raise RuntimeError(
-                    "Code interpreter stream ended without a result event"
-                )
-
-            full_stdout = "".join(stdout_parts)
-            full_stderr = "".join(stderr_parts)
+            )
 
             # Truncate output for LLM consumption
             truncated_stdout = _truncate_output(
-                full_stdout, CODE_INTERPRETER_MAX_OUTPUT_LENGTH, "stdout"
+                response.stdout, CODE_INTERPRETER_MAX_OUTPUT_LENGTH, "stdout"
             )
             truncated_stderr = _truncate_output(
-                full_stderr, CODE_INTERPRETER_MAX_OUTPUT_LENGTH, "stderr"
+                response.stderr, CODE_INTERPRETER_MAX_OUTPUT_LENGTH, "stderr"
             )
 
             # Handle generated files
@@ -245,7 +202,7 @@ class PythonTool(Tool[PythonToolOverrideKwargs]):
             file_ids_to_cleanup: list[str] = []
             file_store = get_default_file_store()
 
-            for workspace_file in result_event.files:
+            for workspace_file in response.files:
                 if workspace_file.kind != "file" or not workspace_file.file_id:
                     continue
 
@@ -301,23 +258,26 @@ class PythonTool(Tool[PythonToolOverrideKwargs]):
                         f"Failed to delete Code Interpreter staged file {file_mapping['file_id']}: {e}"
                     )
 
-            # Emit file_ids once files are processed
-            if generated_file_ids:
-                self.emitter.emit(
-                    Packet(
-                        placement=placement,
-                        obj=PythonToolDelta(file_ids=generated_file_ids),
-                    )
+            # Emit delta with stdout/stderr and generated files
+            self.emitter.emit(
+                Packet(
+                    placement=placement,
+                    obj=PythonToolDelta(
+                        stdout=truncated_stdout,
+                        stderr=truncated_stderr,
+                        file_ids=generated_file_ids,
+                    ),
                 )
+            )
 
             # Build result
             result = LlmPythonExecutionResult(
                 stdout=truncated_stdout,
                 stderr=truncated_stderr,
-                exit_code=result_event.exit_code,
-                timed_out=result_event.timed_out,
+                exit_code=response.exit_code,
+                timed_out=response.timed_out,
                 generated_files=generated_files,
-                error=None if result_event.exit_code == 0 else truncated_stderr,
+                error=None if response.exit_code == 0 else truncated_stderr,
             )
 
             # Serialize result for LLM

backend/onyx/tools/tool_implementations/search/search_tool.py

@@ -247,8 +247,6 @@ class SearchTool(Tool[SearchToolOverrideKwargs]):
         user_selected_filters: BaseFilters | None,
         # If the chat is part of a project
         project_id: int | None,
-        # If set, search scopes to files attached to this persona
-        persona_id: int | None = None,
         bypass_acl: bool = False,
         # Slack context for federated Slack search (tokens fetched internally)
         slack_context: SlackContext | None = None,
@@ -263,7 +261,6 @@ class SearchTool(Tool[SearchToolOverrideKwargs]):
         self.document_index = document_index
         self.user_selected_filters = user_selected_filters
         self.project_id = project_id
-        self.persona_id = persona_id
         self.bypass_acl = bypass_acl
         self.slack_context = slack_context
         self.enable_slack_search = enable_slack_search
@@ -459,7 +456,6 @@ class SearchTool(Tool[SearchToolOverrideKwargs]):
                 limit=num_hits,
             ),
             project_id=self.project_id,
-            persona_id=self.persona_id,
             document_index=self.document_index,
             user=self.user,
             persona=self.persona,

backend/requirements/default.txt

@@ -6,8 +6,6 @@ aioboto3==15.1.0
     # via onyx
 aiobotocore==2.24.0
     # via aioboto3
-aiofile==3.9.0
-    # via py-key-value-aio
 aiofiles==25.1.0
     # via
     #   aioboto3
@@ -42,10 +40,8 @@ anyio==4.11.0
     #   httpx
     #   mcp
     #   openai
-    #   py-key-value-aio
     #   sse-starlette
     #   starlette
-    #   watchfiles
 argon2-cffi==23.1.0
     # via pwdlib
 argon2-cffi-bindings==25.1.0
@@ -78,7 +74,9 @@ backports-tarfile==1.2.0 ; python_full_version < '3.12'
 bcrypt==4.3.0
     # via pwdlib
 beartype==0.22.6
-    # via py-key-value-aio
+    # via
+    #   py-key-value-aio
+    #   py-key-value-shared
 beautifulsoup4==4.12.3
     # via
     #   atlassian-python-api
@@ -112,8 +110,6 @@ cachetools==6.2.2
     # via
     #   google-auth
     #   py-key-value-aio
-caio==0.9.25
-    # via aiofile
 celery==5.5.1
     # via onyx
 certifi==2025.11.12
@@ -174,6 +170,7 @@ cloudpickle==3.1.2
     # via
     #   dask
     #   distributed
+    #   pydocket
 cobble==0.1.4
     # via mammoth
 cohere==5.6.1
@@ -221,6 +218,8 @@ deprecated==1.3.1
     #   pygithub
 discord-py==2.4.0
     # via onyx
+diskcache==5.6.3
+    # via py-key-value-aio
 distributed==2026.1.1
     # via onyx
 distro==1.9.0
@@ -257,6 +256,8 @@ exceptiongroup==1.3.0
     # via
     #   braintrust
     #   fastmcp
+fakeredis==2.33.0
+    # via pydocket
 fastapi==0.128.0
     # via
     #   fastapi-limiter
@@ -272,7 +273,7 @@ fastapi-users-db-sqlalchemy==7.0.0
     # via onyx
 fastavro==1.12.1
     # via cohere
-fastmcp==3.0.2
+fastmcp==2.14.2
     # via onyx
 fastuuid==0.14.0
     # via litellm
@@ -477,9 +478,7 @@ jsonpatch==1.33
 jsonpointer==3.0.0
     # via jsonpatch
 jsonref==1.1.0
-    # via
-    #   fastmcp
-    #   onyx
+    # via onyx
 jsonschema==4.25.1
     # via
     #   litellm
@@ -514,6 +513,8 @@ locket==1.0.0
     # via
     #   distributed
     #   partd
+lupa==2.6
+    # via fakeredis
 lxml==5.3.0
     # via
     #   htmldate
@@ -555,7 +556,7 @@ marshmallow==3.26.2
     # via dataclasses-json
 matrix-client==0.3.2
     # via zulip
-mcp==1.26.0
+mcp==1.25.0
     # via
     #   claude-agent-sdk
     #   fastmcp
@@ -612,7 +613,7 @@ oauthlib==3.2.2
     #   kubernetes
     #   onyx
     #   requests-oauthlib
-office365-rest-python-client==2.6.2
+office365-rest-python-client==2.5.9
     # via onyx
 olefile==0.47
     # via
@@ -641,16 +642,22 @@ opensearch-py==3.0.0
 opentelemetry-api==1.39.1
     # via
     #   ddtrace
-    #   fastmcp
     #   langfuse
     #   openinference-instrumentation
     #   opentelemetry-exporter-otlp-proto-http
+    #   opentelemetry-exporter-prometheus
+    #   opentelemetry-instrumentation
     #   opentelemetry-sdk
     #   opentelemetry-semantic-conventions
+    #   pydocket
 opentelemetry-exporter-otlp-proto-common==1.39.1
     # via opentelemetry-exporter-otlp-proto-http
 opentelemetry-exporter-otlp-proto-http==1.39.1
     # via langfuse
+opentelemetry-exporter-prometheus==0.60b1
+    # via pydocket
+opentelemetry-instrumentation==0.60b1
+    # via pydocket
 opentelemetry-proto==1.39.1
     # via
     #   onyx
@@ -661,15 +668,17 @@ opentelemetry-sdk==1.39.1
     #   langfuse
     #   openinference-instrumentation
     #   opentelemetry-exporter-otlp-proto-http
+    #   opentelemetry-exporter-prometheus
 opentelemetry-semantic-conventions==0.60b1
-    # via opentelemetry-sdk
+    # via
+    #   opentelemetry-instrumentation
+    #   opentelemetry-sdk
 orjson==3.11.4 ; platform_python_implementation != 'PyPy'
     # via langsmith
 packaging==24.2
     # via
     #   dask
     #   distributed
-    #   fastmcp
     #   google-cloud-aiplatform
     #   google-cloud-bigquery
     #   huggingface-hub
@@ -680,6 +689,7 @@ packaging==24.2
     #   langsmith
     #   marshmallow
     #   onnxruntime
+    #   opentelemetry-instrumentation
     #   pytest
     #   pywikibot
 pandas==2.3.3
@@ -692,6 +702,8 @@ passlib==1.7.4
     # via onyx
 pathable==0.4.4
     # via jsonschema-path
+pathvalidate==3.3.1
+    # via py-key-value-aio
 pdfminer-six==20251107
     # via markitdown
 pillow==12.1.1
@@ -711,7 +723,9 @@ ply==3.11
 prometheus-client==0.23.1
     # via
     #   onyx
+    #   opentelemetry-exporter-prometheus
     #   prometheus-fastapi-instrumentator
+    #   pydocket
 prometheus-fastapi-instrumentator==7.1.0
     # via onyx
 prompt-toolkit==3.0.52
@@ -750,8 +764,12 @@ pwdlib==0.3.0
     # via fastapi-users
 py==1.11.0
     # via retry
-py-key-value-aio==0.4.4
-    # via fastmcp
+py-key-value-aio==0.3.0
+    # via
+    #   fastmcp
+    #   pydocket
+py-key-value-shared==0.3.0
+    # via py-key-value-aio
 pyairtable==3.0.1
     # via onyx
 pyasn1==0.6.2
@@ -788,6 +806,8 @@ pydantic-core==2.33.2
     # via pydantic
 pydantic-settings==2.12.0
     # via mcp
+pydocket==0.16.3
+    # via fastmcp
 pyee==13.0.0
     # via playwright
 pygithub==2.5.0
@@ -859,6 +879,8 @@ python-http-client==3.3.7
     # via sendgrid
 python-iso639==2025.11.16
     # via unstructured
+python-json-logger==4.0.0
+    # via pydocket
 python-magic==0.4.27
     # via unstructured
 python-multipart==0.0.22
@@ -896,7 +918,6 @@ pyyaml==6.0.3
     # via
     #   dask
     #   distributed
-    #   fastmcp
     #   huggingface-hub
     #   jsonschema-path
     #   kubernetes
@@ -907,8 +928,11 @@ rapidfuzz==3.13.0
     #   unstructured
 redis==5.0.8
     # via
+    #   fakeredis
     #   fastapi-limiter
     #   onyx
+    #   py-key-value-aio
+    #   pydocket
 referencing==0.36.2
     # via
     #   jsonschema
@@ -983,6 +1007,7 @@ rich==14.2.0
     # via
     #   cyclopts
     #   fastmcp
+    #   pydocket
     #   rich-rst
     #   typer
 rich-rst==1.3.2
@@ -1031,7 +1056,9 @@ sniffio==1.3.1
     #   anyio
     #   openai
 sortedcontainers==2.4.0
-    # via distributed
+    # via
+    #   distributed
+    #   fakeredis
 soupsieve==2.8
     # via beautifulsoup4
 sqlalchemy==2.0.15
@@ -1097,7 +1124,9 @@ tqdm==4.67.1
 trafilatura==1.12.2
     # via onyx
 typer==0.20.0
-    # via mcp
+    # via
+    #   mcp
+    #   pydocket
 types-awscrt==0.28.4
     # via botocore-stubs
 types-openpyxl==3.0.4.7
@@ -1133,10 +1162,11 @@ typing-extensions==4.15.0
     #   opentelemetry-exporter-otlp-proto-http
     #   opentelemetry-sdk
     #   opentelemetry-semantic-conventions
-    #   py-key-value-aio
+    #   py-key-value-shared
     #   pyairtable
     #   pydantic
     #   pydantic-core
+    #   pydocket
     #   pyee
     #   pygithub
     #   python-docx
@@ -1204,8 +1234,6 @@ vine==5.1.0
     #   kombu
 voyageai==0.2.3
     # via onyx
-watchfiles==1.1.1
-    # via fastmcp
 wcwidth==0.2.14
     # via prompt-toolkit
 webencodings==0.5.1
@@ -1226,6 +1254,7 @@ wrapt==1.17.3
     #   deprecated
     #   langfuse
     #   openinference-instrumentation
+    #   opentelemetry-instrumentation
     #   unstructured
 xlrd==2.0.2
     # via markitdown

backend/requirements/dev.txt

@@ -288,7 +288,7 @@ matplotlib-inline==0.2.1
     # via
     #   ipykernel
     #   ipython
-mcp==1.26.0
+mcp==1.25.0
     # via claude-agent-sdk
 multidict==6.7.0
     # via
@@ -317,7 +317,7 @@ oauthlib==3.2.2
     # via
     #   kubernetes
     #   requests-oauthlib
-onyx-devtools==0.6.2
+onyx-devtools==0.6.0
     # via onyx
 openai==2.14.0
     # via

backend/requirements/ee.txt

@@ -211,7 +211,7 @@ litellm==1.81.6
     # via onyx
 markupsafe==3.0.3
     # via jinja2
-mcp==1.26.0
+mcp==1.25.0
     # via claude-agent-sdk
 monotonic==1.6
     # via posthog

backend/requirements/model_server.txt

@@ -246,7 +246,7 @@ litellm==1.81.6
     # via onyx
 markupsafe==3.0.3
     # via jinja2
-mcp==1.26.0
+mcp==1.25.0
     # via claude-agent-sdk
 mpmath==1.3.0
     # via sympy

backend/scripts/query_time_check/seed_dummy_docs.py

@@ -95,7 +95,6 @@ def generate_dummy_chunk(
     return DocMetadataAwareIndexChunk.from_index_chunk(
         index_chunk=chunk,
         user_project=[],
-        personas=[],
         access=DocumentAccess.build(
             user_emails=user_emails,
             user_groups=user_groups,

backend/scripts/restart_containers.sh

@@ -3,8 +3,8 @@ set -e
 
 cleanup() {
   echo "Error occurred. Cleaning up..."
-  docker stop onyx_postgres onyx_vespa onyx_redis onyx_minio onyx_code_interpreter 2>/dev/null || true
-  docker rm onyx_postgres onyx_vespa onyx_redis onyx_minio onyx_code_interpreter 2>/dev/null || true
+  docker stop onyx_postgres onyx_vespa onyx_redis onyx_minio 2>/dev/null || true
+  docker rm onyx_postgres onyx_vespa onyx_redis onyx_minio 2>/dev/null || true
 }
 
 # Trap errors and output a message, then cleanup
@@ -20,8 +20,8 @@ MINIO_VOLUME=${4:-""}  # Default is empty if not provided
 
 # Stop and remove the existing containers
 echo "Stopping and removing existing containers..."
-docker stop onyx_postgres onyx_vespa onyx_redis onyx_minio onyx_code_interpreter 2>/dev/null || true
-docker rm onyx_postgres onyx_vespa onyx_redis onyx_minio onyx_code_interpreter 2>/dev/null || true
+docker stop onyx_postgres onyx_vespa onyx_redis onyx_minio 2>/dev/null || true
+docker rm onyx_postgres onyx_vespa onyx_redis onyx_minio 2>/dev/null || true
 
 # Start the PostgreSQL container with optional volume
 echo "Starting PostgreSQL container..."
@@ -55,10 +55,6 @@ else
     docker run --detach --name onyx_minio --publish 9004:9000 --publish 9005:9001 -e MINIO_ROOT_USER=minioadmin -e MINIO_ROOT_PASSWORD=minioadmin minio/minio server /data --console-address ":9001"
 fi
 
-# Start the Code Interpreter container
-echo "Starting Code Interpreter container..."
-docker run --detach --name onyx_code_interpreter --publish 8000:8000 --user root -v /var/run/docker.sock:/var/run/docker.sock onyxdotapp/code-interpreter:latest bash ./entrypoint.sh code-interpreter-api
-
 # Ensure alembic runs in the correct directory (backend/)
 SCRIPT_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )"
 PARENT_DIR="$(dirname "$SCRIPT_DIR")"

backend/tests/daily/conftest.py

@@ -9,7 +9,6 @@ from collections.abc import AsyncGenerator
 from collections.abc import Generator
 from contextlib import asynccontextmanager
 from unittest.mock import MagicMock
-from unittest.mock import patch
 
 import pytest
 from dotenv import load_dotenv
@@ -47,15 +46,11 @@ def mock_current_admin_user() -> MagicMock:
 
 @pytest.fixture(scope="function")
 def client() -> Generator[TestClient, None, None]:
-    # Initialize TestClient with the FastAPI app using a no-op test lifespan.
-    # Patch out prometheus metrics setup to avoid "Duplicated timeseries in
-    # CollectorRegistry" errors when multiple tests each create a new app
-    # (prometheus registers metrics globally and rejects duplicate names).
+    # Initialize TestClient with the FastAPI app using a no-op test lifespan
     get_app = fetch_versioned_implementation(
         module="onyx.main", attribute="get_application"
     )
-    with patch("onyx.main.setup_prometheus_metrics"):
-        app: FastAPI = get_app(lifespan_override=test_lifespan)
+    app: FastAPI = get_app(lifespan_override=test_lifespan)
 
     # Override the database session dependency with a mock
     # (these tests don't actually need DB access)

backend/tests/external_dependency_unit/mock_search_pipeline.py

@@ -144,8 +144,7 @@ def use_mock_search_pipeline(
         auto_detect_filters: bool = False,  # noqa: ARG001
         llm: LLM | None = None,  # noqa: ARG001
         project_id: int | None = None,  # noqa: ARG001
-        persona_id: int | None = None,  # noqa: ARG001
-        # Pre-fetched data (used by SearchTool to avoid DB access in parallel calls)
+        # Pre-fetched data (used by SearchTool to avoid DB access in parallel)
         acl_filters: list[str] | None = None,  # noqa: ARG001
         embedding_model: EmbeddingModel | None = None,  # noqa: ARG001
         prefetched_federated_retrieval_infos: (  # noqa: ARG001

backend/tests/external_dependency_unit/opensearch/test_assistant_knowledge_filter.py

@@ -38,7 +38,6 @@ def _get_search_filters(
         tags=[],
         document_sets=[],
         project_id=None,
-        persona_id=None,
         time_cutoff=None,
         min_chunk_index=None,
         max_chunk_index=None,

backend/tests/external_dependency_unit/opensearch/test_opensearch_client.py

@@ -1,4 +1,4 @@
-"""External dependency unit tests for OpenSearchIndexClient.
+"""External dependency unit tests for OpenSearchClient.
 
 These tests assume OpenSearch is running and test all implemented methods
 using real schemas, pipelines, and search queries from the codebase.
@@ -19,7 +19,7 @@ from onyx.access.utils import prefix_user_email
 from onyx.configs.constants import DocumentSource
 from onyx.context.search.models import IndexFilters
 from onyx.document_index.interfaces_new import TenantState
-from onyx.document_index.opensearch.client import OpenSearchIndexClient
+from onyx.document_index.opensearch.client import OpenSearchClient
 from onyx.document_index.opensearch.client import wait_for_opensearch_with_timeout
 from onyx.document_index.opensearch.constants import DEFAULT_MAX_CHUNK_SIZE
 from onyx.document_index.opensearch.opensearch_document_index import (
@@ -125,10 +125,10 @@ def opensearch_available() -> None:
 @pytest.fixture(scope="function")
 def test_client(
     opensearch_available: None,  # noqa: ARG001
-) -> Generator[OpenSearchIndexClient, None, None]:
+) -> Generator[OpenSearchClient, None, None]:
     """Creates an OpenSearch client for testing with automatic cleanup."""
     test_index_name = f"test_index_{uuid.uuid4().hex[:8]}"
-    client = OpenSearchIndexClient(index_name=test_index_name)
+    client = OpenSearchClient(index_name=test_index_name)
 
     yield client  # Test runs here.
 
@@ -142,7 +142,7 @@ def test_client(
 
 
 @pytest.fixture(scope="function")
-def search_pipeline(test_client: OpenSearchIndexClient) -> Generator[None, None, None]:
+def search_pipeline(test_client: OpenSearchClient) -> Generator[None, None, None]:
     """Creates a search pipeline for testing with automatic cleanup."""
     test_client.create_search_pipeline(
         pipeline_id=MIN_MAX_NORMALIZATION_PIPELINE_NAME,
@@ -158,9 +158,9 @@ def search_pipeline(test_client: OpenSearchIndexClient) -> Generator[None, None,
 
 
 class TestOpenSearchClient:
-    """Tests for OpenSearchIndexClient."""
+    """Tests for OpenSearchClient."""
 
-    def test_create_index(self, test_client: OpenSearchIndexClient) -> None:
+    def test_create_index(self, test_client: OpenSearchClient) -> None:
         """Tests creating an index with a real schema."""
         # Precondition.
         mappings = DocumentSchema.get_document_schema(
@@ -176,7 +176,7 @@ class TestOpenSearchClient:
         # Verify index exists.
         assert test_client.validate_index(expected_mappings=mappings) is True
 
-    def test_delete_existing_index(self, test_client: OpenSearchIndexClient) -> None:
+    def test_delete_existing_index(self, test_client: OpenSearchClient) -> None:
         """Tests deleting an existing index returns True."""
         # Precondition.
         mappings = DocumentSchema.get_document_schema(
@@ -193,7 +193,7 @@ class TestOpenSearchClient:
         assert result is True
         assert test_client.validate_index(expected_mappings=mappings) is False
 
-    def test_delete_nonexistent_index(self, test_client: OpenSearchIndexClient) -> None:
+    def test_delete_nonexistent_index(self, test_client: OpenSearchClient) -> None:
         """Tests deleting a nonexistent index returns False."""
         # Under test.
         # Don't create index, just try to delete.
@@ -202,7 +202,7 @@ class TestOpenSearchClient:
         # Postcondition.
         assert result is False
 
-    def test_index_exists(self, test_client: OpenSearchIndexClient) -> None:
+    def test_index_exists(self, test_client: OpenSearchClient) -> None:
         """Tests checking if an index exists."""
         # Precondition.
         # Index should not exist before creation.
@@ -219,7 +219,7 @@ class TestOpenSearchClient:
         # Index should exist after creation.
         assert test_client.index_exists() is True
 
-    def test_validate_index(self, test_client: OpenSearchIndexClient) -> None:
+    def test_validate_index(self, test_client: OpenSearchClient) -> None:
         """Tests validating an index."""
         # Precondition.
         mappings = DocumentSchema.get_document_schema(
@@ -239,120 +239,7 @@ class TestOpenSearchClient:
         # Should return True after creation.
         assert test_client.validate_index(expected_mappings=mappings) is True
 
-    def test_put_mapping_idempotent(self, test_client: OpenSearchIndexClient) -> None:
-        """Tests put_mapping with same schema is idempotent."""
-        # Precondition.
-        mappings = DocumentSchema.get_document_schema(
-            vector_dimension=128, multitenant=True
-        )
-        settings = DocumentSchema.get_index_settings()
-        test_client.create_index(mappings=mappings, settings=settings)
-
-        # Under test.
-        # Applying the same mappings again should succeed.
-        test_client.put_mapping(mappings)
-
-        # Postcondition.
-        # Index should still be valid.
-        assert test_client.validate_index(expected_mappings=mappings)
-
-    def test_put_mapping_adds_new_field(
-        self, test_client: OpenSearchIndexClient
-    ) -> None:
-        """Tests put_mapping successfully adds new fields to existing index."""
-        # Precondition.
-        # Create index with minimal schema (just required fields).
-        initial_mappings = {
-            "dynamic": "strict",
-            "properties": {
-                "document_id": {"type": "keyword"},
-                "chunk_index": {"type": "integer"},
-                "content": {"type": "text"},
-                "content_vector": {
-                    "type": "knn_vector",
-                    "dimension": 128,
-                    "method": {
-                        "name": "hnsw",
-                        "space_type": "cosinesimil",
-                        "engine": "lucene",
-                        "parameters": {"ef_construction": 512, "m": 16},
-                    },
-                },
-            },
-        }
-        settings = DocumentSchema.get_index_settings()
-        test_client.create_index(mappings=initial_mappings, settings=settings)
-
-        # Under test.
-        # Add a new field using put_mapping.
-        updated_mappings = {
-            "properties": {
-                "document_id": {"type": "keyword"},
-                "chunk_index": {"type": "integer"},
-                "content": {"type": "text"},
-                "content_vector": {
-                    "type": "knn_vector",
-                    "dimension": 128,
-                    "method": {
-                        "name": "hnsw",
-                        "space_type": "cosinesimil",
-                        "engine": "lucene",
-                        "parameters": {"ef_construction": 512, "m": 16},
-                    },
-                },
-                # New field
-                "new_test_field": {"type": "keyword"},
-            },
-        }
-        # Should not raise.
-        test_client.put_mapping(updated_mappings)
-
-        # Postcondition.
-        # Validate the new schema includes the new field.
-        assert test_client.validate_index(expected_mappings=updated_mappings)
-
-    def test_put_mapping_fails_on_type_change(
-        self, test_client: OpenSearchIndexClient
-    ) -> None:
-        """Tests put_mapping fails when trying to change existing field type."""
-        # Precondition.
-        initial_mappings = {
-            "dynamic": "strict",
-            "properties": {
-                "document_id": {"type": "keyword"},
-                "test_field": {"type": "keyword"},
-            },
-        }
-        settings = DocumentSchema.get_index_settings()
-        test_client.create_index(mappings=initial_mappings, settings=settings)
-
-        # Under test and postcondition.
-        # Try to change test_field type from keyword to text.
-        conflicting_mappings = {
-            "properties": {
-                "document_id": {"type": "keyword"},
-                "test_field": {"type": "text"},  # Changed from keyword to text
-            },
-        }
-        # Should raise because field type cannot be changed.
-        with pytest.raises(Exception, match="mapper|illegal_argument_exception"):
-            test_client.put_mapping(conflicting_mappings)
-
-    def test_put_mapping_on_nonexistent_index(
-        self, test_client: OpenSearchIndexClient
-    ) -> None:
-        """Tests put_mapping on non-existent index raises an error."""
-        # Precondition.
-        # Index does not exist yet.
-        mappings = DocumentSchema.get_document_schema(
-            vector_dimension=128, multitenant=True
-        )
-
-        # Under test and postcondition.
-        with pytest.raises(Exception, match="index_not_found_exception|404"):
-            test_client.put_mapping(mappings)
-
-    def test_create_duplicate_index(self, test_client: OpenSearchIndexClient) -> None:
+    def test_create_duplicate_index(self, test_client: OpenSearchClient) -> None:
         """Tests creating an index twice raises an error."""
         # Precondition.
         mappings = DocumentSchema.get_document_schema(
@@ -367,14 +254,14 @@ class TestOpenSearchClient:
         with pytest.raises(Exception, match="already exists"):
             test_client.create_index(mappings=mappings, settings=settings)
 
-    def test_update_settings(self, test_client: OpenSearchIndexClient) -> None:
+    def test_update_settings(self, test_client: OpenSearchClient) -> None:
         """Tests that update_settings raises NotImplementedError."""
         # Under test and postcondition.
         with pytest.raises(NotImplementedError):
             test_client.update_settings(settings={})
 
     def test_create_and_delete_search_pipeline(
-        self, test_client: OpenSearchIndexClient
+        self, test_client: OpenSearchClient
     ) -> None:
         """Tests creating and deleting a search pipeline."""
         # Under test and postcondition.
@@ -391,7 +278,7 @@ class TestOpenSearchClient:
         )
 
     def test_index_document(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """Tests indexing a document."""
         # Precondition.
@@ -419,7 +306,7 @@ class TestOpenSearchClient:
         )
 
     def test_bulk_index_documents(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """Tests bulk indexing documents."""
         # Precondition.
@@ -450,7 +337,7 @@ class TestOpenSearchClient:
         )
 
     def test_index_duplicate_document(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """Tests indexing a duplicate document raises an error."""
         # Precondition.
@@ -478,7 +365,7 @@ class TestOpenSearchClient:
             test_client.index_document(document=doc, tenant_state=tenant_state)
 
     def test_get_document(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """Tests getting a document."""
         # Precondition.
@@ -514,7 +401,7 @@ class TestOpenSearchClient:
         assert retrieved_doc == original_doc
 
     def test_get_nonexistent_document(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """Tests getting a nonexistent document raises an error."""
         # Precondition.
@@ -532,7 +419,7 @@ class TestOpenSearchClient:
             )
 
     def test_delete_existing_document(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """Tests deleting an existing document returns True."""
         # Precondition.
@@ -568,7 +455,7 @@ class TestOpenSearchClient:
             test_client.get_document(document_chunk_id=doc_chunk_id)
 
     def test_delete_nonexistent_document(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """Tests deleting a nonexistent document returns False."""
         # Precondition.
@@ -589,7 +476,7 @@ class TestOpenSearchClient:
         assert result is False
 
     def test_delete_by_query(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """Tests deleting documents by query."""
         # Precondition.
@@ -665,7 +552,7 @@ class TestOpenSearchClient:
         assert len(keep_ids) == 1
 
     def test_update_document(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """Tests updating a document's properties."""
         # Precondition.
@@ -714,7 +601,7 @@ class TestOpenSearchClient:
         assert updated_doc.public == doc.public
 
     def test_update_nonexistent_document(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """Tests updating a nonexistent document raises an error."""
         # Precondition.
@@ -736,7 +623,7 @@ class TestOpenSearchClient:
 
     def test_hybrid_search_with_pipeline(
         self,
-        test_client: OpenSearchIndexClient,
+        test_client: OpenSearchClient,
         search_pipeline: None,  # noqa: ARG002
         monkeypatch: pytest.MonkeyPatch,
     ) -> None:
@@ -817,7 +704,7 @@ class TestOpenSearchClient:
 
     def test_search_empty_index(
         self,
-        test_client: OpenSearchIndexClient,
+        test_client: OpenSearchClient,
         search_pipeline: None,  # noqa: ARG002
         monkeypatch: pytest.MonkeyPatch,
     ) -> None:
@@ -856,7 +743,7 @@ class TestOpenSearchClient:
 
     def test_hybrid_search_with_pipeline_and_filters(
         self,
-        test_client: OpenSearchIndexClient,
+        test_client: OpenSearchClient,
         search_pipeline: None,  # noqa: ARG002
         monkeypatch: pytest.MonkeyPatch,
     ) -> None:
@@ -976,7 +863,7 @@ class TestOpenSearchClient:
 
     def test_hybrid_search_with_pipeline_and_filters_returns_chunks_with_related_content_first(
         self,
-        test_client: OpenSearchIndexClient,
+        test_client: OpenSearchClient,
         search_pipeline: None,  # noqa: ARG002
         monkeypatch: pytest.MonkeyPatch,
     ) -> None:
@@ -1106,7 +993,7 @@ class TestOpenSearchClient:
             previous_score = current_score
 
     def test_delete_by_query_multitenant_isolation(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """
         Tests delete_by_query respects tenant boundaries in multi-tenant mode.
@@ -1200,7 +1087,7 @@ class TestOpenSearchClient:
         assert set(remaining_y_ids) == expected_y_ids
 
     def test_delete_by_query_nonexistent_document(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """
         Tests delete_by_query for non-existent document returns 0 deleted.
@@ -1229,7 +1116,7 @@ class TestOpenSearchClient:
         assert num_deleted == 0
 
     def test_search_for_document_ids(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """Tests search_for_document_ids method returns correct chunk IDs."""
         # Precondition.
@@ -1294,7 +1181,7 @@ class TestOpenSearchClient:
         assert set(chunk_ids) == expected_ids
 
     def test_search_with_no_document_access_can_retrieve_all_documents(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """
         Tests search with no document access can retrieve all documents, even
@@ -1372,7 +1259,7 @@ class TestOpenSearchClient:
 
     def test_time_cutoff_filter(
         self,
-        test_client: OpenSearchIndexClient,
+        test_client: OpenSearchClient,
         search_pipeline: None,  # noqa: ARG002
         monkeypatch: pytest.MonkeyPatch,
     ) -> None:
@@ -1465,7 +1352,7 @@ class TestOpenSearchClient:
         )
 
     def test_random_search(
-        self, test_client: OpenSearchIndexClient, monkeypatch: pytest.MonkeyPatch
+        self, test_client: OpenSearchClient, monkeypatch: pytest.MonkeyPatch
     ) -> None:
         """Tests the random search query works."""
         # Precondition.

backend/tests/external_dependency_unit/opensearch_migration/test_opensearch_migration_tasks.py

@@ -37,7 +37,6 @@ from onyx.db.opensearch_migration import build_sanitized_to_original_doc_id_mapp
 from onyx.db.search_settings import get_active_search_settings
 from onyx.document_index.interfaces_new import TenantState
 from onyx.document_index.opensearch.client import OpenSearchClient
-from onyx.document_index.opensearch.client import OpenSearchIndexClient
 from onyx.document_index.opensearch.client import wait_for_opensearch_with_timeout
 from onyx.document_index.opensearch.constants import DEFAULT_MAX_CHUNK_SIZE
 from onyx.document_index.opensearch.schema import DocumentChunk
@@ -75,7 +74,7 @@ CHUNK_COUNT = 5
 
 
 def _get_document_chunks_from_opensearch(
-    opensearch_client: OpenSearchIndexClient, document_id: str, current_tenant_id: str
+    opensearch_client: OpenSearchClient, document_id: str, current_tenant_id: str
 ) -> list[DocumentChunk]:
     opensearch_client.refresh_index()
     filters = IndexFilters(access_control_list=None, tenant_id=current_tenant_id)
@@ -96,7 +95,7 @@ def _get_document_chunks_from_opensearch(
 
 
 def _delete_document_chunks_from_opensearch(
-    opensearch_client: OpenSearchIndexClient, document_id: str, current_tenant_id: str
+    opensearch_client: OpenSearchClient, document_id: str, current_tenant_id: str
 ) -> None:
     opensearch_client.refresh_index()
     query_body = DocumentQuery.delete_from_document_id_query(
@@ -284,10 +283,10 @@ def vespa_document_index(
 def opensearch_client(
     db_session: Session,
     full_deployment_setup: None,  # noqa: ARG001
-) -> Generator[OpenSearchIndexClient, None, None]:
+) -> Generator[OpenSearchClient, None, None]:
     """Creates an OpenSearch client for the test tenant."""
     active = get_active_search_settings(db_session)
-    yield OpenSearchIndexClient(index_name=active.primary.index_name)  # Test runs here.
+    yield OpenSearchClient(index_name=active.primary.index_name)  # Test runs here.
 
 
 @pytest.fixture(scope="module")
@@ -331,7 +330,7 @@ def patch_get_vespa_chunks_page_size() -> Generator[int, None, None]:
 def test_documents(
     db_session: Session,
     vespa_document_index: VespaDocumentIndex,
-    opensearch_client: OpenSearchIndexClient,
+    opensearch_client: OpenSearchClient,
     patch_get_vespa_chunks_page_size: int,
 ) -> Generator[list[Document], None, None]:
     """
@@ -412,7 +411,7 @@ class TestMigrateChunksFromVespaToOpenSearchTask:
         db_session: Session,
         test_documents: list[Document],
         vespa_document_index: VespaDocumentIndex,
-        opensearch_client: OpenSearchIndexClient,
+        opensearch_client: OpenSearchClient,
         test_embedding_dimension: int,
         clean_migration_tables: None,  # noqa: ARG002
         enable_opensearch_indexing_for_onyx: None,  # noqa: ARG002
@@ -481,7 +480,7 @@ class TestMigrateChunksFromVespaToOpenSearchTask:
         db_session: Session,
         test_documents: list[Document],
         vespa_document_index: VespaDocumentIndex,
-        opensearch_client: OpenSearchIndexClient,
+        opensearch_client: OpenSearchClient,
         test_embedding_dimension: int,
         clean_migration_tables: None,  # noqa: ARG002
         enable_opensearch_indexing_for_onyx: None,  # noqa: ARG002
@@ -619,7 +618,7 @@ class TestMigrateChunksFromVespaToOpenSearchTask:
         db_session: Session,
         test_documents: list[Document],
         vespa_document_index: VespaDocumentIndex,
-        opensearch_client: OpenSearchIndexClient,
+        opensearch_client: OpenSearchClient,
         test_embedding_dimension: int,
         clean_migration_tables: None,  # noqa: ARG002
         enable_opensearch_indexing_for_onyx: None,  # noqa: ARG002
@@ -713,7 +712,7 @@ class TestMigrateChunksFromVespaToOpenSearchTask:
         db_session: Session,
         test_documents: list[Document],
         vespa_document_index: VespaDocumentIndex,
-        opensearch_client: OpenSearchIndexClient,
+        opensearch_client: OpenSearchClient,
         test_embedding_dimension: int,
         clean_migration_tables: None,  # noqa: ARG002
         enable_opensearch_indexing_for_onyx: None,  # noqa: ARG002