Address comments

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

.git-blame-ignore-revs

@@ -6,3 +6,4 @@
 
 3134e5f840c12c8f32613ce520101a047c89dcc2  # refactor(whitespace): rm temporary react fragments (#7161)
 ed3f72bc75f3e3a9ae9e4d8cd38278f9c97e78b4  # refactor(whitespace): rm react fragment #7190
+7b927e79c25f4ddfd18a067f489e122acd2c89de  # chore(format): format files where `ruff` and `black` agree (#9339)

.github/CODEOWNERS

@@ -8,3 +8,6 @@
 # Agent context files
 /CLAUDE.md @Weves
 /AGENTS.md @Weves
+
+# Beta cherry-pick workflow owners
+/.github/workflows/post-merge-beta-cherry-pick.yml @justin-tahara @jmelahman

.github/actions/slack-notify/action.yml

@@ -1,11 +1,17 @@
-name: "Slack Notify on Failure"
-description: "Sends a Slack notification when a workflow fails"
+name: "Slack Notify"
+description: "Sends a Slack notification for workflow events"
 inputs:
   webhook-url:
     description: "Slack webhook URL (can also use SLACK_WEBHOOK_URL env var)"
     required: false
+  details:
+    description: "Additional message body content"
+    required: false
   failed-jobs:
-    description: "List of failed job names (newline-separated)"
+    description: "Deprecated alias for details"
+    required: false
+  mention:
+    description: "GitHub username to resolve to a Slack @-mention. Replaces {mention} in details."
     required: false
   title:
     description: "Title for the notification"
@@ -21,7 +27,9 @@ runs:
       shell: bash
       env:
         SLACK_WEBHOOK_URL: ${{ inputs.webhook-url }}
+        DETAILS: ${{ inputs.details }}
         FAILED_JOBS: ${{ inputs.failed-jobs }}
+        MENTION_USER: ${{ inputs.mention }}
         TITLE: ${{ inputs.title }}
         REF_NAME: ${{ inputs.ref-name }}
         REPO: ${{ github.repository }}
@@ -44,6 +52,39 @@ runs:
           REF_NAME="$GITHUB_REF_NAME"
         fi
 
+        if [ -z "$DETAILS" ]; then
+          DETAILS="$FAILED_JOBS"
+        fi
+
+        # Resolve {mention} placeholder if a GitHub username was provided.
+        # Looks up the username in user-mappings.json (co-located with this action)
+        # and replaces {mention} with <@SLACK_ID> for a Slack @-mention.
+        # Falls back to the plain GitHub username if not found in the mapping.
+        if [ -n "$MENTION_USER" ]; then
+          MAPPINGS_FILE="${GITHUB_ACTION_PATH}/user-mappings.json"
+          slack_id="$(jq -r --arg gh "$MENTION_USER" 'to_entries[] | select(.value | ascii_downcase == ($gh | ascii_downcase)) | .key' "$MAPPINGS_FILE" 2>/dev/null | head -1)"
+
+          if [ -n "$slack_id" ]; then
+            mention_text="<@${slack_id}>"
+          else
+            mention_text="${MENTION_USER}"
+          fi
+
+          DETAILS="${DETAILS//\{mention\}/$mention_text}"
+          TITLE="${TITLE//\{mention\}/}"
+        else
+          DETAILS="${DETAILS//\{mention\}/}"
+          TITLE="${TITLE//\{mention\}/}"
+        fi
+
+        normalize_multiline() {
+          printf '%s' "$1" | awk 'BEGIN { ORS=""; first=1 } { if (!first) printf "\\n"; printf "%s", $0; first=0 }'
+        }
+
+        DETAILS="$(normalize_multiline "$DETAILS")"
+        REF_NAME="$(normalize_multiline "$REF_NAME")"
+        TITLE="$(normalize_multiline "$TITLE")"
+
         # Escape JSON special characters
         escape_json() {
           local input="$1"
@@ -59,12 +100,12 @@ runs:
         }
 
         REF_NAME_ESC=$(escape_json "$REF_NAME")
-        FAILED_JOBS_ESC=$(escape_json "$FAILED_JOBS")
+        DETAILS_ESC=$(escape_json "$DETAILS")
         WORKFLOW_URL_ESC=$(escape_json "$WORKFLOW_URL")
         TITLE_ESC=$(escape_json "$TITLE")
 
         # Build JSON payload piece by piece
-        # Note: FAILED_JOBS_ESC already contains \n sequences that should remain as \n in JSON
+        # Note: DETAILS_ESC already contains \n sequences that should remain as \n in JSON
         PAYLOAD="{"
         PAYLOAD="${PAYLOAD}\"text\":\"${TITLE_ESC}\","
         PAYLOAD="${PAYLOAD}\"blocks\":[{"
@@ -79,10 +120,10 @@ runs:
         PAYLOAD="${PAYLOAD}{\"type\":\"mrkdwn\",\"text\":\"*Run ID:*\\n#${RUN_NUMBER}\"}"
         PAYLOAD="${PAYLOAD}]"
         PAYLOAD="${PAYLOAD}}"
-        if [ -n "$FAILED_JOBS" ]; then
+        if [ -n "$DETAILS" ]; then
           PAYLOAD="${PAYLOAD},{"
           PAYLOAD="${PAYLOAD}\"type\":\"section\","
-          PAYLOAD="${PAYLOAD}\"text\":{\"type\":\"mrkdwn\",\"text\":\"*Failed Jobs:*\\n${FAILED_JOBS_ESC}\"}"
+          PAYLOAD="${PAYLOAD}\"text\":{\"type\":\"mrkdwn\",\"text\":\"${DETAILS_ESC}\"}"
           PAYLOAD="${PAYLOAD}}"
         fi
         PAYLOAD="${PAYLOAD},{"
@@ -99,4 +140,3 @@ runs:
         curl -X POST -H 'Content-type: application/json' \
           --data "$PAYLOAD" \
           "$SLACK_WEBHOOK_URL"
-

.github/actions/slack-notify/user-mappings.json

@@ -0,0 +1,18 @@
+{
+  "U05SAGZPEA1": "yuhongsun96",
+  "U05SAH6UGUD": "Weves",
+  "U07PWEQB7A5": "evan-onyx",
+  "U07V1SM68KF": "joachim-danswer",
+  "U08JZ9N3QNN": "raunakab",
+  "U08L24NCLJE": "Subash-Mohan",
+  "U090B9M07B2": "wenxi-onyx",
+  "U094RASDP0Q": "duo-onyx",
+  "U096L8ZQ85B": "justin-tahara",
+  "U09AHV8UBQX": "jessicasingh7",
+  "U09KAL5T3C2": "nmgarza5",
+  "U09KPGVQ70R": "acaprau",
+  "U09QR8KTSJH": "rohoswagger",
+  "U09RB4NTXA4": "jmelahman",
+  "U0A6K9VCY6A": "Danelegend",
+  "U0AGC4KH71A": "Bo-Onyx"
+}

.github/workflows/deployment.yml

@@ -29,20 +29,32 @@ jobs:
       build-backend-craft: ${{ steps.check.outputs.build-backend-craft }}
       build-model-server: ${{ steps.check.outputs.build-model-server }}
       is-cloud-tag: ${{ steps.check.outputs.is-cloud-tag }}
-      is-stable: ${{ steps.check.outputs.is-stable }}
       is-beta: ${{ steps.check.outputs.is-beta }}
-      is-stable-standalone: ${{ steps.check.outputs.is-stable-standalone }}
       is-beta-standalone: ${{ steps.check.outputs.is-beta-standalone }}
-      is-craft-latest: ${{ steps.check.outputs.is-craft-latest }}
+      is-latest: ${{ steps.check.outputs.is-latest }}
       is-test-run: ${{ steps.check.outputs.is-test-run }}
       sanitized-tag: ${{ steps.check.outputs.sanitized-tag }}
       short-sha: ${{ steps.check.outputs.short-sha }}
     steps:
+      - name: Checkout (for git tags)
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v6
+        with:
+          persist-credentials: false
+          fetch-depth: 0
+          fetch-tags: true
+
+      - name: Setup uv
+        uses: astral-sh/setup-uv@37802adc94f370d6bfd71619e3f0bf239e1f3b78 # ratchet:astral-sh/setup-uv@v7
+        with:
+          version: "0.9.9"
+          enable-cache: false
+
       - name: Check which components to build and version info
         id: check
         env:
           EVENT_NAME: ${{ github.event_name }}
         run: |
+          set -eo pipefail
           TAG="${GITHUB_REF_NAME}"
           # Sanitize tag name by replacing slashes with hyphens (for Docker tag compatibility)
           SANITIZED_TAG=$(echo "$TAG" | tr '/' '-')
@@ -54,9 +66,8 @@ jobs:
           IS_VERSION_TAG=false
           IS_STABLE=false
           IS_BETA=false
-          IS_STABLE_STANDALONE=false
           IS_BETA_STANDALONE=false
-          IS_CRAFT_LATEST=false
+          IS_LATEST=false
           IS_PROD_TAG=false
           IS_TEST_RUN=false
           BUILD_DESKTOP=false
@@ -67,9 +78,6 @@ jobs:
           BUILD_MODEL_SERVER=true
 
           # Determine tag type based on pattern matching (do regex checks once)
-          if [[ "$TAG" == craft-* ]]; then
-            IS_CRAFT_LATEST=true
-          fi
           if [[ "$TAG" == *cloud* ]]; then
             IS_CLOUD=true
           fi
@@ -97,20 +105,28 @@ jobs:
             fi
           fi
 
-          # Craft-latest builds backend with Craft enabled
-          if [[ "$IS_CRAFT_LATEST" == "true" ]]; then
-            BUILD_BACKEND_CRAFT=true
-            BUILD_BACKEND=false
-          fi
-
           # Standalone version checks (for backend/model-server - version excluding cloud tags)
-          if [[ "$IS_STABLE" == "true" ]] && [[ "$IS_CLOUD" != "true" ]]; then
-            IS_STABLE_STANDALONE=true
-          fi
           if [[ "$IS_BETA" == "true" ]] && [[ "$IS_CLOUD" != "true" ]]; then
             IS_BETA_STANDALONE=true
           fi
 
+          # Determine if this tag should get the "latest" Docker tag.
+          # Only the highest semver stable tag (vX.Y.Z exactly) gets "latest".
+          if [[ "$IS_STABLE" == "true" ]]; then
+            HIGHEST_STABLE=$(uv run --no-sync --with onyx-devtools ods latest-stable-tag) || {
+              echo "::error::Failed to determine highest stable tag via 'ods latest-stable-tag'"
+              exit 1
+            }
+            if [[ "$TAG" == "$HIGHEST_STABLE" ]]; then
+              IS_LATEST=true
+            fi
+          fi
+
+          # Build craft-latest backend alongside the regular latest.
+          if [[ "$IS_LATEST" == "true" ]]; then
+            BUILD_BACKEND_CRAFT=true
+          fi
+
           # Determine if this is a production tag
           # Production tags are: version tags (v1.2.3*) or nightly tags
           if [[ "$IS_VERSION_TAG" == "true" ]] || [[ "$IS_NIGHTLY" == "true" ]]; then
@@ -129,11 +145,9 @@ jobs:
             echo "build-backend-craft=$BUILD_BACKEND_CRAFT"
             echo "build-model-server=$BUILD_MODEL_SERVER"
             echo "is-cloud-tag=$IS_CLOUD"
-            echo "is-stable=$IS_STABLE"
             echo "is-beta=$IS_BETA"
-            echo "is-stable-standalone=$IS_STABLE_STANDALONE"
             echo "is-beta-standalone=$IS_BETA_STANDALONE"
-            echo "is-craft-latest=$IS_CRAFT_LATEST"
+            echo "is-latest=$IS_LATEST"
             echo "is-test-run=$IS_TEST_RUN"
             echo "sanitized-tag=$SANITIZED_TAG"
             echo "short-sha=$SHORT_SHA"
@@ -151,7 +165,7 @@ jobs:
           fetch-depth: 0
 
       - name: Setup uv
-        uses: astral-sh/setup-uv@61cb8a9741eeb8a550a1b8544337180c0fc8476b # ratchet:astral-sh/setup-uv@v7
+        uses: astral-sh/setup-uv@37802adc94f370d6bfd71619e3f0bf239e1f3b78 # ratchet:astral-sh/setup-uv@v7
         with:
           version: "0.9.9"
           # NOTE: This isn't caching much and zizmor suggests this could be poisoned, so disable.
@@ -293,7 +307,7 @@ jobs:
             xdg-utils
 
       - name: setup node
-        uses: actions/setup-node@6044e13b5dc448c55e2357c09f80417699197238 # ratchet:actions/setup-node@v6.2.0
+        uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # ratchet:actions/setup-node@v6.3.0
         with:
           node-version: 24
           package-manager-cache: false
@@ -441,7 +455,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ needs.determine-builds.outputs.is-test-run == 'true' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
           flavor: |
@@ -515,7 +529,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ needs.determine-builds.outputs.is-test-run == 'true' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
           flavor: |
@@ -593,14 +607,15 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ needs.determine-builds.outputs.is-test-run == 'true' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
           flavor: |
             latest=false
           tags: |
             type=raw,value=${{ needs.determine-builds.outputs.is-test-run == 'true' && format('web-{0}', needs.determine-builds.outputs.sanitized-tag) || github.ref_name }}
-            type=raw,value=${{ needs.determine-builds.outputs.is-test-run != 'true' && needs.determine-builds.outputs.is-stable == 'true' && 'latest' || '' }}
+            type=raw,value=${{ needs.determine-builds.outputs.is-test-run != 'true' && needs.determine-builds.outputs.is-latest == 'true' && 'latest' || '' }}
+            type=raw,value=${{ needs.determine-builds.outputs.is-test-run != 'true' && needs.determine-builds.outputs.is-latest == 'true' && 'craft-latest' || '' }}
             type=raw,value=${{ needs.determine-builds.outputs.is-test-run != 'true' && env.EDGE_TAG == 'true' && 'edge' || '' }}
             type=raw,value=${{ needs.determine-builds.outputs.is-test-run != 'true' && needs.determine-builds.outputs.is-beta == 'true' && 'beta' || '' }}
 
@@ -654,7 +669,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ needs.determine-builds.outputs.is-test-run == 'true' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
           flavor: |
@@ -736,7 +751,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ needs.determine-builds.outputs.is-test-run == 'true' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
           flavor: |
@@ -822,7 +837,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ needs.determine-builds.outputs.is-test-run == 'true' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
           flavor: |
@@ -880,7 +895,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ needs.determine-builds.outputs.is-test-run == 'true' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
           flavor: |
@@ -953,7 +968,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ needs.determine-builds.outputs.is-test-run == 'true' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
           flavor: |
@@ -1030,14 +1045,14 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ needs.determine-builds.outputs.is-test-run == 'true' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
           flavor: |
             latest=false
           tags: |
             type=raw,value=${{ needs.determine-builds.outputs.is-test-run == 'true' && format('backend-{0}', needs.determine-builds.outputs.sanitized-tag) || github.ref_name }}
-            type=raw,value=${{ needs.determine-builds.outputs.is-test-run != 'true' && needs.determine-builds.outputs.is-stable-standalone == 'true' && 'latest' || '' }}
+            type=raw,value=${{ needs.determine-builds.outputs.is-test-run != 'true' && needs.determine-builds.outputs.is-latest == 'true' && 'latest' || '' }}
             type=raw,value=${{ needs.determine-builds.outputs.is-test-run != 'true' && env.EDGE_TAG == 'true' && 'edge' || '' }}
             type=raw,value=${{ needs.determine-builds.outputs.is-test-run != 'true' && needs.determine-builds.outputs.is-beta-standalone == 'true' && 'beta' || '' }}
 
@@ -1091,7 +1106,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ env.REGISTRY_IMAGE }}
           flavor: |
@@ -1164,7 +1179,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ env.REGISTRY_IMAGE }}
           flavor: |
@@ -1242,15 +1257,13 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ env.REGISTRY_IMAGE }}
           flavor: |
             latest=false
           tags: |
             type=raw,value=craft-latest
-            # TODO: Consider aligning craft-latest tags with regular backend builds (e.g., latest, edge, beta)
-            # to keep tagging strategy consistent across all backend images
 
       - name: Create and push manifest
         env:
@@ -1303,7 +1316,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ needs.determine-builds.outputs.is-test-run == 'true' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
           flavor: |
@@ -1383,7 +1396,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ needs.determine-builds.outputs.is-test-run == 'true' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
           flavor: |
@@ -1466,14 +1479,15 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ needs.determine-builds.outputs.is-test-run == 'true' && env.RUNS_ON_ECR_CACHE || env.REGISTRY_IMAGE }}
           flavor: |
             latest=false
           tags: |
             type=raw,value=${{ needs.determine-builds.outputs.is-test-run == 'true' && format('model-server-{0}', needs.determine-builds.outputs.sanitized-tag) || github.ref_name }}
-            type=raw,value=${{ needs.determine-builds.outputs.is-test-run != 'true' && needs.determine-builds.outputs.is-stable-standalone == 'true' && 'latest' || '' }}
+            type=raw,value=${{ needs.determine-builds.outputs.is-test-run != 'true' && needs.determine-builds.outputs.is-latest == 'true' && 'latest' || '' }}
+            type=raw,value=${{ needs.determine-builds.outputs.is-test-run != 'true' && needs.determine-builds.outputs.is-latest == 'true' && 'craft-latest' || '' }}
             type=raw,value=${{ needs.determine-builds.outputs.is-test-run != 'true' && env.EDGE_TAG == 'true' && 'edge' || '' }}
             type=raw,value=${{ needs.determine-builds.outputs.is-test-run != 'true' && needs.determine-builds.outputs.is-beta-standalone == 'true' && 'beta' || '' }}
 

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

@@ -1,67 +1,112 @@
 name: Post-Merge Beta Cherry-Pick
 
 on:
-  push:
-    branches:
-      - main
+  pull_request_target:
+    types:
+      - closed
 
+# SECURITY NOTE:
+# This workflow intentionally uses pull_request_target so post-merge automation can
+# use base-repo credentials. Do not checkout PR head refs in this workflow
+# (e.g. github.event.pull_request.head.sha). Only trusted base refs are allowed.
 permissions:
   contents: read
 
 jobs:
-  cherry-pick-to-latest-release:
-    permissions:
-      contents: write
-      pull-requests: write
+  resolve-cherry-pick-request:
+    if: >-
+      github.event.pull_request.merged == true
+      && github.event.pull_request.base.ref == 'main'
+      && github.event.pull_request.head.repo.full_name == github.repository
     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 }}
+      merge_commit_sha: ${{ steps.gate.outputs.merge_commit_sha }}
+      merged_by: ${{ steps.gate.outputs.merged_by }}
+      gate_error: ${{ steps.gate.outputs.gate_error }}
     runs-on: ubuntu-latest
-    timeout-minutes: 45
+    timeout-minutes: 10
     steps:
       - name: Resolve merged PR and checkbox state
         id: gate
         env:
           GH_TOKEN: ${{ github.token }}
+          PR_NUMBER: ${{ github.event.pull_request.number }}
+          # SECURITY: keep PR body in env/plain-text handling; avoid directly
+          # inlining github.event.pull_request.body into shell commands.
+          PR_BODY: ${{ github.event.pull_request.body }}
+          MERGE_COMMIT_SHA: ${{ github.event.pull_request.merge_commit_sha }}
+          MERGED_BY: ${{ github.event.pull_request.merged_by.login }}
+          # Explicit merger allowlist used because pull_request_target runs with
+          # the default GITHUB_TOKEN, which cannot reliably read org/team
+          # membership for this repository context.
+          ALLOWED_MERGERS: |
+            acaprau
+            bo-onyx
+            danelegend
+            duo-onyx
+            evan-onyx
+            jessicasingh7
+            jmelahman
+            joachim-danswer
+            justin-tahara
+            nmgarza5
+            raunakab
+            rohoswagger
+            subash-mohan
+            trial2onyx
+            wenxi-onyx
+            weves
+            yuhongsun96
         run: |
-          # For the commit that triggered this workflow (HEAD on main), fetch all
-          # associated PRs and keep only the PR that was actually merged into main
-          # with this exact merge commit SHA.
-          pr_numbers="$(gh api "repos/${GITHUB_REPOSITORY}/commits/${GITHUB_SHA}/pulls" | jq -r --arg sha "${GITHUB_SHA}" '.[] | select(.merged_at != null and .base.ref == "main" and .merge_commit_sha == $sha) | .number')"
-          match_count="$(printf '%s\n' "$pr_numbers" | sed '/^[[:space:]]*$/d' | wc -l | tr -d ' ')"
-          pr_number="$(printf '%s\n' "$pr_numbers" | sed '/^[[:space:]]*$/d' | head -n 1)"
+          echo "pr_number=${PR_NUMBER}" >> "$GITHUB_OUTPUT"
+          echo "merged_by=${MERGED_BY}" >> "$GITHUB_OUTPUT"
 
-          if [ "${match_count}" -gt 1 ]; then
-            echo "::warning::Multiple merged PRs matched commit ${GITHUB_SHA}. Using PR #${pr_number}."
-          fi
-
-          if [ -z "$pr_number" ]; then
-            echo "No merged PR associated with commit ${GITHUB_SHA}; skipping."
+          if ! echo "${PR_BODY}" | grep -qiE "\\[x\\][[:space:]]*(\\[[^]]+\\][[:space:]]*)?Please cherry-pick this PR to the latest release version"; then
             echo "should_cherrypick=false" >> "$GITHUB_OUTPUT"
+            echo "Cherry-pick checkbox not checked for PR #${PR_NUMBER}. Skipping."
             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' "$pr_json" | jq -r '.merged_by.login // ""')"
+          # Keep should_cherrypick output before any possible exit 1 below so
+          # notify-slack can still gate on this output even if this job fails.
+          echo "should_cherrypick=true" >> "$GITHUB_OUTPUT"
+          echo "Cherry-pick checkbox checked for PR #${PR_NUMBER}."
 
-          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"
-            echo "Cherry-pick checkbox checked for PR #${pr_number}."
-            exit 0
+          if [ -z "${MERGE_COMMIT_SHA}" ] || [ "${MERGE_COMMIT_SHA}" = "null" ]; then
+            echo "gate_error=missing-merge-commit-sha" >> "$GITHUB_OUTPUT"
+            echo "::error::PR #${PR_NUMBER} requested cherry-pick, but merge_commit_sha is missing."
+            exit 1
           fi
 
-          echo "should_cherrypick=false" >> "$GITHUB_OUTPUT"
-          echo "Cherry-pick checkbox not checked for PR #${pr_number}. Skipping."
+          echo "merge_commit_sha=${MERGE_COMMIT_SHA}" >> "$GITHUB_OUTPUT"
 
+          normalized_merged_by="$(printf '%s' "${MERGED_BY}" | tr '[:upper:]' '[:lower:]')"
+          normalized_allowed_mergers="$(printf '%s\n' "${ALLOWED_MERGERS}" | tr '[:upper:]' '[:lower:]')"
+          if ! printf '%s\n' "${normalized_allowed_mergers}" | grep -Fxq "${normalized_merged_by}"; then
+            echo "gate_error=not-allowed-merger" >> "$GITHUB_OUTPUT"
+            echo "::error::${MERGED_BY} is not in the explicit cherry-pick merger allowlist. Failing cherry-pick gate."
+            exit 1
+          fi
+
+          exit 0
+
+  cherry-pick-to-latest-release:
+    needs:
+      - resolve-cherry-pick-request
+    if: needs.resolve-cherry-pick-request.outputs.should_cherrypick == 'true' && needs.resolve-cherry-pick-request.result == 'success'
+    permissions:
+      contents: write
+      pull-requests: write
+    outputs:
+      cherry_pick_pr_url: ${{ steps.run_cherry_pick.outputs.pr_url }}
+      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:
       - name: Checkout repository
-        if: steps.gate.outputs.should_cherrypick == 'true'
+        # SECURITY: keep checkout pinned to trusted base branch; do not switch to PR head refs.
         uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v6
         with:
           fetch-depth: 0
@@ -69,34 +114,44 @@ jobs:
           ref: main
 
       - name: Install the latest version of uv
-        if: steps.gate.outputs.should_cherrypick == 'true'
-        uses: astral-sh/setup-uv@61cb8a9741eeb8a550a1b8544337180c0fc8476b # ratchet:astral-sh/setup-uv@v7
+        uses: astral-sh/setup-uv@37802adc94f370d6bfd71619e3f0bf239e1f3b78 # ratchet:astral-sh/setup-uv@v7
         with:
           enable-cache: false
           version: "0.9.9"
 
       - name: Configure git identity
-        if: steps.gate.outputs.should_cherrypick == 'true'
         run: |
           git config user.name "github-actions[bot]"
           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 }}
+          CHERRY_PICK_ASSIGNEE: ${{ needs.resolve-cherry-pick-request.outputs.merged_by }}
+          MERGE_COMMIT_SHA: ${{ needs.resolve-cherry-pick-request.outputs.merge_commit_sha }}
         run: |
-          set -o pipefail
           output_file="$(mktemp)"
-          uv run --no-sync --with onyx-devtools ods cherry-pick "${GITHUB_SHA}" --yes --no-verify 2>&1 | tee "$output_file"
-          exit_code="${PIPESTATUS[0]}"
+          set +e
+          uv run --no-sync --with onyx-devtools ods cherry-pick "${MERGE_COMMIT_SHA}" --yes --no-verify 2>&1 | tee "$output_file"
+          pipe_statuses=("${PIPESTATUS[@]}")
+          exit_code="${pipe_statuses[0]}"
+          tee_exit="${pipe_statuses[1]:-0}"
+          set -e
+          if [ "${tee_exit}" -ne 0 ]; then
+            echo "status=failure" >> "$GITHUB_OUTPUT"
+            echo "reason=output-capture-failed" >> "$GITHUB_OUTPUT"
+            echo "::error::tee failed to capture cherry-pick output (exit ${tee_exit}); cannot classify result."
+            exit 1
+          fi
 
           if [ "${exit_code}" -eq 0 ]; then
+            pr_url="$(sed -n 's/^.*PR created successfully: \(https:\/\/github\.com\/[^[:space:]]\+\/pull\/[0-9]\+\).*$/\1/p' "$output_file" | tail -n 1)"
             echo "status=success" >> "$GITHUB_OUTPUT"
+            if [ -n "${pr_url}" ]; then
+              echo "pr_url=${pr_url}" >> "$GITHUB_OUTPUT"
+            fi
             exit 0
           fi
 
@@ -115,17 +170,18 @@ jobs:
           } >> "$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'
+        if: steps.run_cherry_pick.outputs.status == 'failure'
         env:
           CHERRY_PICK_REASON: ${{ steps.run_cherry_pick.outputs.reason }}
         run: |
           echo "::error::Automated cherry-pick failed (${CHERRY_PICK_REASON})."
           exit 1
 
-  notify-slack-on-cherry-pick-failure:
+  notify-slack-on-cherry-pick-success:
     needs:
+      - resolve-cherry-pick-request
       - 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'
+    if: needs.resolve-cherry-pick-request.outputs.should_cherrypick == 'true' && needs.resolve-cherry-pick-request.result == 'success' && needs.cherry-pick-to-latest-release.result == 'success'
     runs-on: ubuntu-slim
     timeout-minutes: 10
     steps:
@@ -134,32 +190,107 @@ jobs:
         with:
           persist-credentials: false
 
+      - name: Fail if Slack webhook secret is missing
+        env:
+          CHERRY_PICK_PRS_WEBHOOK: ${{ secrets.CHERRY_PICK_PRS_WEBHOOK }}
+        run: |
+          if [ -z "${CHERRY_PICK_PRS_WEBHOOK}" ]; then
+            echo "::error::CHERRY_PICK_PRS_WEBHOOK is not configured."
+            exit 1
+          fi
+
+      - name: Build cherry-pick success summary
+        id: success-summary
+        env:
+          SOURCE_PR_NUMBER: ${{ needs.resolve-cherry-pick-request.outputs.pr_number }}
+          MERGE_COMMIT_SHA: ${{ needs.resolve-cherry-pick-request.outputs.merge_commit_sha }}
+          CHERRY_PICK_PR_URL: ${{ needs.cherry-pick-to-latest-release.outputs.cherry_pick_pr_url }}
+        run: |
+          source_pr_url="https://github.com/${GITHUB_REPOSITORY}/pull/${SOURCE_PR_NUMBER}"
+          details="*Cherry-pick PR opened successfully.*\\n• author: {mention}\\n• source PR: ${source_pr_url}"
+          if [ -n "${CHERRY_PICK_PR_URL}" ]; then
+            details="${details}\\n• cherry-pick PR: ${CHERRY_PICK_PR_URL}"
+          fi
+          if [ -n "${MERGE_COMMIT_SHA}" ]; then
+            details="${details}\\n• merge SHA: ${MERGE_COMMIT_SHA}"
+          fi
+
+          echo "details=${details}" >> "$GITHUB_OUTPUT"
+
+      - name: Notify #cherry-pick-prs about cherry-pick success
+        uses: ./.github/actions/slack-notify
+        with:
+          webhook-url: ${{ secrets.CHERRY_PICK_PRS_WEBHOOK }}
+          mention: ${{ needs.resolve-cherry-pick-request.outputs.merged_by }}
+          details: ${{ steps.success-summary.outputs.details }}
+          title: "✅ Automated Cherry-Pick PR Opened"
+          ref-name: ${{ github.event.pull_request.base.ref }}
+
+  notify-slack-on-cherry-pick-failure:
+    needs:
+      - resolve-cherry-pick-request
+      - cherry-pick-to-latest-release
+    if: always() && needs.resolve-cherry-pick-request.outputs.should_cherrypick == 'true' && (needs.resolve-cherry-pick-request.result == 'failure' || needs.cherry-pick-to-latest-release.result == 'failure')
+    runs-on: ubuntu-slim
+    timeout-minutes: 10
+    steps:
+      - name: Checkout
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v6
+        with:
+          persist-credentials: false
+
+      - name: Fail if Slack webhook secret is missing
+        env:
+          CHERRY_PICK_PRS_WEBHOOK: ${{ secrets.CHERRY_PICK_PRS_WEBHOOK }}
+        run: |
+          if [ -z "${CHERRY_PICK_PRS_WEBHOOK}" ]; then
+            echo "::error::CHERRY_PICK_PRS_WEBHOOK is not configured."
+            exit 1
+          fi
+
       - name: Build cherry-pick failure summary
         id: failure-summary
         env:
-          SOURCE_PR_NUMBER: ${{ needs.cherry-pick-to-latest-release.outputs.pr_number }}
+          SOURCE_PR_NUMBER: ${{ needs.resolve-cherry-pick-request.outputs.pr_number }}
+          MERGE_COMMIT_SHA: ${{ needs.resolve-cherry-pick-request.outputs.merge_commit_sha }}
+          GATE_ERROR: ${{ needs.resolve-cherry-pick-request.outputs.gate_error }}
           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 [ "${CHERRY_PICK_REASON}" = "merge-conflict" ]; then
+          if [ "${GATE_ERROR}" = "missing-merge-commit-sha" ]; then
+            reason_text="requested cherry-pick but merge commit SHA was missing"
+          elif [ "${GATE_ERROR}" = "not-allowed-merger" ]; then
+            reason_text="merger is not in the explicit cherry-pick allowlist"
+          elif [ "${CHERRY_PICK_REASON}" = "output-capture-failed" ]; then
+            reason_text="failed to capture cherry-pick output for classification"
+          elif [ "${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: ${reason_text}"
+          if [ -n "${GATE_ERROR}" ]; then
+            failed_job_label="resolve-cherry-pick-request"
+          else
+            failed_job_label="cherry-pick-to-latest-release"
+          fi
+          details="• author: {mention}\\n• ${failed_job_label}\\n• source PR: ${source_pr_url}\\n• reason: ${reason_text}"
+          if [ -n "${MERGE_COMMIT_SHA}" ]; then
+            details="${details}\\n• merge SHA: ${MERGE_COMMIT_SHA}"
+          fi
           if [ -n "${details_excerpt}" ]; then
-            failed_jobs="${failed_jobs}\\n• excerpt: ${details_excerpt}"
+            details="${details}\\n• excerpt: ${details_excerpt}"
           fi
 
-          echo "jobs=${failed_jobs}" >> "$GITHUB_OUTPUT"
+          echo "details=${details}" >> "$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 }}
+          mention: ${{ needs.resolve-cherry-pick-request.outputs.merged_by }}
+          details: ${{ steps.failure-summary.outputs.details }}
           title: "🚨 Automated Cherry-Pick Failed"
-          ref-name: ${{ github.ref_name }}
+          ref-name: ${{ github.event.pull_request.base.ref }}

.github/workflows/pr-desktop-build.yml

@@ -50,7 +50,7 @@ jobs:
           persist-credentials: false
 
       - name: Setup node
-        uses: actions/setup-node@6044e13b5dc448c55e2357c09f80417699197238
+        uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f
         with:
           node-version: 24
           cache: "npm" # zizmor: ignore[cache-poisoning]
@@ -105,7 +105,7 @@ jobs:
 
       - name: Upload build artifacts
         if: always()
-        uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         with:
           name: desktop-build-${{ matrix.platform }}-${{ github.run_id }}
           path: |

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

@@ -7,6 +7,15 @@ on:
   merge_group:
   pull_request:
     branches: [main]
+    paths:
+      - "backend/**"
+      - "pyproject.toml"
+      - "uv.lock"
+      - ".github/workflows/pr-external-dependency-unit-tests.yml"
+      - ".github/actions/setup-python-and-install-dependencies/**"
+      - ".github/actions/setup-playwright/**"
+      - "deployment/docker_compose/docker-compose.yml"
+      - "deployment/docker_compose/docker-compose.dev.yml"
   push:
     tags:
       - "v*.*.*"
@@ -174,7 +183,7 @@ jobs:
 
       - name: Upload Docker logs
         if: failure()
-        uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         with:
           name: docker-logs-${{ matrix.test-dir }}
           path: docker-logs/

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

@@ -25,7 +25,7 @@ jobs:
     outputs:
       modules: ${{ steps.set-modules.outputs.modules }}
     steps:
-      - uses: actions/checkout@8e8c483db84b4bee98b60c0593521ed34d9990e8
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd
         with:
           persist-credentials: false
       - id: set-modules
@@ -39,7 +39,7 @@ jobs:
       matrix:
         modules: ${{ fromJSON(needs.detect-modules.outputs.modules) }}
     steps:
-      - uses: actions/checkout@8e8c483db84b4bee98b60c0593521ed34d9990e8 # ratchet:actions/checkout@v6
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v6
         with:
           persist-credentials: false
       - uses: actions/setup-go@4dc6199c7b1a012772edbd06daecab0f50c9053c # zizmor: ignore[cache-poisoning]

.github/workflows/pr-helm-chart-testing.yml

@@ -133,7 +133,7 @@ jobs:
           echo "=== Validating chart dependencies ==="
           cd deployment/helm/charts/onyx
           helm dependency update
-          helm lint .
+          helm lint . --set auth.userauth.values.user_auth_secret=placeholder
 
       - name: Run chart-testing (install) with enhanced monitoring
         timeout-minutes: 25
@@ -194,6 +194,7 @@ jobs:
               --set=vespa.enabled=false \
               --set=opensearch.enabled=true \
               --set=auth.opensearch.enabled=true \
+              --set=auth.userauth.values.user_auth_secret=test-secret \
               --set=slackbot.enabled=false \
               --set=postgresql.enabled=true \
               --set=postgresql.cluster.storage.storageClass=standard \
@@ -230,6 +231,10 @@ jobs:
         if: steps.list-changed.outputs.changed == 'true'
         run: |
           echo "=== Post-install verification ==="
+          if ! kubectl cluster-info >/dev/null 2>&1; then
+            echo "ERROR: Kubernetes cluster is not reachable after install"
+            exit 1
+          fi
           kubectl get pods --all-namespaces
           kubectl get services --all-namespaces
           # Only show issues if they exist
@@ -239,6 +244,10 @@ jobs:
         if: failure() && steps.list-changed.outputs.changed == 'true'
         run: |
           echo "=== Cleanup on failure ==="
+          if ! kubectl cluster-info >/dev/null 2>&1; then
+            echo "Skipping failure cleanup: Kubernetes cluster is not reachable"
+            exit 0
+          fi
           echo "=== Final cluster state ==="
           kubectl get pods --all-namespaces
           kubectl get events --all-namespaces --sort-by=.lastTimestamp | tail -10

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

@@ -316,6 +316,7 @@ jobs:
           # Base config shared by both editions
           cat <<EOF > deployment/docker_compose/.env
           COMPOSE_PROFILES=s3-filestore
+          OPENSEARCH_FOR_ONYX_ENABLED=false
           AUTH_TYPE=basic
           POSTGRES_POOL_PRE_PING=true
           POSTGRES_USE_NULL_POOL=true
@@ -418,6 +419,7 @@ jobs:
               -e POSTGRES_POOL_PRE_PING=true \
               -e POSTGRES_USE_NULL_POOL=true \
               -e VESPA_HOST=index \
+              -e ENABLE_OPENSEARCH_INDEXING_FOR_ONYX=false \
               -e REDIS_HOST=cache \
               -e API_SERVER_HOST=api_server \
               -e OPENAI_API_KEY=${OPENAI_API_KEY} \
@@ -464,7 +466,7 @@ jobs:
 
       - name: Upload logs
         if: always()
-        uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         with:
           name: docker-all-logs-${{ matrix.edition }}-${{ matrix.test-dir.name }}
           path: ${{ github.workspace }}/docker-compose.log
@@ -585,7 +587,7 @@ jobs:
 
       - name: Upload logs (onyx-lite)
         if: always()
-        uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         with:
           name: docker-all-logs-onyx-lite
           path: ${{ github.workspace }}/docker-compose-onyx-lite.log
@@ -637,6 +639,7 @@ jobs:
           ONYX_BACKEND_IMAGE=${ECR_CACHE}:integration-test-backend-test-${RUN_ID} \
           ONYX_MODEL_SERVER_IMAGE=${ECR_CACHE}:integration-test-model-server-test-${RUN_ID} \
           DEV_MODE=true \
+          OPENSEARCH_FOR_ONYX_ENABLED=false \
           docker compose -f docker-compose.multitenant-dev.yml up \
             relational_db \
             index \
@@ -691,6 +694,7 @@ jobs:
             -e POSTGRES_DB=postgres \
             -e POSTGRES_USE_NULL_POOL=true \
             -e VESPA_HOST=index \
+            -e ENABLE_OPENSEARCH_INDEXING_FOR_ONYX=false \
             -e REDIS_HOST=cache \
             -e API_SERVER_HOST=api_server \
             -e OPENAI_API_KEY=${OPENAI_API_KEY} \
@@ -721,7 +725,7 @@ jobs:
 
       - name: Upload logs (multi-tenant)
         if: always()
-        uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         with:
           name: docker-all-logs-multitenant
           path: ${{ github.workspace }}/docker-compose-multitenant.log

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

@@ -28,7 +28,7 @@ jobs:
           persist-credentials: false
 
       - name: Setup node
-        uses: actions/setup-node@6044e13b5dc448c55e2357c09f80417699197238 # ratchet:actions/setup-node@v4
+        uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # ratchet:actions/setup-node@v4
         with:
           node-version: 22
           cache: "npm" # zizmor: ignore[cache-poisoning] test-only workflow; no deploy artifacts
@@ -44,7 +44,7 @@ jobs:
 
       - name: Upload coverage reports
         if: always()
-        uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         with:
           name: jest-coverage-${{ github.run_id }}
           path: ./web/coverage

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

@@ -12,6 +12,9 @@ on:
   push:
     tags:
       - "v*.*.*"
+    # TODO: Remove this if we enable merge-queues for release branches.
+    branches:
+      - "release/**"
 
 permissions:
   contents: read
@@ -269,7 +272,7 @@ jobs:
 
       - name: Setup node
         # zizmor: ignore[cache-poisoning] ephemeral runners; no release artifacts
-        uses: actions/setup-node@6044e13b5dc448c55e2357c09f80417699197238 # ratchet:actions/setup-node@v4
+        uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # ratchet:actions/setup-node@v4
         with:
           node-version: 22
           cache: "npm" # zizmor: ignore[cache-poisoning]
@@ -442,7 +445,7 @@ jobs:
         run: |
           npx playwright test --project ${PROJECT}
 
-      - uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+      - uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         if: always()
         with:
           # Includes test results and trace.zip files
@@ -451,7 +454,7 @@ jobs:
           retention-days: 30
 
       - name: Upload screenshots
-        uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         if: always()
         with:
           name: playwright-screenshots-${{ matrix.project }}-${{ github.run_id }}
@@ -468,7 +471,7 @@ jobs:
 
       - name: Install the latest version of uv
         if: always()
-        uses: astral-sh/setup-uv@61cb8a9741eeb8a550a1b8544337180c0fc8476b # ratchet:astral-sh/setup-uv@v7
+        uses: astral-sh/setup-uv@37802adc94f370d6bfd71619e3f0bf239e1f3b78 # ratchet:astral-sh/setup-uv@v7
         with:
           enable-cache: false
           version: "0.9.9"
@@ -531,7 +534,7 @@ jobs:
             "s3://${PLAYWRIGHT_S3_BUCKET}/reports/pr-${PR_NUMBER}/${RUN_ID}/${PROJECT}/"
 
       - name: Upload visual diff summary
-        uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         if: always()
         with:
           name: screenshot-diff-summary-${{ matrix.project }}
@@ -540,7 +543,7 @@ jobs:
           retention-days: 5
 
       - name: Upload visual diff report artifact
-        uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         if: always()
         with:
           name: screenshot-diff-report-${{ matrix.project }}-${{ github.run_id }}
@@ -587,7 +590,7 @@ jobs:
 
       - name: Upload logs
         if: success() || failure()
-        uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         with:
           name: docker-logs-${{ matrix.project }}-${{ github.run_id }}
           path: ${{ github.workspace }}/docker-compose.log
@@ -611,7 +614,7 @@ jobs:
 
       - name: Setup node
         # zizmor: ignore[cache-poisoning] ephemeral runners; no release artifacts
-        uses: actions/setup-node@6044e13b5dc448c55e2357c09f80417699197238 # ratchet:actions/setup-node@v4
+        uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # ratchet:actions/setup-node@v4
         with:
           node-version: 22
           cache: "npm" # zizmor: ignore[cache-poisoning]
@@ -671,7 +674,7 @@ jobs:
         working-directory: ./web
         run: npx playwright test --project lite
 
-      - uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+      - uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         if: always()
         with:
           name: playwright-test-results-lite-${{ github.run_id }}
@@ -689,7 +692,7 @@ jobs:
 
       - name: Upload logs
         if: success() || failure()
-        uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         with:
           name: docker-logs-lite-${{ github.run_id }}
           path: ${{ github.workspace }}/docker-compose.log
@@ -707,7 +710,7 @@ jobs:
       pull-requests: write
     steps:
       - name: Download visual diff summaries
-        uses: actions/download-artifact@37930b1c2abaa49bbe596cd826c3c89aef350131
+        uses: actions/download-artifact@70fc10c6e5e1ce46ad2ea6f2b72d43f7d47b13c3
         with:
           pattern: screenshot-diff-summary-*
           path: summaries/

.github/workflows/pr-python-connector-tests.yml

@@ -7,6 +7,13 @@ on:
   merge_group:
   pull_request:
     branches: [main]
+    paths:
+      - "backend/**"
+      - "pyproject.toml"
+      - "uv.lock"
+      - ".github/workflows/pr-python-connector-tests.yml"
+      - ".github/actions/setup-python-and-install-dependencies/**"
+      - ".github/actions/setup-playwright/**"
   push:
     tags:
       - "v*.*.*"

.github/workflows/pr-python-model-tests.yml

@@ -73,7 +73,7 @@ jobs:
         uses: docker/setup-buildx-action@8d2750c68a42422c14e847fe6c8ac0403b4cbd6f
 
       - name: Build and load
-        uses: docker/bake-action@5be5f02ff8819ecd3092ea6b2e6261c31774f2b4 # ratchet:docker/bake-action@v6
+        uses: docker/bake-action@82490499d2e5613fcead7e128237ef0b0ea210f7 # ratchet:docker/bake-action@v7.0.0
         env:
           TAG: model-server-${{ github.run_id }}
         with:
@@ -122,7 +122,7 @@ jobs:
 
       - name: Upload logs
         if: always()
-        uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         with:
           name: docker-all-logs
           path: ${{ github.workspace }}/docker-compose.log

.github/workflows/pr-quality-checks.yml

@@ -28,9 +28,9 @@ jobs:
         with:
           python-version: "3.11"
       - name: Setup Terraform
-        uses: hashicorp/setup-terraform@b9cd54a3c349d3f38e8881555d616ced269862dd # ratchet:hashicorp/setup-terraform@v3
+        uses: hashicorp/setup-terraform@5e8dbf3c6d9deaf4193ca7a8fb23f2ac83bb6c85 # ratchet:hashicorp/setup-terraform@v4.0.0
       - name: Setup node
-        uses: actions/setup-node@6044e13b5dc448c55e2357c09f80417699197238 # ratchet:actions/setup-node@v6
+        uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # ratchet:actions/setup-node@v6
         with: # zizmor: ignore[cache-poisoning]
           node-version: 22
           cache: "npm"

.github/workflows/preview.yml

@@ -22,7 +22,7 @@ jobs:
           persist-credentials: false
 
       - name: Setup node
-        uses: actions/setup-node@6044e13b5dc448c55e2357c09f80417699197238 # ratchet:actions/setup-node@v4
+        uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # ratchet:actions/setup-node@v4
         with:
           node-version: 22
           cache: "npm"

.github/workflows/release-cli.yml

@@ -26,7 +26,7 @@ jobs:
       - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v6
         with:
           persist-credentials: false
-      - uses: astral-sh/setup-uv@5a095e7a2014a4212f075830d4f7277575a9d098 # ratchet:astral-sh/setup-uv@v7
+      - uses: astral-sh/setup-uv@37802adc94f370d6bfd71619e3f0bf239e1f3b78 # ratchet:astral-sh/setup-uv@v7
         with:
           enable-cache: false
           version: "0.9.9"

.github/workflows/release-devtools.yml

@@ -26,7 +26,7 @@ jobs:
       - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v6
         with:
           persist-credentials: false
-      - uses: astral-sh/setup-uv@61cb8a9741eeb8a550a1b8544337180c0fc8476b # ratchet:astral-sh/setup-uv@v7
+      - uses: astral-sh/setup-uv@37802adc94f370d6bfd71619e3f0bf239e1f3b78 # ratchet:astral-sh/setup-uv@v7
         with:
           enable-cache: false
           version: "0.9.9"

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

@@ -319,7 +319,7 @@ jobs:
 
       - name: Upload logs
         if: always()
-        uses: actions/upload-artifact@b7c566a772e6b6bfb58ed0dc250532a479d7789f
+        uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f
         with:
           name: docker-all-logs-nightly-${{ matrix.provider }}-llm-provider
           path: |

.github/workflows/sandbox-deployment.yml

@@ -125,7 +125,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ env.REGISTRY_IMAGE }}
           flavor: |
@@ -195,7 +195,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ env.REGISTRY_IMAGE }}
           flavor: |
@@ -268,7 +268,7 @@ jobs:
 
       - name: Docker meta
         id: meta
-        uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # ratchet:docker/metadata-action@v5
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # ratchet:docker/metadata-action@v6.0.0
         with:
           images: ${{ env.REGISTRY_IMAGE }}
           flavor: |

.github/workflows/storybook-deploy.yml

@@ -0,0 +1,69 @@
+name: Storybook Deploy
+env:
+  VERCEL_ORG_ID: ${{ secrets.VERCEL_ORG_ID }}
+  VERCEL_PROJECT_ID: prj_sG49mVsA25UsxIPhN2pmBJlikJZM
+  VERCEL_CLI: vercel@50.14.1
+  VERCEL_TOKEN: ${{ secrets.VERCEL_TOKEN }}
+
+concurrency:
+  group: storybook-deploy-production
+  cancel-in-progress: true
+
+on:
+  workflow_dispatch:
+  push:
+    branches:
+      - main
+    paths:
+      - "web/lib/opal/**"
+      - "web/src/refresh-components/**"
+      - "web/.storybook/**"
+      - "web/package.json"
+      - "web/package-lock.json"
+permissions:
+  contents: read
+jobs:
+  Deploy-Storybook:
+    runs-on: ubuntu-latest
+    timeout-minutes: 30
+    steps:
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v4
+        with:
+          persist-credentials: false
+
+      - name: Setup node
+        uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # ratchet:actions/setup-node@v4
+        with:
+          node-version: 22
+          cache: "npm"
+          cache-dependency-path: ./web/package-lock.json
+
+      - name: Install dependencies
+        working-directory: web
+        run: npm ci
+
+      - name: Build Storybook
+        working-directory: web
+        run: npm run storybook:build
+
+      - name: Deploy to Vercel (Production)
+        working-directory: web
+        run: npx --yes "$VERCEL_CLI" deploy storybook-static/ --prod --yes --token="$VERCEL_TOKEN"
+
+  notify-slack-on-failure:
+    needs: Deploy-Storybook
+    if: always() && needs.Deploy-Storybook.result == 'failure'
+    runs-on: ubuntu-latest
+    timeout-minutes: 10
+    steps:
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # ratchet:actions/checkout@v4
+        with:
+          persist-credentials: false
+          sparse-checkout: .github/actions/slack-notify
+
+      - name: Send Slack notification
+        uses: ./.github/actions/slack-notify
+        with:
+          webhook-url: ${{ secrets.MONITOR_DEPLOYMENTS_WEBHOOK }}
+          failed-jobs: "• Deploy-Storybook"
+          title: "🚨 Storybook Deploy Failed"

.github/workflows/zizmor.yml

@@ -24,7 +24,7 @@ jobs:
           persist-credentials: false
 
       - name: Install the latest version of uv
-        uses: astral-sh/setup-uv@61cb8a9741eeb8a550a1b8544337180c0fc8476b # ratchet:astral-sh/setup-uv@v7
+        uses: astral-sh/setup-uv@37802adc94f370d6bfd71619e3f0bf239e1f3b78 # ratchet:astral-sh/setup-uv@v7
         with:
           enable-cache: false
           version: "0.9.9"

.greptile/config.json

@@ -0,0 +1,64 @@
+{
+    "labels": [],
+    "comment": "",
+    "fixWithAI": true,
+    "hideFooter": false,
+    "strictness": 3,
+    "statusCheck": true,
+    "commentTypes": [
+      "logic",
+      "syntax",
+      "style"
+    ],
+    "instructions": "",
+    "disabledLabels": [],
+    "excludeAuthors": [
+      "dependabot[bot]",
+      "renovate[bot]"
+    ],
+    "ignoreKeywords": "",
+    "ignorePatterns": "",
+    "includeAuthors": [],
+    "summarySection": {
+      "included": true,
+      "collapsible": false,
+      "defaultOpen": false
+    },
+    "excludeBranches": [],
+    "fileChangeLimit": 300,
+    "includeBranches": [],
+    "includeKeywords": "",
+    "triggerOnUpdates": true,
+    "updateExistingSummaryComment": true,
+    "updateSummaryOnly": false,
+    "issuesTableSection": {
+      "included": true,
+      "collapsible": false,
+      "defaultOpen": false
+    },
+    "statusCommentsEnabled": true,
+    "confidenceScoreSection": {
+      "included": true,
+      "collapsible": false
+    },
+    "sequenceDiagramSection": {
+      "included": true,
+      "collapsible": false,
+      "defaultOpen": false
+    },
+    "shouldUpdateDescription": false,
+    "rules": [
+      {
+        "scope": ["web/**"],
+        "rule": "In Onyx's Next.js app, the `app/ee/admin/` directory is a filesystem convention for Enterprise Edition route overrides — it does NOT add an `/ee/` prefix to the URL. Both `app/admin/groups/page.tsx` and `app/ee/admin/groups/page.tsx` serve the same URL `/admin/groups`. Hardcoded `/admin/...` paths in router.push() calls are correct and do NOT break EE deployments. Do not flag hardcoded admin paths as bugs."
+      },
+      {
+        "scope": ["web/**"],
+        "rule": "In Onyx, each API key creates a unique user row in the database with a unique `user_id` (UUID). There is a 1:1 mapping between API keys and their backing user records. Multiple API keys do NOT share the same `user_id`. Do not flag potential duplicate row IDs when using `user_id` from API key descriptors."
+      },
+      {
+        "scope": ["backend/**/*.py"],
+        "rule": "Never raise HTTPException directly in business code. Use `raise OnyxError(OnyxErrorCode.XXX, \"message\")` from `onyx.error_handling.exceptions`. A global FastAPI exception handler converts OnyxError into structured JSON responses with {\"error_code\": \"...\", \"detail\": \"...\"}. Error codes are defined in `onyx.error_handling.error_codes.OnyxErrorCode`. For upstream errors with dynamic HTTP status codes, use `status_code_override`: `raise OnyxError(OnyxErrorCode.BAD_GATEWAY, detail, status_code_override=upstream_status)`."
+      }
+    ]
+}

.greptile/files.json

@@ -0,0 +1,57 @@
+[
+  {
+    "scope": [],
+    "path": "contributing_guides/best_practices.md",
+    "description": "Best practices for contributing to the codebase"
+  },
+  {
+    "scope": ["web/**"],
+    "path": "web/AGENTS.md",
+    "description": "Frontend coding standards for the web directory"
+  },
+  {
+    "scope": ["web/**"],
+    "path": "web/tests/README.md",
+    "description": "Frontend testing guide and conventions"
+  },
+  {
+    "scope": ["web/**"],
+    "path": "web/CLAUDE.md",
+    "description": "Single source of truth for frontend coding standards"
+  },
+  {
+    "scope": ["web/**"],
+    "path": "web/lib/opal/README.md",
+    "description": "Opal component library usage guide"
+  },
+  {
+    "scope": ["backend/**"],
+    "path": "backend/tests/README.md",
+    "description": "Backend testing guide covering all 4 test types, fixtures, and conventions"
+  },
+  {
+    "scope": ["backend/onyx/connectors/**"],
+    "path": "backend/onyx/connectors/README.md",
+    "description": "Connector development guide covering design, interfaces, and required changes"
+  },
+  {
+    "scope": [],
+    "path": "CLAUDE.md",
+    "description": "Project instructions and coding standards"
+  },
+  {
+    "scope": [],
+    "path": "backend/alembic/README.md",
+    "description": "Migration guidance, including multi-tenant migration behavior"
+  },
+  {
+    "scope": [],
+    "path": "deployment/helm/charts/onyx/values-lite.yaml",
+    "description": "Lite deployment Helm values and service assumptions"
+  },
+  {
+    "scope": [],
+    "path": "deployment/docker_compose/docker-compose.onyx-lite.yml",
+    "description": "Lite deployment Docker Compose overlay and disabled service behavior"
+  }
+]

.greptile/rules.md

@@ -0,0 +1,29 @@
+# Greptile Review Rules
+
+## Type Annotations
+
+Use explicit type annotations for variables to enhance code clarity, especially when moving type hints around in the code.
+
+## Best Practices
+
+Use `contributing_guides/best_practices.md` as core review context. Prefer consistency with existing patterns, fix issues in code you touch, avoid tacking new features onto muddy interfaces, fail loudly instead of silently swallowing errors, keep code strictly typed, preserve clear state boundaries, remove duplicate or dead logic, break up overly long functions, avoid hidden import-time side effects, respect module boundaries, and favor correctness-by-construction over relying on callers to use an API correctly.
+
+## TODOs
+
+Whenever a TODO is added, there must always be an associated name or ticket with that TODO in the style of `TODO(name): ...` or `TODO(1234): ...`
+
+## Debugging Code
+
+Remove temporary debugging code before merging to production, especially tenant-specific debugging logs.
+
+## Hardcoded Booleans
+
+When hardcoding a boolean variable to a constant value, remove the variable entirely and clean up all places where it's used rather than just setting it to a constant.
+
+## Multi-tenant vs Single-tenant
+
+Code changes must consider both multi-tenant and single-tenant deployments. In multi-tenant mode, preserve tenant isolation, ensure tenant context is propagated correctly, and avoid assumptions that only hold for a single shared schema or globally shared state. In single-tenant mode, avoid introducing unnecessary tenant-specific requirements or cloud-only control-plane dependencies.
+
+## Full vs Lite Deployments
+
+Code changes must consider both regular Onyx deployments and Onyx lite deployments. Lite deployments disable the vector DB, Redis, model servers, and background workers by default, use PostgreSQL-backed cache/auth/file storage, and rely on the API server to handle background work. Do not assume those services are available unless the code path is explicitly limited to full deployments.

.vscode/env_template.txt

@@ -7,6 +7,9 @@
 
 
 AUTH_TYPE=basic
+# Recommended for basic auth - used for signing password reset and verification tokens
+# Generate a secure value with: openssl rand -hex 32
+USER_AUTH_SECRET=""
 DEV_MODE=true
 
 

.vscode/launch.json

@@ -117,7 +117,8 @@
       "presentation": {
         "group": "2"
       },
-      "consoleTitle": "API Server Console"
+      "consoleTitle": "API Server Console",
+      "justMyCode": false
     },
     {
       "name": "Slack Bot",
@@ -268,7 +269,8 @@
       "presentation": {
         "group": "2"
       },
-      "consoleTitle": "Celery heavy Console"
+      "consoleTitle": "Celery heavy Console",
+      "justMyCode": false
     },
     {
       "name": "Celery kg_processing",
@@ -355,7 +357,8 @@
       "presentation": {
         "group": "2"
       },
-      "consoleTitle": "Celery user_file_processing Console"
+      "consoleTitle": "Celery user_file_processing Console",
+      "justMyCode": false
     },
     {
       "name": "Celery docfetching",
@@ -413,7 +416,8 @@
       "presentation": {
         "group": "2"
       },
-      "consoleTitle": "Celery docprocessing Console"
+      "consoleTitle": "Celery docprocessing Console",
+      "justMyCode": false
     },
     {
       "name": "Celery beat",

AGENTS.md

@@ -167,284 +167,7 @@ web/
 
 ## Frontend Standards
 
-### 1. Import Standards
-
-**Always use absolute imports with the `@` prefix.**
-
-**Reason:** Moving files around becomes easier since you don't also have to update those import statements. This makes modifications to the codebase much nicer.
-
-```typescript
-// ✅ Good
-import { Button } from "@/components/ui/button";
-import { useAuth } from "@/hooks/useAuth";
-import { Text } from "@/refresh-components/texts/Text";
-
-// ❌ Bad
-import { Button } from "../../../components/ui/button";
-import { useAuth } from "./hooks/useAuth";
-```
-
-### 2. React Component Functions
-
-**Prefer regular functions over arrow functions for React components.**
-
-**Reason:** Functions just become easier to read.
-
-```typescript
-// ✅ Good
-function UserProfile({ userId }: UserProfileProps) {
-  return <div>User Profile</div>
-}
-
-// ❌ Bad
-const UserProfile = ({ userId }: UserProfileProps) => {
-  return <div>User Profile</div>
-}
-```
-
-### 3. Props Interface Extraction
-
-**Extract prop types into their own interface definitions.**
-
-**Reason:** Functions just become easier to read.
-
-```typescript
-// ✅ Good
-interface UserCardProps {
-  user: User
-  showActions?: boolean
-  onEdit?: (userId: string) => void
-}
-
-function UserCard({ user, showActions = false, onEdit }: UserCardProps) {
-  return <div>User Card</div>
-}
-
-// ❌ Bad
-function UserCard({
-  user,
-  showActions = false,
-  onEdit
-}: {
-  user: User
-  showActions?: boolean
-  onEdit?: (userId: string) => void
-}) {
-  return <div>User Card</div>
-}
-```
-
-### 4. Spacing Guidelines
-
-**Prefer padding over margins for spacing.**
-
-**Reason:** We want to consolidate usage to paddings instead of margins.
-
-```typescript
-// ✅ Good
-<div className="p-4 space-y-2">
-  <div className="p-2">Content</div>
-</div>
-
-// ❌ Bad
-<div className="m-4 space-y-2">
-  <div className="m-2">Content</div>
-</div>
-```
-
-### 5. Tailwind Dark Mode
-
-**Strictly forbid using the `dark:` modifier in Tailwind classes, except for logo icon handling.**
-
-**Reason:** The `colors.css` file already, VERY CAREFULLY, defines what the exact opposite colour of each light-mode colour is. Overriding this behaviour is VERY bad and will lead to horrible UI breakages.
-
-**Exception:** The `createLogoIcon` helper in `web/src/components/icons/icons.tsx` uses `dark:` modifiers (`dark:invert`, `dark:hidden`, `dark:block`) to handle third-party logo icons that cannot automatically adapt through `colors.css`. This is the ONLY acceptable use of dark mode modifiers.
-
-```typescript
-// ✅ Good - Standard components use `tailwind-themes/tailwind.config.js` / `src/app/css/colors.css`
-<div className="bg-background-neutral-03 text-text-02">
-  Content
-</div>
-
-// ✅ Good - Logo icons with dark mode handling via createLogoIcon
-export const GithubIcon = createLogoIcon(githubLightIcon, {
-  monochromatic: true,  // Will apply dark:invert internally
-});
-
-export const GitbookIcon = createLogoIcon(gitbookLightIcon, {
-  darkSrc: gitbookDarkIcon,  // Will use dark:hidden/dark:block internally
-});
-
-// ❌ Bad - Manual dark mode overrides
-<div className="bg-white dark:bg-black text-black dark:text-white">
-  Content
-</div>
-```
-
-### 6. Class Name Utilities
-
-**Use the `cn` utility instead of raw string formatting for classNames.**
-
-**Reason:** `cn`s are easier to read. They also allow for more complex types (i.e., string-arrays) to get formatted properly (it flattens each element in that string array down). As a result, it can allow things such as conditionals (i.e., `myCondition && "some-tailwind-class"`, which evaluates to `false` when `myCondition` is `false`) to get filtered out.
-
-```typescript
-import { cn } from '@/lib/utils'
-
-// ✅ Good
-<div className={cn(
-  'base-class',
-  isActive && 'active-class',
-  className
-)}>
-  Content
-</div>
-
-// ❌ Bad
-<div className={`base-class ${isActive ? 'active-class' : ''} ${className}`}>
-  Content
-</div>
-```
-
-### 7. Custom Hooks Organization
-
-**Follow a "hook-per-file" layout. Each hook should live in its own file within `web/src/hooks`.**
-
-**Reason:** This is just a layout preference. Keeps code clean.
-
-```typescript
-// web/src/hooks/useUserData.ts
-export function useUserData(userId: string) {
-  // hook implementation
-}
-
-// web/src/hooks/useLocalStorage.ts
-export function useLocalStorage<T>(key: string, initialValue: T) {
-  // hook implementation
-}
-```
-
-### 8. Icon Usage
-
-**ONLY use icons from the `web/src/icons` directory. Do NOT use icons from `react-icons`, `lucide`, or other external libraries.**
-
-**Reason:** We have a very carefully curated selection of icons that match our Onyx guidelines. We do NOT want to muddy those up with different aesthetic stylings.
-
-```typescript
-// ✅ Good
-import SvgX from "@/icons/x";
-import SvgMoreHorizontal from "@/icons/more-horizontal";
-
-// ❌ Bad
-import { User } from "lucide-react";
-import { FiSearch } from "react-icons/fi";
-```
-
-**Missing Icons**: If an icon is needed but doesn't exist in the `web/src/icons` directory, import it from Figma using the Figma MCP tool and add it to the icons directory.
-If you need help with this step, reach out to `raunak@onyx.app`.
-
-### 9. Text Rendering
-
-**Prefer using the `refresh-components/texts/Text` component for all text rendering. Avoid "naked" text nodes.**
-
-**Reason:** The `Text` component is fully compliant with the stylings provided in Figma. It provides easy utilities to specify the text-colour and font-size in the form of flags. Super duper easy.
-
-```typescript
-// ✅ Good
-import { Text } from '@/refresh-components/texts/Text'
-
-function UserCard({ name }: { name: string }) {
-  return (
-    <Text
-      {/* The `text03` flag makes the text it renders to be coloured the 3rd-scale grey */}
-      text03
-      {/* The `mainAction` flag makes the text it renders to be "main-action" font + line-height + weightage, as described in the Figma */}
-      mainAction
-    >
-      {name}
-    </Text>
-  )
-}
-
-// ❌ Bad
-function UserCard({ name }: { name: string }) {
-  return (
-    <div>
-      <h2>{name}</h2>
-      <p>User details</p>
-    </div>
-  )
-}
-```
-
-### 10. Component Usage
-
-**Heavily avoid raw HTML input components. Always use components from the `web/src/refresh-components` or `web/lib/opal/src` directory.**
-
-**Reason:** We've put in a lot of effort to unify the components that are rendered in the Onyx app. Using raw components breaks the entire UI of the application, and leaves it in a muddier state than before.
-
-```typescript
-// ✅ Good
-import Button from '@/refresh-components/buttons/Button'
-import InputTypeIn from '@/refresh-components/inputs/InputTypeIn'
-import SvgPlusCircle from '@/icons/plus-circle'
-
-function ContactForm() {
-  return (
-    <form>
-      <InputTypeIn placeholder="Search..." />
-      <Button type="submit" leftIcon={SvgPlusCircle}>Submit</Button>
-    </form>
-  )
-}
-
-// ❌ Bad
-function ContactForm() {
-  return (
-    <form>
-      <input placeholder="Name" />
-      <textarea placeholder="Message" />
-      <button type="submit">Submit</button>
-    </form>
-  )
-}
-```
-
-### 11. Colors
-
-**Always use custom overrides for colors and borders rather than built in Tailwind CSS colors. These overrides live in `web/tailwind-themes/tailwind.config.js`.**
-
-**Reason:** Our custom color system uses CSS variables that automatically handle dark mode and maintain design consistency across the app. Standard Tailwind colors bypass this system.
-
-**Available color categories:**
-
-- **Text:** `text-01` through `text-05`, `text-inverted-XX`
-- **Backgrounds:** `background-neutral-XX`, `background-tint-XX` (and inverted variants)
-- **Borders:** `border-01` through `border-05`, `border-inverted-XX`
-- **Actions:** `action-link-XX`, `action-danger-XX`
-- **Status:** `status-info-XX`, `status-success-XX`, `status-warning-XX`, `status-error-XX`
-- **Theme:** `theme-primary-XX`, `theme-red-XX`, `theme-blue-XX`, etc.
-
-```typescript
-// ✅ Good - Use custom Onyx color classes
-<div className="bg-background-neutral-01 border border-border-02" />
-<div className="bg-background-tint-02 border border-border-01" />
-<div className="bg-status-success-01" />
-<div className="bg-action-link-01" />
-<div className="bg-theme-primary-05" />
-
-// ❌ Bad - Do NOT use standard Tailwind colors
-<div className="bg-gray-100 border border-gray-300 text-gray-600" />
-<div className="bg-white border border-slate-200" />
-<div className="bg-green-100 text-green-700" />
-<div className="bg-blue-100 text-blue-600" />
-<div className="bg-indigo-500" />
-```
-
-### 12. Data Fetching
-
-**Prefer using `useSWR` for data fetching. Data should generally be fetched on the client side. Components that need data should display a loader / placeholder while waiting for that data. Prefer loading data within the component that needs it rather than at the top level and passing it down.**
-
-**Reason:** Client side fetching allows us to load the skeleton of the page without waiting for data to load, leading to a snappier UX. Loading data where needed reduces dependencies between a component and its parent component(s).
+Frontend standards for the `web/` and `desktop/` projects live in `web/AGENTS.md`.
 
 ## Database & Migrations
 
@@ -598,7 +321,7 @@ Before writing your plan, make sure to do research. Explore the relevant section
 Never hardcode status codes or use `starlette.status` / `fastapi.status` constants directly.**
 
 A global FastAPI exception handler converts `OnyxError` into a JSON response with the standard
-`{"error_code": "...", "message": "..."}` shape. This eliminates boilerplate and keeps error
+`{"error_code": "...", "detail": "..."}` shape. This eliminates boilerplate and keeps error
 handling consistent across the entire backend.
 
 ```python

backend/Dockerfile

@@ -46,7 +46,11 @@ RUN apt-get update && \
         pkg-config \
         gcc \
         nano \
-        vim && \
+        vim \
+        # Install procps so kubernetes exec sessions can use ps aux for debugging
+        procps \
+        libjemalloc2 \
+        && \
     rm -rf /var/lib/apt/lists/* && \
     apt-get clean
 
@@ -141,6 +145,7 @@ COPY --chown=onyx:onyx ./scripts/debugging /app/scripts/debugging
 COPY --chown=onyx:onyx ./scripts/force_delete_connector_by_id.py /app/scripts/force_delete_connector_by_id.py
 COPY --chown=onyx:onyx ./scripts/supervisord_entrypoint.sh /app/scripts/supervisord_entrypoint.sh
 COPY --chown=onyx:onyx ./scripts/setup_craft_templates.sh /app/scripts/setup_craft_templates.sh
+COPY --chown=onyx:onyx ./scripts/reencrypt_secrets.py /app/scripts/reencrypt_secrets.py
 RUN chmod +x /app/scripts/supervisord_entrypoint.sh /app/scripts/setup_craft_templates.sh
 
 # Run Craft template setup at build time when ENABLE_CRAFT=true
@@ -164,6 +169,13 @@ ENV PYTHONPATH=/app
 ARG ONYX_VERSION=0.0.0-dev
 ENV ONYX_VERSION=${ONYX_VERSION}
 
+# Use jemalloc instead of glibc malloc to reduce memory fragmentation
+# in long-running Python processes (API server, Celery workers).
+# The soname is architecture-independent; the dynamic linker resolves
+# the correct path from standard library directories.
+# Placed after all RUN steps so build-time processes are unaffected.
+ENV LD_PRELOAD=libjemalloc.so.2
+
 # Default command which does nothing
 # This container is used by api server and background which specify their own CMD
 CMD ["tail", "-f", "/dev/null"]

backend/alembic/env.py

@@ -244,7 +244,10 @@ def do_run_migrations(
 
 
 def provide_iam_token_for_alembic(
-    dialect: Any, conn_rec: Any, cargs: Any, cparams: Any  # noqa: ARG001
+    dialect: Any,  # noqa: ARG001
+    conn_rec: Any,  # noqa: ARG001
+    cargs: Any,  # noqa: ARG001
+    cparams: Any,
 ) -> None:
     if USE_IAM_AUTH:
         # Database connection settings
@@ -360,8 +363,7 @@ async def run_async_migrations() -> None:
         # upgrade_all_tenants=true or schemas in multi-tenant mode
         # and for non-multi-tenant mode, we should use schemas with the default schema
         raise ValueError(
-            "No migration target specified. Use either upgrade_all_tenants=true for all tenants "
-            "or schemas for specific schemas."
+            "No migration target specified. Use either upgrade_all_tenants=true for all tenants or schemas for specific schemas."
         )
 
     await engine.dispose()
@@ -457,8 +459,7 @@ def run_migrations_offline() -> None:
     else:
         # This should not happen in the new design
         raise ValueError(
-            "No migration target specified. Use either upgrade_all_tenants=true for all tenants "
-            "or schemas for specific schemas."
+            "No migration target specified. Use either upgrade_all_tenants=true for all tenants or schemas for specific schemas."
         )
 
 

backend/alembic/run_multitenant_migrations.py

@@ -13,6 +13,7 @@ Usage examples::
     # custom settings
     python alembic/run_multitenant_migrations.py -j 8 -b 100
 """
+
 from __future__ import annotations
 
 import argparse
@@ -117,8 +118,7 @@ def run_migrations_parallel(
     batches = [schemas[i : i + batch_size] for i in range(0, len(schemas), batch_size)]
     total_batches = len(batches)
     print(
-        f"{len(schemas)} schemas in {total_batches} batch(es) "
-        f"with {max_workers} workers (batch size: {batch_size})...",
+        f"{len(schemas)} schemas in {total_batches} batch(es) with {max_workers} workers (batch size: {batch_size})...",
         flush=True,
     )
     all_success = True
@@ -166,8 +166,7 @@ def run_migrations_parallel(
                 with lock:
                     in_flight[batch_idx] = batch
                 print(
-                    f"Batch {batch_idx + 1}/{total_batches} started "
-                    f"({len(batch)} schemas): {', '.join(batch)}",
+                    f"Batch {batch_idx + 1}/{total_batches} started ({len(batch)} schemas): {', '.join(batch)}",
                     flush=True,
                 )
                 result = run_alembic_for_batch(batch)
@@ -201,7 +200,7 @@ def run_migrations_parallel(
 
                 except Exception as e:
                     print(
-                        f"Batch {batch_idx + 1}/{total_batches} " f"✗ exception: {e}",
+                        f"Batch {batch_idx + 1}/{total_batches} ✗ exception: {e}",
                         flush=True,
                     )
                     all_success = False
@@ -268,14 +267,12 @@ def main() -> int:
 
     if not schemas_to_migrate:
         print(
-            f"All {len(tenant_schemas)} tenants are already at head "
-            f"revision ({head_rev})."
+            f"All {len(tenant_schemas)} tenants are already at head revision ({head_rev})."
         )
         return 0
 
     print(
-        f"{len(schemas_to_migrate)}/{len(tenant_schemas)} tenants need "
-        f"migration (head: {head_rev})."
+        f"{len(schemas_to_migrate)}/{len(tenant_schemas)} tenants need migration (head: {head_rev})."
     )
 
     success = run_migrations_parallel(

backend/alembic/versions/25a5501dc766_group_permissions_phase1.py

@@ -0,0 +1,109 @@
+"""group_permissions_phase1
+
+Revision ID: 25a5501dc766
+Revises: b728689f45b1
+Create Date: 2026-03-23 11:41:25.557442
+
+"""
+
+from alembic import op
+import fastapi_users_db_sqlalchemy
+import sqlalchemy as sa
+
+from onyx.db.enums import AccountType
+from onyx.db.enums import GrantSource
+from onyx.db.enums import Permission
+
+
+# revision identifiers, used by Alembic.
+revision = "25a5501dc766"
+down_revision = "b728689f45b1"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # 1. Add account_type column to user table (nullable for now).
+    #    TODO(subash): backfill account_type for existing rows and add NOT NULL.
+    op.add_column(
+        "user",
+        sa.Column(
+            "account_type",
+            sa.Enum(AccountType, native_enum=False),
+            nullable=True,
+        ),
+    )
+
+    # 2. Add is_default column to user_group table
+    op.add_column(
+        "user_group",
+        sa.Column(
+            "is_default",
+            sa.Boolean(),
+            nullable=False,
+            server_default=sa.false(),
+        ),
+    )
+
+    # 3. Create permission_grant table
+    op.create_table(
+        "permission_grant",
+        sa.Column("id", sa.Integer(), autoincrement=True, nullable=False),
+        sa.Column("group_id", sa.Integer(), nullable=False),
+        sa.Column(
+            "permission",
+            sa.Enum(Permission, native_enum=False),
+            nullable=False,
+        ),
+        sa.Column(
+            "grant_source",
+            sa.Enum(GrantSource, native_enum=False),
+            nullable=False,
+        ),
+        sa.Column(
+            "granted_by",
+            fastapi_users_db_sqlalchemy.generics.GUID(),
+            nullable=True,
+        ),
+        sa.Column(
+            "granted_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.func.now(),
+            nullable=False,
+        ),
+        sa.Column(
+            "is_deleted",
+            sa.Boolean(),
+            nullable=False,
+            server_default=sa.false(),
+        ),
+        sa.PrimaryKeyConstraint("id"),
+        sa.ForeignKeyConstraint(
+            ["group_id"],
+            ["user_group.id"],
+            ondelete="CASCADE",
+        ),
+        sa.ForeignKeyConstraint(
+            ["granted_by"],
+            ["user.id"],
+            ondelete="SET NULL",
+        ),
+        sa.UniqueConstraint(
+            "group_id", "permission", name="uq_permission_grant_group_permission"
+        ),
+    )
+
+    # 4. Index on user__user_group(user_id) — existing composite PK
+    #    has user_group_id as leading column; user-filtered queries need this
+    op.create_index(
+        "ix_user__user_group_user_id",
+        "user__user_group",
+        ["user_id"],
+    )
+
+
+def downgrade() -> None:
+    op.drop_index("ix_user__user_group_user_id", table_name="user__user_group")
+    op.drop_table("permission_grant")
+    op.drop_column("user_group", "is_default")
+    op.drop_column("user", "account_type")

backend/alembic/versions/27fb147a843f_add_timestamps_to_user_table.py

@@ -0,0 +1,43 @@
+"""add timestamps to user table
+
+Revision ID: 27fb147a843f
+Revises: b5c4d7e8f9a1
+Create Date: 2026-03-08 17:18:40.828644
+
+"""
+
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision = "27fb147a843f"
+down_revision = "b5c4d7e8f9a1"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "user",
+        sa.Column(
+            "created_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.func.now(),
+            nullable=False,
+        ),
+    )
+    op.add_column(
+        "user",
+        sa.Column(
+            "updated_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.func.now(),
+            nullable=False,
+        ),
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("user", "updated_at")
+    op.drop_column("user", "created_at")

backend/alembic/versions/2b75d0a8ffcb_user_file_schema_cleanup.py

@@ -50,8 +50,7 @@ def upgrade() -> None:
 
         if orphaned_count > 0:
             logger.warning(
-                f"WARNING: {orphaned_count} chat_session records still have "
-                f"folder_id without project_id. Proceeding anyway."
+                f"WARNING: {orphaned_count} chat_session records still have folder_id without project_id. Proceeding anyway."
             )
 
     # === Step 2: Drop chat_session.folder_id ===

backend/alembic/versions/3a78dba1080a_user_file_legacy_data_cleanup.py

@@ -75,8 +75,7 @@ def batch_delete(
 
     if failed_batches:
         logger.warning(
-            f"Failed to delete {len(failed_batches)} batches from {table_name}. "
-            f"Total deleted: {total_deleted}/{total_count}"
+            f"Failed to delete {len(failed_batches)} batches from {table_name}. Total deleted: {total_deleted}/{total_count}"
         )
         # Fail the migration to avoid silently succeeding on partial cleanup
         raise RuntimeError(

backend/alembic/versions/40926a4dab77_reset_userfile_document_id_migrated_.py

@@ -18,8 +18,7 @@ depends_on = None
 def upgrade() -> None:
     # Set all existing records to not migrated
     op.execute(
-        "UPDATE user_file SET document_id_migrated = FALSE "
-        "WHERE document_id_migrated IS DISTINCT FROM FALSE;"
+        "UPDATE user_file SET document_id_migrated = FALSE WHERE document_id_migrated IS DISTINCT FROM FALSE;"
     )
 
 

backend/alembic/versions/495cb26ce93e_create_knowlege_graph_tables.py

@@ -35,7 +35,6 @@ def upgrade() -> None:
     # environment variables MUST be set. Otherwise, an exception will be raised.
 
     if not MULTI_TENANT:
-
         # Enable pg_trgm extension if not already enabled
         op.execute("CREATE EXTENSION IF NOT EXISTS pg_trgm")
 
@@ -481,8 +480,7 @@ def upgrade() -> None:
         f"ON kg_entity USING GIN (name {POSTGRES_DEFAULT_SCHEMA}.gin_trgm_ops)"
     )
     op.execute(
-        "CREATE INDEX IF NOT EXISTS idx_kg_entity_normalization_trigrams "
-        "ON kg_entity USING GIN (name_trigrams)"
+        "CREATE INDEX IF NOT EXISTS idx_kg_entity_normalization_trigrams ON kg_entity USING GIN (name_trigrams)"
     )
 
     # Create kg_entity trigger to update kg_entity.name and its trigrams

backend/alembic/versions/4d58345da04a_lowercase_user_emails.py

@@ -51,10 +51,7 @@ def upgrade() -> None:
                 next_email = f"{username.lower()}_{attempt}@{domain.lower()}"
                 # Email conflict occurred, append `_1`, `_2`, etc., to the username
                 logger.warning(
-                    f"Conflict while lowercasing email: "
-                    f"old_email={email} "
-                    f"conflicting_email={new_email} "
-                    f"next_email={next_email}"
+                    f"Conflict while lowercasing email: old_email={email} conflicting_email={new_email} next_email={next_email}"
                 )
                 new_email = next_email
                 attempt += 1

backend/alembic/versions/689433b0d8de_add_hook_and_hook_execution_log_tables.py

@@ -0,0 +1,103 @@
+"""add_hook_and_hook_execution_log_tables
+
+Revision ID: 689433b0d8de
+Revises: 93a2e195e25c
+Create Date: 2026-03-13 11:25:06.547474
+
+"""
+
+from alembic import op
+import sqlalchemy as sa
+from sqlalchemy.dialects.postgresql import UUID as PGUUID
+
+
+# revision identifiers, used by Alembic.
+revision = "689433b0d8de"
+down_revision = "93a2e195e25c"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.create_table(
+        "hook",
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column("name", sa.String(), nullable=False),
+        sa.Column(
+            "hook_point",
+            sa.Enum("document_ingestion", "query_processing", native_enum=False),
+            nullable=False,
+        ),
+        sa.Column("endpoint_url", sa.Text(), nullable=True),
+        sa.Column("api_key", sa.LargeBinary(), nullable=True),
+        sa.Column("is_reachable", sa.Boolean(), nullable=True),
+        sa.Column(
+            "fail_strategy",
+            sa.Enum("hard", "soft", native_enum=False),
+            nullable=False,
+        ),
+        sa.Column("timeout_seconds", sa.Float(), nullable=False),
+        sa.Column(
+            "is_active", sa.Boolean(), nullable=False, server_default=sa.text("false")
+        ),
+        sa.Column(
+            "deleted", sa.Boolean(), nullable=False, server_default=sa.text("false")
+        ),
+        sa.Column("creator_id", PGUUID(as_uuid=True), nullable=True),
+        sa.Column(
+            "created_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.text("now()"),
+            nullable=False,
+        ),
+        sa.Column(
+            "updated_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.text("now()"),
+            nullable=False,
+        ),
+        sa.ForeignKeyConstraint(["creator_id"], ["user.id"], ondelete="SET NULL"),
+        sa.PrimaryKeyConstraint("id"),
+    )
+    op.create_index(
+        "ix_hook_one_non_deleted_per_point",
+        "hook",
+        ["hook_point"],
+        unique=True,
+        postgresql_where=sa.text("deleted = false"),
+    )
+
+    op.create_table(
+        "hook_execution_log",
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column("hook_id", sa.Integer(), nullable=False),
+        sa.Column(
+            "is_success",
+            sa.Boolean(),
+            nullable=False,
+        ),
+        sa.Column("error_message", sa.Text(), nullable=True),
+        sa.Column("status_code", sa.Integer(), nullable=True),
+        sa.Column("duration_ms", sa.Integer(), nullable=True),
+        sa.Column(
+            "created_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.text("now()"),
+            nullable=False,
+        ),
+        sa.ForeignKeyConstraint(["hook_id"], ["hook.id"], ondelete="CASCADE"),
+        sa.PrimaryKeyConstraint("id"),
+    )
+    op.create_index("ix_hook_execution_log_hook_id", "hook_execution_log", ["hook_id"])
+    op.create_index(
+        "ix_hook_execution_log_created_at", "hook_execution_log", ["created_at"]
+    )
+
+
+def downgrade() -> None:
+    op.drop_index("ix_hook_execution_log_created_at", table_name="hook_execution_log")
+    op.drop_index("ix_hook_execution_log_hook_id", table_name="hook_execution_log")
+    op.drop_table("hook_execution_log")
+
+    op.drop_index("ix_hook_one_non_deleted_per_point", table_name="hook")
+    op.drop_table("hook")

backend/alembic/versions/72aa7de2e5cf_make_processing_mode_default_all_caps.py

@@ -24,12 +24,10 @@ depends_on = None
 def upgrade() -> None:
     # Convert existing lowercase values to uppercase to match enum member names
     op.execute(
-        "UPDATE connector_credential_pair SET processing_mode = 'REGULAR' "
-        "WHERE processing_mode = 'regular'"
+        "UPDATE connector_credential_pair SET processing_mode = 'REGULAR' WHERE processing_mode = 'regular'"
     )
     op.execute(
-        "UPDATE connector_credential_pair SET processing_mode = 'FILE_SYSTEM' "
-        "WHERE processing_mode = 'file_system'"
+        "UPDATE connector_credential_pair SET processing_mode = 'FILE_SYSTEM' WHERE processing_mode = 'file_system'"
     )
 
     # Update the server default to use uppercase

backend/alembic/versions/7b9b952abdf6_update_entities.py

@@ -289,8 +289,7 @@ def upgrade() -> None:
         attributes_str = json.dumps(attributes).replace("'", "''")
         op.execute(
             sa.text(
-                f"UPDATE kg_entity_type SET attributes = '{attributes_str}'"
-                f"WHERE id_name = '{entity_type}'"
+                f"UPDATE kg_entity_type SET attributes = '{attributes_str}'WHERE id_name = '{entity_type}'"
             ),
         )
 
@@ -312,7 +311,6 @@ def downgrade() -> None:
         attributes_str = json.dumps(attributes).replace("'", "''")
         op.execute(
             sa.text(
-                f"UPDATE kg_entity_type SET attributes = '{attributes_str}'"
-                f"WHERE id_name = '{entity_type}'"
+                f"UPDATE kg_entity_type SET attributes = '{attributes_str}'WHERE id_name = '{entity_type}'"
             ),
         )

backend/alembic/versions/90e3b9af7da4_tag_fix.py

@@ -160,7 +160,7 @@ def remove_old_tags() -> None:
                     f"""
                     DELETE FROM document__tag
                     WHERE document_id = '{document_id}'
-                    AND tag_id IN ({','.join(to_delete)})
+                    AND tag_id IN ({",".join(to_delete)})
                     """
                 )
             )
@@ -239,7 +239,7 @@ def _get_batch_documents_with_multiple_tags(
         ).fetchall()
         if not batch:
             break
-        doc_ids = [document_id for document_id, in batch]
+        doc_ids = [document_id for (document_id,) in batch]
         yield doc_ids
         offset_clause = f"AND document__tag.document_id > '{doc_ids[-1]}'"
 

backend/alembic/versions/93a2e195e25c_add_voice_provider_and_user_voice_prefs.py

@@ -0,0 +1,117 @@
+"""add_voice_provider_and_user_voice_prefs
+
+Revision ID: 93a2e195e25c
+Revises: 27fb147a843f
+Create Date: 2026-02-23 15:16:39.507304
+
+"""
+
+from alembic import op
+import sqlalchemy as sa
+from sqlalchemy import column
+from sqlalchemy import true
+from sqlalchemy.dialects import postgresql
+
+
+# revision identifiers, used by Alembic.
+revision = "93a2e195e25c"
+down_revision = "27fb147a843f"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # Create voice_provider table
+    op.create_table(
+        "voice_provider",
+        sa.Column("id", sa.Integer(), primary_key=True),
+        sa.Column("name", sa.String(), unique=True, nullable=False),
+        sa.Column("provider_type", sa.String(), nullable=False),
+        sa.Column("api_key", sa.LargeBinary(), nullable=True),
+        sa.Column("api_base", sa.String(), nullable=True),
+        sa.Column("custom_config", postgresql.JSONB(), nullable=True),
+        sa.Column("stt_model", sa.String(), nullable=True),
+        sa.Column("tts_model", sa.String(), nullable=True),
+        sa.Column("default_voice", sa.String(), nullable=True),
+        sa.Column(
+            "is_default_stt", sa.Boolean(), nullable=False, server_default="false"
+        ),
+        sa.Column(
+            "is_default_tts", sa.Boolean(), nullable=False, server_default="false"
+        ),
+        sa.Column("deleted", sa.Boolean(), nullable=False, server_default="false"),
+        sa.Column(
+            "time_created",
+            sa.DateTime(timezone=True),
+            server_default=sa.func.now(),
+            nullable=False,
+        ),
+        sa.Column(
+            "time_updated",
+            sa.DateTime(timezone=True),
+            server_default=sa.func.now(),
+            onupdate=sa.func.now(),
+            nullable=False,
+        ),
+    )
+
+    # Add partial unique indexes to enforce only one default STT/TTS provider
+    op.create_index(
+        "ix_voice_provider_one_default_stt",
+        "voice_provider",
+        ["is_default_stt"],
+        unique=True,
+        postgresql_where=column("is_default_stt") == true(),
+    )
+    op.create_index(
+        "ix_voice_provider_one_default_tts",
+        "voice_provider",
+        ["is_default_tts"],
+        unique=True,
+        postgresql_where=column("is_default_tts") == true(),
+    )
+
+    # Add voice preference columns to user table
+    op.add_column(
+        "user",
+        sa.Column(
+            "voice_auto_send",
+            sa.Boolean(),
+            default=False,
+            nullable=False,
+            server_default="false",
+        ),
+    )
+    op.add_column(
+        "user",
+        sa.Column(
+            "voice_auto_playback",
+            sa.Boolean(),
+            default=False,
+            nullable=False,
+            server_default="false",
+        ),
+    )
+    op.add_column(
+        "user",
+        sa.Column(
+            "voice_playback_speed",
+            sa.Float(),
+            default=1.0,
+            nullable=False,
+            server_default="1.0",
+        ),
+    )
+
+
+def downgrade() -> None:
+    # Remove user voice preference columns
+    op.drop_column("user", "voice_playback_speed")
+    op.drop_column("user", "voice_auto_playback")
+    op.drop_column("user", "voice_auto_send")
+
+    op.drop_index("ix_voice_provider_one_default_tts", table_name="voice_provider")
+    op.drop_index("ix_voice_provider_one_default_stt", table_name="voice_provider")
+
+    # Drop voice_provider table
+    op.drop_table("voice_provider")

backend/alembic/versions/a01bf2971c5d_update_default_tool_descriptions.py

@@ -24,8 +24,7 @@ TOOL_DESCRIPTIONS = {
         "The action will be used when the user asks the agent to generate an image."
     ),
     "WebSearchTool": (
-        "The Web Search Action allows the agent "
-        "to perform internet searches for up-to-date information."
+        "The Web Search Action allows the agent to perform internet searches for up-to-date information."
     ),
     "KnowledgeGraphTool": (
         "The Knowledge Graph Search Action allows the agent to search the "

backend/alembic/versions/a3f8b2c1d4e5_add_preferred_response_id_to_chat_message.py

@@ -0,0 +1,36 @@
+"""add preferred_response_id and model_display_name to chat_message
+
+Revision ID: a3f8b2c1d4e5
+Create Date: 2026-03-22
+
+"""
+
+from alembic import op
+import sqlalchemy as sa
+
+# revision identifiers, used by Alembic.
+revision = "a3f8b2c1d4e5"
+down_revision = "25a5501dc766"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "chat_message",
+        sa.Column(
+            "preferred_response_id",
+            sa.Integer(),
+            sa.ForeignKey("chat_message.id", ondelete="SET NULL"),
+            nullable=True,
+        ),
+    )
+    op.add_column(
+        "chat_message",
+        sa.Column("model_display_name", sa.String(), nullable=True),
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("chat_message", "model_display_name")
+    op.drop_column("chat_message", "preferred_response_id")

backend/alembic/versions/b5c4d7e8f9a1_add_hierarchy_node_cc_pair_table.py

@@ -0,0 +1,51 @@
+"""add hierarchy_node_by_connector_credential_pair table
+
+Revision ID: b5c4d7e8f9a1
+Revises: a3b8d9e2f1c4
+Create Date: 2026-03-04
+
+"""
+
+import sqlalchemy as sa
+from alembic import op
+
+revision = "b5c4d7e8f9a1"
+down_revision = "a3b8d9e2f1c4"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.create_table(
+        "hierarchy_node_by_connector_credential_pair",
+        sa.Column("hierarchy_node_id", sa.Integer(), nullable=False),
+        sa.Column("connector_id", sa.Integer(), nullable=False),
+        sa.Column("credential_id", sa.Integer(), nullable=False),
+        sa.ForeignKeyConstraint(
+            ["hierarchy_node_id"],
+            ["hierarchy_node.id"],
+            ondelete="CASCADE",
+        ),
+        sa.ForeignKeyConstraint(
+            ["connector_id", "credential_id"],
+            [
+                "connector_credential_pair.connector_id",
+                "connector_credential_pair.credential_id",
+            ],
+            ondelete="CASCADE",
+        ),
+        sa.PrimaryKeyConstraint("hierarchy_node_id", "connector_id", "credential_id"),
+    )
+    op.create_index(
+        "ix_hierarchy_node_cc_pair_connector_credential",
+        "hierarchy_node_by_connector_credential_pair",
+        ["connector_id", "credential_id"],
+    )
+
+
+def downgrade() -> None:
+    op.drop_index(
+        "ix_hierarchy_node_cc_pair_connector_credential",
+        table_name="hierarchy_node_by_connector_credential_pair",
+    )
+    op.drop_table("hierarchy_node_by_connector_credential_pair")

backend/alembic/versions/b728689f45b1_rename_persona_is_visible_to_is_listed_.py

@@ -0,0 +1,26 @@
+"""rename persona is_visible to is_listed and featured to is_featured
+
+Revision ID: b728689f45b1
+Revises: 689433b0d8de
+Create Date: 2026-03-23 12:36:26.607305
+
+"""
+
+from alembic import op
+
+
+# revision identifiers, used by Alembic.
+revision = "b728689f45b1"
+down_revision = "689433b0d8de"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.alter_column("persona", "is_visible", new_column_name="is_listed")
+    op.alter_column("persona", "featured", new_column_name="is_featured")
+
+
+def downgrade() -> None:
+    op.alter_column("persona", "is_listed", new_column_name="is_visible")
+    op.alter_column("persona", "is_featured", new_column_name="featured")

backend/alembic/versions/c9e2cd766c29_add_s3_file_store_table.py

@@ -140,8 +140,7 @@ def _migrate_files_to_postgres() -> None:
     # Fetch rows that have external storage pointers (bucket/object_key not NULL)
     result = session.execute(
         text(
-            "SELECT file_id, bucket_name, object_key FROM file_record "
-            "WHERE bucket_name IS NOT NULL AND object_key IS NOT NULL"
+            "SELECT file_id, bucket_name, object_key FROM file_record WHERE bucket_name IS NOT NULL AND object_key IS NOT NULL"
         )
     )
 
@@ -182,8 +181,7 @@ def _migrate_files_to_postgres() -> None:
             # Update DB row: set lobj_oid, clear bucket/object_key
             session.execute(
                 text(
-                    "UPDATE file_record SET lobj_oid = :lobj_oid, bucket_name = NULL, "
-                    "object_key = NULL WHERE file_id = :file_id"
+                    "UPDATE file_record SET lobj_oid = :lobj_oid, bucket_name = NULL, object_key = NULL WHERE file_id = :file_id"
                 ),
                 {"lobj_oid": lobj_oid, "file_id": file_id},
             )
@@ -224,8 +222,7 @@ def _migrate_files_to_external_storage() -> None:
     # Find all files currently stored in PostgreSQL (lobj_oid is not null)
     result = session.execute(
         text(
-            "SELECT file_id FROM file_record WHERE lobj_oid IS NOT NULL "
-            "AND bucket_name IS NULL AND object_key IS NULL"
+            "SELECT file_id FROM file_record WHERE lobj_oid IS NOT NULL AND bucket_name IS NULL AND object_key IS NULL"
         )
     )
 

backend/alembic/versions/d09fc20a3c66_seed_builtin_tools.py

@@ -39,8 +39,7 @@ BUILT_IN_TOOLS = [
         "name": "WebSearchTool",
         "display_name": "Web Search",
         "description": (
-            "The Web Search Action allows the assistant "
-            "to perform internet searches for up-to-date information."
+            "The Web Search Action allows the assistant to perform internet searches for up-to-date information."
         ),
         "in_code_tool_id": "WebSearchTool",
     },

backend/alembic/versions/e7f8a9b0c1d2_create_anonymous_user.py

@@ -36,6 +36,56 @@ TABLES_WITH_USER_ID = [
 ]
 
 
+def _dedupe_null_notifications(connection: sa.Connection) -> None:
+    # Multiple NULL-owned notifications can exist because the unique index treats
+    # NULL user_id values as distinct. Before migrating them to the anonymous
+    # user, collapse duplicates and remove rows that would conflict with an
+    # already-existing anonymous notification.
+    result = connection.execute(
+        sa.text(
+            """
+            WITH ranked_null_notifications AS (
+                SELECT
+                    id,
+                    ROW_NUMBER() OVER (
+                        PARTITION BY notif_type, COALESCE(additional_data, '{}'::jsonb)
+                        ORDER BY first_shown DESC, last_shown DESC, id DESC
+                    ) AS row_num
+                FROM notification
+                WHERE user_id IS NULL
+            )
+            DELETE FROM notification
+            WHERE id IN (
+                SELECT id
+                FROM ranked_null_notifications
+                WHERE row_num > 1
+            )
+            """
+        )
+    )
+    if result.rowcount > 0:
+        print(f"Deleted {result.rowcount} duplicate NULL-owned notifications")
+
+    result = connection.execute(
+        sa.text(
+            """
+            DELETE FROM notification AS null_owned
+            USING notification AS anonymous_owned
+            WHERE null_owned.user_id IS NULL
+              AND anonymous_owned.user_id = :user_id
+              AND null_owned.notif_type = anonymous_owned.notif_type
+              AND COALESCE(null_owned.additional_data, '{}'::jsonb) =
+                  COALESCE(anonymous_owned.additional_data, '{}'::jsonb)
+            """
+        ),
+        {"user_id": ANONYMOUS_USER_UUID},
+    )
+    if result.rowcount > 0:
+        print(
+            f"Deleted {result.rowcount} NULL-owned notifications that conflict with existing anonymous-owned notifications"
+        )
+
+
 def upgrade() -> None:
     """
     Create the anonymous user for anonymous access feature.
@@ -65,7 +115,12 @@ def upgrade() -> None:
 
     # Migrate any remaining user_id=NULL records to anonymous user
     for table in TABLES_WITH_USER_ID:
-        try:
+        # Dedup notifications outside the savepoint so deletions persist
+        # even if the subsequent UPDATE rolls back
+        if table == "notification":
+            _dedupe_null_notifications(connection)
+
+        with connection.begin_nested():
             # Exclude public credential (id=0) which must remain user_id=NULL
             # Exclude builtin tools (in_code_tool_id IS NOT NULL) which must remain user_id=NULL
             # Exclude builtin personas (builtin_persona=True) which must remain user_id=NULL
@@ -80,6 +135,7 @@ def upgrade() -> None:
                 condition = "user_id IS NULL AND is_public = false"
             else:
                 condition = "user_id IS NULL"
+
             result = connection.execute(
                 sa.text(
                     f"""
@@ -92,19 +148,19 @@ def upgrade() -> None:
             )
             if result.rowcount > 0:
                 print(f"Updated {result.rowcount} rows in {table} to anonymous user")
-        except Exception as e:
-            print(f"Skipping {table}: {e}")
 
 
 def downgrade() -> None:
     """
     Set anonymous user's records back to NULL and delete the anonymous user.
+
+    Note: Duplicate NULL-owned notifications removed during upgrade are not restored.
     """
     connection = op.get_bind()
 
     # Set records back to NULL
     for table in TABLES_WITH_USER_ID:
-        try:
+        with connection.begin_nested():
             connection.execute(
                 sa.text(
                     f"""
@@ -115,8 +171,6 @@ def downgrade() -> None:
                 ),
                 {"user_id": ANONYMOUS_USER_UUID},
             )
-        except Exception:
-            pass
 
     # Delete the anonymous user
     connection.execute(

backend/alembic_tenants/versions/3b9f09038764_add_read_only_kg_user.py

@@ -11,7 +11,6 @@ from sqlalchemy import text
 from alembic import op
 from onyx.configs.app_configs import DB_READONLY_PASSWORD
 from onyx.configs.app_configs import DB_READONLY_USER
-from shared_configs.configs import MULTI_TENANT
 
 
 # revision identifiers, used by Alembic.
@@ -22,59 +21,52 @@ depends_on = None
 
 
 def upgrade() -> None:
-    if MULTI_TENANT:
+    # Enable pg_trgm extension if not already enabled
+    op.execute("CREATE EXTENSION IF NOT EXISTS pg_trgm")
 
-        # Enable pg_trgm extension if not already enabled
-        op.execute("CREATE EXTENSION IF NOT EXISTS pg_trgm")
+    # Create the read-only db user if it does not already exist.
+    if not (DB_READONLY_USER and DB_READONLY_PASSWORD):
+        raise Exception("DB_READONLY_USER or DB_READONLY_PASSWORD is not set")
 
-        # Create read-only db user here only in multi-tenant mode. For single-tenant mode,
-        # the user is created in the standard migration.
-        if not (DB_READONLY_USER and DB_READONLY_PASSWORD):
-            raise Exception("DB_READONLY_USER or DB_READONLY_PASSWORD is not set")
-
-        op.execute(
-            text(
-                f"""
-                DO $$
-                BEGIN
-                    -- Check if the read-only user already exists
-                    IF NOT EXISTS (SELECT FROM pg_catalog.pg_roles WHERE rolname = '{DB_READONLY_USER}') THEN
-                        -- Create the read-only user with the specified password
-                        EXECUTE format('CREATE USER %I WITH PASSWORD %L', '{DB_READONLY_USER}', '{DB_READONLY_PASSWORD}');
-                        -- First revoke all privileges to ensure a clean slate
-                        EXECUTE format('REVOKE ALL ON DATABASE %I FROM %I', current_database(), '{DB_READONLY_USER}');
-                        -- Grant only the CONNECT privilege to allow the user to connect to the database
-                        -- but not perform any operations without additional specific grants
-                        EXECUTE format('GRANT CONNECT ON DATABASE %I TO %I', current_database(), '{DB_READONLY_USER}');
-                    END IF;
-                END
-                $$;
-                """
-            )
-        )
-
-
-def downgrade() -> None:
-    if MULTI_TENANT:
-        # Drop read-only db user here only in single tenant mode. For multi-tenant mode,
-        # the user is dropped in the alembic_tenants migration.
-
-        op.execute(
-            text(
-                f"""
+    op.execute(
+        text(
+            f"""
             DO $$
             BEGIN
-                IF EXISTS (SELECT FROM pg_catalog.pg_roles WHERE rolname = '{DB_READONLY_USER}') THEN
-                    -- First revoke all privileges from the database
+                -- Check if the read-only user already exists
+                IF NOT EXISTS (SELECT FROM pg_catalog.pg_roles WHERE rolname = '{DB_READONLY_USER}') THEN
+                    -- Create the read-only user with the specified password
+                    EXECUTE format('CREATE USER %I WITH PASSWORD %L', '{DB_READONLY_USER}', '{DB_READONLY_PASSWORD}');
+                    -- First revoke all privileges to ensure a clean slate
                     EXECUTE format('REVOKE ALL ON DATABASE %I FROM %I', current_database(), '{DB_READONLY_USER}');
-                    -- Then revoke all privileges from the public schema
-                    EXECUTE format('REVOKE ALL ON SCHEMA public FROM %I', '{DB_READONLY_USER}');
-                    -- Then drop the user
-                    EXECUTE format('DROP USER %I', '{DB_READONLY_USER}');
+                    -- Grant only the CONNECT privilege to allow the user to connect to the database
+                    -- but not perform any operations without additional specific grants
+                    EXECUTE format('GRANT CONNECT ON DATABASE %I TO %I', current_database(), '{DB_READONLY_USER}');
                 END IF;
             END
             $$;
-        """
-            )
+            """
         )
-        op.execute(text("DROP EXTENSION IF EXISTS pg_trgm"))
+    )
+
+
+def downgrade() -> None:
+    op.execute(
+        text(
+            f"""
+        DO $$
+        BEGIN
+            IF EXISTS (SELECT FROM pg_catalog.pg_roles WHERE rolname = '{DB_READONLY_USER}') THEN
+                -- First revoke all privileges from the database
+                EXECUTE format('REVOKE ALL ON DATABASE %I FROM %I', current_database(), '{DB_READONLY_USER}');
+                -- Then revoke all privileges from the public schema
+                EXECUTE format('REVOKE ALL ON SCHEMA public FROM %I', '{DB_READONLY_USER}');
+                -- Then drop the user
+                EXECUTE format('DROP USER %I', '{DB_READONLY_USER}');
+            END IF;
+        END
+        $$;
+    """
+        )
+    )
+    op.execute(text("DROP EXTENSION IF EXISTS pg_trgm"))

backend/ee/onyx/access/access.py

@@ -9,12 +9,15 @@ from onyx.access.access import (
     _get_access_for_documents as get_access_for_documents_without_groups,
 )
 from onyx.access.access import _get_acl_for_user as get_acl_for_user_without_groups
+from onyx.access.access import collect_user_file_access
 from onyx.access.models import DocumentAccess
 from onyx.access.utils import prefix_external_group
 from onyx.access.utils import prefix_user_group
 from onyx.db.document import get_document_sources
 from onyx.db.document import get_documents_by_ids
 from onyx.db.models import User
+from onyx.db.models import UserFile
+from onyx.db.user_file import fetch_user_files_with_access_relationships
 from onyx.utils.logger import setup_logger
 
 
@@ -116,6 +119,68 @@ def _get_access_for_documents(
     return access_map
 
 
+def _collect_user_file_group_names(user_file: UserFile) -> set[str]:
+    """Extract user-group names from the already-loaded Persona.groups
+    relationships on a UserFile (skipping deleted personas)."""
+    groups: set[str] = set()
+    for persona in user_file.assistants:
+        if persona.deleted:
+            continue
+        for group in persona.groups:
+            groups.add(group.name)
+    return groups
+
+
+def get_access_for_user_files_impl(
+    user_file_ids: list[str],
+    db_session: Session,
+) -> dict[str, DocumentAccess]:
+    """EE version: extends the MIT user file ACL with user group names
+    from personas shared via user groups.
+
+    Uses a single DB query (via fetch_user_files_with_access_relationships)
+    that eagerly loads both the MIT-needed and EE-needed relationships.
+
+    NOTE: is imported in onyx.access.access by `fetch_versioned_implementation`
+    DO NOT REMOVE."""
+    user_files = fetch_user_files_with_access_relationships(
+        user_file_ids, db_session, eager_load_groups=True
+    )
+    return build_access_for_user_files_impl(user_files)
+
+
+def build_access_for_user_files_impl(
+    user_files: list[UserFile],
+) -> dict[str, DocumentAccess]:
+    """EE version: works on pre-loaded UserFile objects.
+    Expects Persona.groups to be eagerly loaded.
+
+    NOTE: is imported in onyx.access.access by `fetch_versioned_implementation`
+    DO NOT REMOVE."""
+    result: dict[str, DocumentAccess] = {}
+    for user_file in user_files:
+        if user_file.user is None:
+            result[str(user_file.id)] = DocumentAccess.build(
+                user_emails=[],
+                user_groups=[],
+                is_public=True,
+                external_user_emails=[],
+                external_user_group_ids=[],
+            )
+            continue
+
+        emails, is_public = collect_user_file_access(user_file)
+        group_names = _collect_user_file_group_names(user_file)
+        result[str(user_file.id)] = DocumentAccess.build(
+            user_emails=list(emails),
+            user_groups=list(group_names),
+            is_public=is_public,
+            external_user_emails=[],
+            external_user_group_ids=[],
+        )
+    return result
+
+
 def _get_acl_for_user(user: User, db_session: Session) -> set[str]:
     """Returns a list of ACL entries that the user has access to. This is meant to be
     used downstream to filter out documents that the user does not have access to. The

backend/ee/onyx/auth/users.py

@@ -1,3 +1,4 @@
+import os
 from datetime import datetime
 
 import jwt
@@ -20,7 +21,12 @@ logger = setup_logger()
 
 
 def verify_auth_setting() -> None:
-    # All the Auth flows are valid for EE version
+    # All the Auth flows are valid for EE version, but warn about deprecated 'disabled'
+    raw_auth_type = (os.environ.get("AUTH_TYPE") or "").lower()
+    if raw_auth_type == "disabled":
+        logger.warning(
+            "AUTH_TYPE='disabled' is no longer supported. Using 'basic' instead. Please update your configuration."
+        )
     logger.notice(f"Using Auth Type: {AUTH_TYPE.value}")
 
 

backend/ee/onyx/background/celery/tasks/cloud/tasks.py

@@ -59,7 +59,6 @@ def cloud_beat_task_generator(
         # gated_tenants = get_gated_tenants()
 
         for tenant_id in tenant_ids:
-
             # Same comment here as the above NOTE
             # if tenant_id in gated_tenants:
             #     continue

backend/ee/onyx/background/celery/tasks/doc_permission_syncing/tasks.py

@@ -424,10 +424,7 @@ def connector_permission_sync_generator_task(
             raise ValueError(error_msg)
 
         if not redis_connector.permissions.fenced:  # The fence must exist
-            error_msg = (
-                f"connector_permission_sync_generator_task - fence not found: "
-                f"fence={redis_connector.permissions.fence_key}"
-            )
+            error_msg = f"connector_permission_sync_generator_task - fence not found: fence={redis_connector.permissions.fence_key}"
             _fail_doc_permission_sync_attempt(attempt_id, error_msg)
             raise ValueError(error_msg)
 
@@ -441,8 +438,7 @@ def connector_permission_sync_generator_task(
 
         if payload.celery_task_id is None:
             logger.info(
-                f"connector_permission_sync_generator_task - Waiting for fence: "
-                f"fence={redis_connector.permissions.fence_key}"
+                f"connector_permission_sync_generator_task - Waiting for fence: fence={redis_connector.permissions.fence_key}"
             )
             sleep(1)
             continue
@@ -608,8 +604,7 @@ def connector_permission_sync_generator_task(
                 docs_with_permission_errors=docs_with_errors,
             )
             task_logger.info(
-                f"Completed doc permission sync attempt {attempt_id}: "
-                f"{tasks_generated} docs, {docs_with_errors} errors"
+                f"Completed doc permission sync attempt {attempt_id}: {tasks_generated} docs, {docs_with_errors} errors"
             )
 
             redis_connector.permissions.generator_complete = tasks_generated
@@ -716,9 +711,7 @@ def element_update_permissions(
 
             elapsed = time.monotonic() - start
             task_logger.info(
-                f"{element_type}={element_id} "
-                f"action=update_permissions "
-                f"elapsed={elapsed:.2f}"
+                f"{element_type}={element_id} action=update_permissions elapsed={elapsed:.2f}"
             )
     except Exception as e:
         task_logger.exception(
@@ -900,8 +893,7 @@ def validate_permission_sync_fence(
         tasks_not_in_celery += 1
 
     task_logger.info(
-        "validate_permission_sync_fence task check: "
-        f"tasks_scanned={tasks_scanned} tasks_not_in_celery={tasks_not_in_celery}"
+        f"validate_permission_sync_fence task check: tasks_scanned={tasks_scanned} tasks_not_in_celery={tasks_not_in_celery}"
     )
 
     # we're active if there are still tasks to run and those tasks all exist in celery
@@ -1007,7 +999,10 @@ class PermissionSyncCallback(IndexingHeartbeatInterface):
 
 
 def monitor_ccpair_permissions_taskset(
-    tenant_id: str, key_bytes: bytes, r: Redis, db_session: Session  # noqa: ARG001
+    tenant_id: str,
+    key_bytes: bytes,
+    r: Redis,  # noqa: ARG001
+    db_session: Session,
 ) -> None:
     fence_key = key_bytes.decode("utf-8")
     cc_pair_id_str = RedisConnector.get_id_from_fence_key(fence_key)
@@ -1031,8 +1026,7 @@ def monitor_ccpair_permissions_taskset(
         payload = redis_connector.permissions.payload
     except ValidationError:
         task_logger.exception(
-            "Permissions sync payload failed to validate. "
-            "Schema may have been updated."
+            "Permissions sync payload failed to validate. Schema may have been updated."
         )
         return
 
@@ -1041,11 +1035,7 @@ def monitor_ccpair_permissions_taskset(
 
     remaining = redis_connector.permissions.get_remaining()
     task_logger.info(
-        f"Permissions sync progress: "
-        f"cc_pair={cc_pair_id} "
-        f"id={payload.id} "
-        f"remaining={remaining} "
-        f"initial={initial}"
+        f"Permissions sync progress: cc_pair={cc_pair_id} id={payload.id} remaining={remaining} initial={initial}"
     )
 
     # Add telemetry for permission syncing progress
@@ -1064,10 +1054,7 @@ def monitor_ccpair_permissions_taskset(
 
     mark_cc_pair_as_permissions_synced(db_session, int(cc_pair_id), payload.started)
     task_logger.info(
-        f"Permissions sync finished: "
-        f"cc_pair={cc_pair_id} "
-        f"id={payload.id} "
-        f"num_synced={initial}"
+        f"Permissions sync finished: cc_pair={cc_pair_id} id={payload.id} num_synced={initial}"
     )
 
     # Add telemetry for permission syncing complete

backend/ee/onyx/background/celery/tasks/external_group_syncing/tasks.py

@@ -111,23 +111,20 @@ def _is_external_group_sync_due(cc_pair: ConnectorCredentialPair) -> bool:
 
     if cc_pair.access_type != AccessType.SYNC:
         task_logger.error(
-            f"Received non-sync CC Pair {cc_pair.id} for external "
-            f"group sync. Actual access type: {cc_pair.access_type}"
+            f"Received non-sync CC Pair {cc_pair.id} for external group sync. Actual access type: {cc_pair.access_type}"
         )
         return False
 
     if cc_pair.status == ConnectorCredentialPairStatus.DELETING:
         task_logger.debug(
-            f"Skipping group sync for CC Pair {cc_pair.id} - "
-            f"CC Pair is being deleted"
+            f"Skipping group sync for CC Pair {cc_pair.id} - CC Pair is being deleted"
         )
         return False
 
     sync_config = get_source_perm_sync_config(cc_pair.connector.source)
     if sync_config is None:
         task_logger.debug(
-            f"Skipping group sync for CC Pair {cc_pair.id} - "
-            f"no sync config found for {cc_pair.connector.source}"
+            f"Skipping group sync for CC Pair {cc_pair.id} - no sync config found for {cc_pair.connector.source}"
         )
         return False
 
@@ -135,8 +132,7 @@ def _is_external_group_sync_due(cc_pair: ConnectorCredentialPair) -> bool:
     # This is fine because all sources dont necessarily have a concept of groups
     if sync_config.group_sync_config is None:
         task_logger.debug(
-            f"Skipping group sync for CC Pair {cc_pair.id} - "
-            f"no group sync config found for {cc_pair.connector.source}"
+            f"Skipping group sync for CC Pair {cc_pair.id} - no group sync config found for {cc_pair.connector.source}"
         )
         return False
 

backend/ee/onyx/background/celery/tasks/tenant_provisioning/tasks.py

@@ -25,13 +25,13 @@ from onyx.redis.redis_pool import get_redis_client
 from shared_configs.configs import MULTI_TENANT
 from shared_configs.configs import TENANT_ID_PREFIX
 
-# Default number of pre-provisioned tenants to maintain
-DEFAULT_TARGET_AVAILABLE_TENANTS = 5
+# Maximum tenants to provision in a single task run.
+# Each tenant takes ~80s (alembic migrations), so 5 tenants ≈ 7 minutes.
+_MAX_TENANTS_PER_RUN = 5
 
-# Soft time limit for tenant pre-provisioning tasks (in seconds)
-_TENANT_PROVISIONING_SOFT_TIME_LIMIT = 60 * 5  # 5 minutes
-# Hard time limit for tenant pre-provisioning tasks (in seconds)
-_TENANT_PROVISIONING_TIME_LIMIT = 60 * 10  # 10 minutes
+# Time limits sized for worst-case batch: _MAX_TENANTS_PER_RUN × ~90s + buffer.
+_TENANT_PROVISIONING_SOFT_TIME_LIMIT = 60 * 10  # 10 minutes
+_TENANT_PROVISIONING_TIME_LIMIT = 60 * 15  # 15 minutes
 
 
 @shared_task(
@@ -58,7 +58,7 @@ def check_available_tenants(self: Task) -> None:  # noqa: ARG001
     r = get_redis_client(tenant_id=ONYX_CLOUD_TENANT_ID)
     lock_check: RedisLock = r.lock(
         OnyxRedisLocks.CHECK_AVAILABLE_TENANTS_LOCK,
-        timeout=_TENANT_PROVISIONING_SOFT_TIME_LIMIT,
+        timeout=_TENANT_PROVISIONING_TIME_LIMIT,
     )
 
     # These tasks should never overlap
@@ -74,9 +74,7 @@ def check_available_tenants(self: Task) -> None:  # noqa: ARG001
             num_available_tenants = db_session.query(AvailableTenant).count()
 
         # Get the target number of available tenants
-        num_minimum_available_tenants = getattr(
-            TARGET_AVAILABLE_TENANTS, "value", DEFAULT_TARGET_AVAILABLE_TENANTS
-        )
+        num_minimum_available_tenants = TARGET_AVAILABLE_TENANTS
 
         # Calculate how many new tenants we need to provision
         if num_available_tenants < num_minimum_available_tenants:
@@ -90,22 +88,46 @@ def check_available_tenants(self: Task) -> None:  # noqa: ARG001
             f"To provision: {tenants_to_provision}"
         )
 
-        # just provision one tenant each time we run this ... increase if needed.
-        if tenants_to_provision > 0:
-            pre_provision_tenant()
+        batch_size = min(tenants_to_provision, _MAX_TENANTS_PER_RUN)
+        if batch_size < tenants_to_provision:
+            task_logger.info(
+                f"Capping batch to {batch_size} "
+                f"(need {tenants_to_provision}, will catch up next cycle)"
+            )
+
+        provisioned = 0
+        for i in range(batch_size):
+            task_logger.info(f"Provisioning tenant {i + 1}/{batch_size}")
+            try:
+                if pre_provision_tenant():
+                    provisioned += 1
+            except Exception:
+                task_logger.exception(
+                    f"Failed to provision tenant {i + 1}/{batch_size}, "
+                    "continuing with remaining tenants"
+                )
+
+        task_logger.info(f"Provisioning complete: {provisioned}/{batch_size} succeeded")
 
     except Exception:
         task_logger.exception("Error in check_available_tenants task")
 
     finally:
-        lock_check.release()
+        try:
+            lock_check.release()
+        except Exception:
+            task_logger.warning(
+                "Could not release check lock (likely expired), continuing"
+            )
 
 
-def pre_provision_tenant() -> None:
+def pre_provision_tenant() -> bool:
     """
     Pre-provision a new tenant and store it in the NewAvailableTenant table.
     This function fully sets up the tenant with all necessary configurations,
     so it's ready to be assigned to a user immediately.
+
+    Returns True if a tenant was successfully provisioned, False otherwise.
     """
     # The MULTI_TENANT check is now done at the caller level (check_available_tenants)
     # rather than inside this function
@@ -113,15 +135,15 @@ def pre_provision_tenant() -> None:
     r = get_redis_client(tenant_id=ONYX_CLOUD_TENANT_ID)
     lock_provision: RedisLock = r.lock(
         OnyxRedisLocks.CLOUD_PRE_PROVISION_TENANT_LOCK,
-        timeout=_TENANT_PROVISIONING_SOFT_TIME_LIMIT,
+        timeout=_TENANT_PROVISIONING_TIME_LIMIT,
     )
 
     # Allow multiple pre-provisioning tasks to run, but ensure they don't overlap
     if not lock_provision.acquire(blocking=False):
-        task_logger.debug(
-            "Skipping pre_provision_tenant task because it is already running"
+        task_logger.warning(
+            "Skipping pre_provision_tenant — could not acquire provision lock"
         )
-        return
+        return False
 
     tenant_id: str | None = None
     try:
@@ -161,6 +183,7 @@ def pre_provision_tenant() -> None:
                 db_session.add(new_tenant)
                 db_session.commit()
                 task_logger.info(f"Successfully pre-provisioned tenant: {tenant_id}")
+                return True
             except Exception:
                 db_session.rollback()
                 task_logger.error(
@@ -184,5 +207,11 @@ def pre_provision_tenant() -> None:
                 asyncio.run(rollback_tenant_provisioning(tenant_id))
             except Exception:
                 task_logger.exception(f"Error during rollback for tenant: {tenant_id}")
+        return False
     finally:
-        lock_provision.release()
+        try:
+            lock_provision.release()
+        except Exception:
+            task_logger.warning(
+                "Could not release provision lock (likely expired), continuing"
+            )

backend/ee/onyx/background/celery/tasks/ttl_management/tasks.py

@@ -74,8 +74,7 @@ def perform_ttl_management_task(
 
     except Exception:
         logger.exception(
-            "delete_chat_session exceptioned. "
-            f"user_id={user_id} session_id={session_id}"
+            f"delete_chat_session exceptioned. user_id={user_id} session_id={session_id}"
         )
         with get_session_with_current_tenant() as db_session:
             mark_task_as_finished_with_id(

backend/ee/onyx/background/task_name_builders.py

@@ -7,7 +7,8 @@ QUERY_HISTORY_TASK_NAME_PREFIX = OnyxCeleryTask.EXPORT_QUERY_HISTORY_TASK
 
 
 def name_chat_ttl_task(
-    retention_limit_days: float, tenant_id: str | None = None  # noqa: ARG001
+    retention_limit_days: float,
+    tenant_id: str | None = None,  # noqa: ARG001
 ) -> str:
     return f"chat_ttl_{retention_limit_days}_days"
 

backend/ee/onyx/configs/app_configs.py

@@ -118,9 +118,7 @@ JWT_PUBLIC_KEY_URL: str | None = os.getenv("JWT_PUBLIC_KEY_URL", None)
 SUPER_USERS = json.loads(os.environ.get("SUPER_USERS", "[]"))
 SUPER_CLOUD_API_KEY = os.environ.get("SUPER_CLOUD_API_KEY", "api_key")
 
-# The posthog client does not accept empty API keys or hosts however it fails silently
-# when the capture is called. These defaults prevent Posthog issues from breaking the Onyx app
-POSTHOG_API_KEY = os.environ.get("POSTHOG_API_KEY") or "FooBar"
+POSTHOG_API_KEY = os.environ.get("POSTHOG_API_KEY")
 POSTHOG_HOST = os.environ.get("POSTHOG_HOST") or "https://us.i.posthog.com"
 POSTHOG_DEBUG_LOGS_ENABLED = (
     os.environ.get("POSTHOG_DEBUG_LOGS_ENABLED", "").lower() == "true"

backend/ee/onyx/db/analytics.py

@@ -31,7 +31,8 @@ def fetch_query_analytics(
             func.sum(case((ChatMessageFeedback.is_positive, 1), else_=0)),
             func.sum(
                 case(
-                    (ChatMessageFeedback.is_positive == False, 1), else_=0  # noqa: E712
+                    (ChatMessageFeedback.is_positive == False, 1),  # noqa: E712
+                    else_=0,  # noqa: E712
                 )
             ),
             cast(ChatMessage.time_sent, Date),
@@ -66,7 +67,8 @@ def fetch_per_user_query_analytics(
             func.sum(case((ChatMessageFeedback.is_positive, 1), else_=0)),
             func.sum(
                 case(
-                    (ChatMessageFeedback.is_positive == False, 1), else_=0  # noqa: E712
+                    (ChatMessageFeedback.is_positive == False, 1),  # noqa: E712
+                    else_=0,  # noqa: E712
                 )
             ),
             cast(ChatMessage.time_sent, Date),

backend/ee/onyx/db/connector_credential_pair.py

@@ -23,8 +23,7 @@ def _delete_connector_credential_pair_user_groups_relationship__no_commit(
     )
     if cc_pair is None:
         raise ValueError(
-            f"ConnectorCredentialPair with connector_id: {connector_id} "
-            f"and credential_id: {credential_id} not found"
+            f"ConnectorCredentialPair with connector_id: {connector_id} and credential_id: {credential_id} not found"
         )
 
     stmt = delete(UserGroup__ConnectorCredentialPair).where(

backend/ee/onyx/db/external_perm.py

@@ -123,8 +123,7 @@ def upsert_external_groups(
             user_id = email_id_map.get(user_email.lower())
             if user_id is None:
                 logger.warning(
-                    f"User in group {external_group.id}"
-                    f" with email {user_email} not found"
+                    f"User in group {external_group.id} with email {user_email} not found"
                 )
                 continue
 

backend/ee/onyx/db/hierarchy.py

@@ -18,7 +18,7 @@ from onyx.db.models import HierarchyNode
 
 
 def _build_hierarchy_access_filter(
-    user_email: str | None,
+    user_email: str,
     external_group_ids: list[str],
 ) -> ColumnElement[bool]:
     """Build SQLAlchemy filter for hierarchy node access.
@@ -43,7 +43,7 @@ def _build_hierarchy_access_filter(
 def _get_accessible_hierarchy_nodes_for_source(
     db_session: Session,
     source: DocumentSource,
-    user_email: str | None,
+    user_email: str,
     external_group_ids: list[str],
 ) -> list[HierarchyNode]:
     """

backend/ee/onyx/db/persona.py

@@ -7,6 +7,7 @@ from onyx.db.models import Persona
 from onyx.db.models import Persona__User
 from onyx.db.models import Persona__UserGroup
 from onyx.db.notification import create_notification
+from onyx.db.persona import mark_persona_user_files_for_sync
 from onyx.server.features.persona.models import PersonaSharedNotificationData
 
 
@@ -26,7 +27,9 @@ def update_persona_access(
 
     NOTE: Callers are responsible for committing."""
 
+    needs_sync = False
     if is_public is not None:
+        needs_sync = True
         persona = db_session.query(Persona).filter(Persona.id == persona_id).first()
         if persona:
             persona.is_public = is_public
@@ -35,6 +38,7 @@ def update_persona_access(
     # and a non-empty list means "replace with these shares".
 
     if user_ids is not None:
+        needs_sync = True
         db_session.query(Persona__User).filter(
             Persona__User.persona_id == persona_id
         ).delete(synchronize_session="fetch")
@@ -54,6 +58,7 @@ def update_persona_access(
                 )
 
     if group_ids is not None:
+        needs_sync = True
         db_session.query(Persona__UserGroup).filter(
             Persona__UserGroup.persona_id == persona_id
         ).delete(synchronize_session="fetch")
@@ -63,3 +68,7 @@ def update_persona_access(
             db_session.add(
                 Persona__UserGroup(persona_id=persona_id, user_group_id=group_id)
             )
+
+    # When sharing changes, user file ACLs need to be updated in the vector DB
+    if needs_sync:
+        mark_persona_user_files_for_sync(persona_id, db_session)

backend/ee/onyx/db/standard_answer.py

@@ -191,8 +191,7 @@ def create_initial_default_standard_answer_category(db_session: Session) -> None
     if default_category is not None:
         if default_category.name != default_category_name:
             raise ValueError(
-                "DB is not in a valid initial state. "
-                "Default standard answer category does not have expected name."
+                "DB is not in a valid initial state. Default standard answer category does not have expected name."
             )
         return
 

backend/ee/onyx/db/token_limit.py

@@ -115,8 +115,14 @@ def fetch_user_group_token_rate_limits_for_user(
     ordered: bool = True,
     get_editable: bool = True,
 ) -> Sequence[TokenRateLimit]:
-    stmt = select(TokenRateLimit)
-    stmt = stmt.where(User__UserGroup.user_group_id == group_id)
+    stmt = (
+        select(TokenRateLimit)
+        .join(
+            TokenRateLimit__UserGroup,
+            TokenRateLimit.id == TokenRateLimit__UserGroup.rate_limit_id,
+        )
+        .where(TokenRateLimit__UserGroup.user_group_id == group_id)
+    )
     stmt = _add_user_filters(stmt, user, get_editable)
 
     if enabled_only:

backend/ee/onyx/db/user_group.py

@@ -424,8 +424,7 @@ def fetch_user_groups_for_documents(
 def _check_user_group_is_modifiable(user_group: UserGroup) -> None:
     if not user_group.is_up_to_date:
         raise ValueError(
-            "Specified user group is currently syncing. Wait until the current "
-            "sync has finished before editing."
+            "Specified user group is currently syncing. Wait until the current sync has finished before editing."
         )
 
 
@@ -801,6 +800,33 @@ def update_user_group(
     return db_user_group
 
 
+def rename_user_group(
+    db_session: Session,
+    user_group_id: int,
+    new_name: str,
+) -> UserGroup:
+    stmt = select(UserGroup).where(UserGroup.id == user_group_id)
+    db_user_group = db_session.scalar(stmt)
+    if db_user_group is None:
+        raise ValueError(f"UserGroup with id '{user_group_id}' not found")
+
+    _check_user_group_is_modifiable(db_user_group)
+
+    db_user_group.name = new_name
+    db_user_group.time_last_modified_by_user = func.now()
+
+    # CC pair documents in Vespa contain the group name, so we need to
+    # trigger a sync to update them with the new name.
+    _mark_user_group__cc_pair_relationships_outdated__no_commit(
+        db_session=db_session, user_group_id=user_group_id
+    )
+    if not DISABLE_VECTOR_DB:
+        db_user_group.is_up_to_date = False
+
+    db_session.commit()
+    return db_user_group
+
+
 def prepare_user_group_for_deletion(db_session: Session, user_group_id: int) -> None:
     stmt = select(UserGroup).where(UserGroup.id == user_group_id)
     db_user_group = db_session.scalar(stmt)

backend/ee/onyx/external_permissions/github/utils.py

@@ -56,8 +56,7 @@ def _run_with_retry(
         if retry_count < MAX_RETRY_COUNT:
             sleep_after_rate_limit_exception(github_client)
             logger.warning(
-                f"Rate limit exceeded while {description}. Retrying... "
-                f"(attempt {retry_count + 1}/{MAX_RETRY_COUNT})"
+                f"Rate limit exceeded while {description}. Retrying... (attempt {retry_count + 1}/{MAX_RETRY_COUNT})"
             )
             return _run_with_retry(
                 operation, description, github_client, retry_count + 1
@@ -91,7 +90,9 @@ class TeamInfo(BaseModel):
 
 
 def _fetch_organization_members(
-    github_client: Github, org_name: str, retry_count: int = 0  # noqa: ARG001
+    github_client: Github,
+    org_name: str,
+    retry_count: int = 0,  # noqa: ARG001
 ) -> List[UserInfo]:
     """Fetch all organization members including owners and regular members."""
     org_members: List[UserInfo] = []
@@ -124,7 +125,9 @@ def _fetch_organization_members(
 
 
 def _fetch_repository_teams_detailed(
-    repo: Repository, github_client: Github, retry_count: int = 0  # noqa: ARG001
+    repo: Repository,
+    github_client: Github,
+    retry_count: int = 0,  # noqa: ARG001
 ) -> List[TeamInfo]:
     """Fetch teams with access to the repository and their members."""
     teams_data: List[TeamInfo] = []
@@ -167,7 +170,9 @@ def _fetch_repository_teams_detailed(
 
 
 def fetch_repository_team_slugs(
-    repo: Repository, github_client: Github, retry_count: int = 0  # noqa: ARG001
+    repo: Repository,
+    github_client: Github,
+    retry_count: int = 0,  # noqa: ARG001
 ) -> List[str]:
     """Fetch team slugs with access to the repository."""
     logger.info(f"Fetching team slugs for repository {repo.full_name}")

backend/ee/onyx/external_permissions/google_drive/doc_sync.py

@@ -68,6 +68,7 @@ def get_external_access_for_raw_gdrive_file(
     company_domain: str,
     retriever_drive_service: GoogleDriveService | None,
     admin_drive_service: GoogleDriveService,
+    fallback_user_email: str,
     add_prefix: bool = False,
 ) -> ExternalAccess:
     """
@@ -79,6 +80,11 @@ def get_external_access_for_raw_gdrive_file(
                 set add_prefix to True so group IDs are prefixed with the source type.
                 When invoked from doc_sync (permission sync), use the default (False)
                 since upsert_document_external_perms handles prefixing.
+    fallback_user_email: When we cannot retrieve any permission info for a file
+                (e.g. externally-owned files where the API returns no permissions
+                and permissions.list returns 403), fall back to granting access
+                to this user. This is typically the impersonated org user whose
+                drive contained the file.
     """
     doc_id = file.get("id")
     if not doc_id:
@@ -109,14 +115,33 @@ def get_external_access_for_raw_gdrive_file(
         )
         if len(permissions_list) != len(permission_ids) and retriever_drive_service:
             logger.warning(
-                f"Failed to get all permissions for file {doc_id} with retriever service, "
-                "trying admin service"
+                f"Failed to get all permissions for file {doc_id} with retriever service, trying admin service"
             )
             backup_permissions_list = _get_permissions(admin_drive_service)
             permissions_list = _merge_permissions_lists(
                 [permissions_list, backup_permissions_list]
             )
 
+    # For externally-owned files, the Drive API may return no permissions
+    # and permissions.list may return 403. In this case, fall back to
+    # granting access to the user who found the file in their drive.
+    # Note, even if other users also have access to this file,
+    # they will not be granted access in Onyx.
+    # We check permissions_list (the final result after all fetch attempts)
+    # rather than the raw fields, because permission_ids may be present
+    # but the actual fetch can still return empty due to a 403.
+    if not permissions_list:
+        logger.info(
+            f"No permission info available for file {doc_id} "
+            f"(likely owned by a user outside of your organization). "
+            f"Falling back to granting access to retriever user: {fallback_user_email}"
+        )
+        return ExternalAccess(
+            external_user_emails={fallback_user_email},
+            external_user_group_ids=set(),
+            is_public=False,
+        )
+
     folder_ids_to_inherit_permissions_from: set[str] = set()
     user_emails: set[str] = set()
     group_emails: set[str] = set()
@@ -140,9 +165,7 @@ def get_external_access_for_raw_gdrive_file(
                 user_emails.add(permission.email_address)
             else:
                 logger.error(
-                    "Permission is type `user` but no email address is "
-                    f"provided for document {doc_id}"
-                    f"\n {permission}"
+                    f"Permission is type `user` but no email address is provided for document {doc_id}\n {permission}"
                 )
         elif permission.type == PermissionType.GROUP:
             # groups are represented as email addresses within Drive
@@ -150,17 +173,14 @@ def get_external_access_for_raw_gdrive_file(
                 group_emails.add(permission.email_address)
             else:
                 logger.error(
-                    "Permission is type `group` but no email address is "
-                    f"provided for document {doc_id}"
-                    f"\n {permission}"
+                    f"Permission is type `group` but no email address is provided for document {doc_id}\n {permission}"
                 )
         elif permission.type == PermissionType.DOMAIN and company_domain:
             if permission.domain == company_domain:
                 public = True
             else:
                 logger.warning(
-                    "Permission is type domain but does not match company domain:"
-                    f"\n {permission}"
+                    f"Permission is type domain but does not match company domain:\n {permission}"
                 )
         elif permission.type == PermissionType.ANYONE:
             public = True

backend/ee/onyx/external_permissions/google_drive/folder_retrieval.py

@@ -18,10 +18,7 @@ logger = setup_logger()
 # Only include fields we need - folder ID and permissions
 # IMPORTANT: must fetch permissionIds, since sometimes the drive API
 # seems to miss permissions when requesting them directly
-FOLDER_PERMISSION_FIELDS = (
-    "nextPageToken, files(id, name, permissionIds, "
-    "permissions(id, emailAddress, type, domain, permissionDetails))"
-)
+FOLDER_PERMISSION_FIELDS = "nextPageToken, files(id, name, permissionIds, permissions(id, emailAddress, type, domain, permissionDetails))"
 
 
 def get_folder_permissions_by_ids(

backend/ee/onyx/external_permissions/google_drive/group_sync.py

@@ -142,8 +142,7 @@ def _drive_folder_to_onyx_group(
         elif permission.type == PermissionType.GROUP:
             if permission.email_address not in group_email_to_member_emails_map:
                 logger.warning(
-                    f"Group email {permission.email_address} for folder {folder.id} "
-                    "not found in group_email_to_member_emails_map"
+                    f"Group email {permission.email_address} for folder {folder.id} not found in group_email_to_member_emails_map"
                 )
                 continue
             folder_member_emails.update(
@@ -238,8 +237,7 @@ def _drive_member_map_to_onyx_groups(
         for group_email in group_emails:
             if group_email not in group_email_to_member_emails_map:
                 logger.warning(
-                    f"Group email {group_email} for drive {drive_id} not found in "
-                    "group_email_to_member_emails_map"
+                    f"Group email {group_email} for drive {drive_id} not found in group_email_to_member_emails_map"
                 )
                 continue
             drive_member_emails.update(group_email_to_member_emails_map[group_email])
@@ -326,8 +324,7 @@ def _build_onyx_groups(
         for group_email in group_emails:
             if group_email not in group_email_to_member_emails_map:
                 logger.warning(
-                    f"Group email {group_email} for drive {drive_id} not found in "
-                    "group_email_to_member_emails_map"
+                    f"Group email {group_email} for drive {drive_id} not found in group_email_to_member_emails_map"
                 )
                 continue
             drive_member_emails.update(group_email_to_member_emails_map[group_email])

backend/ee/onyx/external_permissions/google_drive/permission_retrieval.py

@@ -55,8 +55,7 @@ def get_permissions_by_ids(
     if len(filtered_permissions) < len(permission_ids):
         missing_ids = permission_id_set - {p.id for p in filtered_permissions if p.id}
         logger.warning(
-            f"Could not find all requested permission IDs for document {doc_id}. "
-            f"Missing IDs: {missing_ids}"
+            f"Could not find all requested permission IDs for document {doc_id}. Missing IDs: {missing_ids}"
         )
 
     return filtered_permissions

backend/ee/onyx/external_permissions/jira/group_sync.py

@@ -1,6 +1,8 @@
 from collections.abc import Generator
+from typing import Any
 
 from jira import JIRA
+from jira.exceptions import JIRAError
 
 from ee.onyx.db.external_perm import ExternalUserGroup
 from onyx.connectors.jira.utils import build_jira_client
@@ -9,107 +11,101 @@ from onyx.utils.logger import setup_logger
 
 logger = setup_logger()
 
+_ATLASSIAN_ACCOUNT_TYPE = "atlassian"
+_GROUP_MEMBER_PAGE_SIZE = 50
 
-def _get_jira_group_members_email(
+# The GET /group/member endpoint was introduced in Jira 6.0.
+# Jira versions older than 6.0 do not have group management REST APIs at all.
+_MIN_JIRA_VERSION_FOR_GROUP_MEMBER = "6.0"
+
+
+def _fetch_group_member_page(
     jira_client: JIRA,
     group_name: str,
-) -> list[str]:
-    """Get all member emails for a Jira group.
+    start_at: int,
+) -> dict[str, Any]:
+    """Fetch a single page from the non-deprecated GET /group/member endpoint.
 
-    Filters out app accounts (bots, integrations) and only returns real user emails.
+    The old GET /group endpoint (used by jira_client.group_members()) is deprecated
+    and decommissioned in Jira Server 10.3+. This uses the replacement endpoint
+    directly via the library's internal _get_json helper, following the same pattern
+    as enhanced_search_ids / bulk_fetch_issues in connector.py.
+
+    There is an open PR to the library to switch to this endpoint since last year:
+    https://github.com/pycontribs/jira/pull/2356
+    so once it is merged and released, we can switch to using the library function.
     """
-    emails: list[str] = []
-
     try:
-        # group_members returns an OrderedDict of account_id -> member_info
-        members = jira_client.group_members(group=group_name)
-
-        if not members:
-            logger.warning(f"No members found for group {group_name}")
-            return emails
-
-        for account_id, member_info in members.items():
-            # member_info is a dict with keys like 'fullname', 'email', 'active'
-            email = member_info.get("email")
-
-            # Skip "hidden" emails - these are typically app accounts
-            if email and email != "hidden":
-                emails.append(email)
-            else:
-                # For cloud, we might need to fetch user details separately
-                try:
-                    user = jira_client.user(id=account_id)
-
-                    # Skip app accounts (bots, integrations, etc.)
-                    if hasattr(user, "accountType") and user.accountType == "app":
-                        logger.info(
-                            f"Skipping app account {account_id} for group {group_name}"
-                        )
-                        continue
-
-                    if hasattr(user, "emailAddress") and user.emailAddress:
-                        emails.append(user.emailAddress)
-                    else:
-                        logger.warning(f"User {account_id} has no email address")
-                except Exception as e:
-                    logger.warning(
-                        f"Could not fetch email for user {account_id} in group {group_name}: {e}"
-                    )
-
-    except Exception as e:
-        logger.error(f"Error fetching members for group {group_name}: {e}")
-
-    return emails
+        return jira_client._get_json(
+            "group/member",
+            params={
+                "groupname": group_name,
+                "includeInactiveUsers": "false",
+                "startAt": start_at,
+                "maxResults": _GROUP_MEMBER_PAGE_SIZE,
+            },
+        )
+    except JIRAError as e:
+        if e.status_code == 404:
+            raise RuntimeError(
+                f"GET /group/member returned 404 for group '{group_name}'. "
+                f"This endpoint requires Jira {_MIN_JIRA_VERSION_FOR_GROUP_MEMBER}+. "
+                f"If you are running a self-hosted Jira instance, please upgrade "
+                f"to at least Jira {_MIN_JIRA_VERSION_FOR_GROUP_MEMBER}."
+            ) from e
+        raise
 
 
-def _build_group_member_email_map(
+def _get_group_member_emails(
     jira_client: JIRA,
-) -> dict[str, set[str]]:
-    """Build a map of group names to member emails."""
-    group_member_emails: dict[str, set[str]] = {}
+    group_name: str,
+) -> set[str]:
+    """Get all member emails for a single Jira group.
 
-    try:
-        # Get all groups from Jira - returns a list of group name strings
-        group_names = jira_client.groups()
+    Uses the non-deprecated GET /group/member endpoint which returns full user
+    objects including accountType, so we can filter out app/customer accounts
+    without making separate user() calls.
+    """
+    emails: set[str] = set()
+    start_at = 0
 
-        if not group_names:
-            logger.warning("No groups found in Jira")
-            return group_member_emails
+    while True:
+        try:
+            page = _fetch_group_member_page(jira_client, group_name, start_at)
+        except Exception as e:
+            logger.error(f"Error fetching members for group {group_name}: {e}")
+            raise
 
-        logger.info(f"Found {len(group_names)} groups in Jira")
-
-        for group_name in group_names:
-            if not group_name:
+        members: list[dict[str, Any]] = page.get("values", [])
+        for member in members:
+            account_type = member.get("accountType")
+            # On Jira DC < 9.0, accountType is absent; include those users.
+            # On Cloud / DC 9.0+, filter to real user accounts only.
+            if account_type is not None and account_type != _ATLASSIAN_ACCOUNT_TYPE:
                 continue
 
-            member_emails = _get_jira_group_members_email(
-                jira_client=jira_client,
-                group_name=group_name,
-            )
-
-            if member_emails:
-                group_member_emails[group_name] = set(member_emails)
-                logger.debug(
-                    f"Found {len(member_emails)} members for group {group_name}"
-                )
+            email = member.get("emailAddress")
+            if email:
+                emails.add(email)
             else:
-                logger.debug(f"No members found for group {group_name}")
+                logger.warning(
+                    f"Atlassian user {member.get('accountId', 'unknown')} in group {group_name} has no visible email address"
+                )
 
-    except Exception as e:
-        logger.error(f"Error building group member email map: {e}")
+        if page.get("isLast", True) or not members:
+            break
+        start_at += len(members)
 
-    return group_member_emails
+    return emails
 
 
 def jira_group_sync(
     tenant_id: str,  # noqa: ARG001
     cc_pair: ConnectorCredentialPair,
 ) -> Generator[ExternalUserGroup, None, None]:
-    """
-    Sync Jira groups and their members.
+    """Sync Jira groups and their members, yielding one group at a time.
 
-    This function fetches all groups from Jira and yields ExternalUserGroup
-    objects containing the group ID and member emails.
+    Streams group-by-group rather than accumulating all groups in memory.
     """
     jira_base_url = cc_pair.connector.connector_specific_config.get("jira_base_url", "")
     scoped_token = cc_pair.connector.connector_specific_config.get(
@@ -130,12 +126,26 @@ def jira_group_sync(
         scoped_token=scoped_token,
     )
 
-    group_member_email_map = _build_group_member_email_map(jira_client=jira_client)
-    if not group_member_email_map:
-        raise ValueError(f"No groups with members found for cc_pair_id={cc_pair.id}")
+    group_names = jira_client.groups()
+    if not group_names:
+        raise ValueError(f"No groups found for cc_pair_id={cc_pair.id}")
 
-    for group_id, group_member_emails in group_member_email_map.items():
-        yield ExternalUserGroup(
-            id=group_id,
-            user_emails=list(group_member_emails),
+    logger.info(f"Found {len(group_names)} groups in Jira")
+
+    for group_name in group_names:
+        if not group_name:
+            continue
+
+        member_emails = _get_group_member_emails(
+            jira_client=jira_client,
+            group_name=group_name,
+        )
+        if not member_emails:
+            logger.debug(f"No members found for group {group_name}")
+            continue
+
+        logger.debug(f"Found {len(member_emails)} members for group {group_name}")
+        yield ExternalUserGroup(
+            id=group_name,
+            user_emails=list(member_emails),
         )

backend/ee/onyx/external_permissions/post_query_censoring.py

@@ -69,8 +69,7 @@ def _post_query_chunk_censoring(
             censored_chunks = censor_chunks_for_source(chunks_for_source, user.email)
         except Exception as e:
             logger.exception(
-                f"Failed to censor chunks for source {source} so throwing out all"
-                f" chunks for this source and continuing: {e}"
+                f"Failed to censor chunks for source {source} so throwing out all chunks for this source and continuing: {e}"
             )
             continue
 

backend/ee/onyx/external_permissions/salesforce/postprocessing.py

@@ -23,7 +23,9 @@ ContentRange = tuple[int, int | None]  # (start_index, end_index) None means to
 
 # NOTE: Used for testing timing
 def _get_dummy_object_access_map(
-    object_ids: set[str], user_email: str, chunks: list[InferenceChunk]  # noqa: ARG001
+    object_ids: set[str],
+    user_email: str,  # noqa: ARG001
+    chunks: list[InferenceChunk],  # noqa: ARG001
 ) -> dict[str, bool]:
     time.sleep(0.15)
     # return {object_id: True for object_id in object_ids}

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

@@ -61,8 +61,7 @@ def _graph_api_get(
             ):
                 wait = min(int(resp.headers.get("Retry-After", str(2**attempt))), 60)
                 logger.warning(
-                    f"Graph API {resp.status_code} on attempt {attempt + 1}, "
-                    f"retrying in {wait}s: {url}"
+                    f"Graph API {resp.status_code} on attempt {attempt + 1}, retrying in {wait}s: {url}"
                 )
                 time.sleep(wait)
                 continue
@@ -72,8 +71,7 @@ def _graph_api_get(
             if attempt < GRAPH_API_MAX_RETRIES:
                 wait = min(2**attempt, 60)
                 logger.warning(
-                    f"Graph API connection error on attempt {attempt + 1}, "
-                    f"retrying in {wait}s: {url}"
+                    f"Graph API connection error on attempt {attempt + 1}, retrying in {wait}s: {url}"
                 )
                 time.sleep(wait)
                 continue
@@ -252,20 +250,24 @@ def _get_sharepoint_list_item_id(drive_item: DriveItem) -> str | None:
         raise e
 
 
-def _is_public_item(drive_item: DriveItem) -> bool:
-    is_public = False
+def _is_public_item(
+    drive_item: DriveItem,
+    treat_sharing_link_as_public: bool = False,
+) -> bool:
+    if not treat_sharing_link_as_public:
+        return False
+
     try:
         permissions = sleep_and_retry(
             drive_item.permissions.get_all(page_loaded=lambda _: None), "is_public_item"
         )
         for permission in permissions:
-            if permission.link and (
-                permission.link.scope == "anonymous"
-                or permission.link.scope == "organization"
+            if permission.link and permission.link.scope in (
+                "anonymous",
+                "organization",
             ):
-                is_public = True
-                break
-        return is_public
+                return True
+        return False
     except Exception as e:
         logger.error(f"Failed to check if item {drive_item.id} is public: {e}")
         return False
@@ -506,6 +508,7 @@ def get_external_access_from_sharepoint(
     drive_item: DriveItem | None,
     site_page: dict[str, Any] | None,
     add_prefix: bool = False,
+    treat_sharing_link_as_public: bool = False,
 ) -> ExternalAccess:
     """
     Get external access information from SharePoint.
@@ -565,8 +568,7 @@ def get_external_access_from_sharepoint(
                     )
 
     if drive_item and drive_name:
-        # Here we check if the item have have any public links, if so we return early
-        is_public = _is_public_item(drive_item)
+        is_public = _is_public_item(drive_item, treat_sharing_link_as_public)
         if is_public:
             logger.info(f"Item {drive_item.id} is public")
             return ExternalAccess(
@@ -767,8 +769,7 @@ def get_sharepoint_external_groups(
 
     if not enumerate_all_ad_groups or get_access_token is None:
         logger.info(
-            "Skipping exhaustive Azure AD group enumeration. "
-            "Only groups found in site role assignments are included."
+            "Skipping exhaustive Azure AD group enumeration. Only groups found in site role assignments are included."
         )
         return external_user_groups
 

backend/ee/onyx/external_permissions/slack/doc_sync.py

@@ -166,8 +166,7 @@ def slack_doc_sync(
     user_id_to_email_map = fetch_user_id_to_email_map(slack_client)
     if not user_id_to_email_map:
         raise ValueError(
-            "No user id to email map found. Please check to make sure that "
-            "your Slack bot token has the `users:read.email` scope"
+            "No user id to email map found. Please check to make sure that your Slack bot token has the `users:read.email` scope"
         )
 
     workspace_permissions = _fetch_workspace_permissions(

backend/ee/onyx/feature_flags/posthog_provider.py

@@ -34,6 +34,9 @@ class PostHogFeatureFlagProvider(FeatureFlagProvider):
         Returns:
             True if the feature is enabled for the user, False otherwise.
         """
+        if not posthog:
+            return False
+
         try:
             posthog.set(
                 distinct_id=user_id,

backend/ee/onyx/search/process_search_query.py

@@ -44,19 +44,21 @@ def _run_single_search(
     user: User,
     db_session: Session,
     num_hits: int | None = None,
+    hybrid_alpha: float | None = None,
 ) -> list[InferenceChunk]:
     """Execute a single search query and return chunks."""
     chunk_search_request = ChunkSearchRequest(
         query=query,
         user_selected_filters=filters,
         limit=num_hits,
+        hybrid_alpha=hybrid_alpha,
     )
 
     return search_pipeline(
         chunk_search_request=chunk_search_request,
         document_index=document_index,
         user=user,
-        persona=None,  # No persona for direct search
+        persona_search_info=None,
         db_session=db_session,
     )
 
@@ -74,7 +76,7 @@ def stream_search_query(
     Core search function that yields streaming packets.
     Used by both streaming and non-streaming endpoints.
     """
-    # Get document index
+    # Get document index.
     search_settings = get_current_search_settings(db_session)
     # This flow is for search so we do not get all indices.
     document_index = get_default_document_index(search_settings, None, db_session)
@@ -119,6 +121,7 @@ def stream_search_query(
             user=user,
             db_session=db_session,
             num_hits=request.num_hits,
+            hybrid_alpha=request.hybrid_alpha,
         )
     else:
         # Multiple queries - run in parallel and merge with RRF
@@ -133,6 +136,7 @@ def stream_search_query(
                     user,
                     db_session,
                     request.num_hits,
+                    request.hybrid_alpha,
                 ),
             )
             for query in all_executed_queries

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

@@ -157,7 +157,11 @@ def fetch_logo_helper(db_session: Session) -> Response:  # noqa: ARG001
             detail="No logo file found",
         )
     else:
-        return Response(content=onyx_file.data, media_type=onyx_file.mime_type)
+        return Response(
+            content=onyx_file.data,
+            media_type=onyx_file.mime_type,
+            headers={"Cache-Control": "no-cache"},
+        )
 
 
 def fetch_logotype_helper(db_session: Session) -> Response:  # noqa: ARG001

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

@@ -27,15 +27,17 @@ class SearchFlowClassificationResponse(BaseModel):
     is_search_flow: bool
 
 
-# NOTE: This model is used for the core flow of the Onyx application, any changes to it should be reviewed and approved by an
-# experienced team member. It is very important to 1. avoid bloat and 2. that this remains backwards compatible across versions.
+# NOTE: This model is used for the core flow of the Onyx application, any
+# changes to it should be reviewed and approved by an experienced team member.
+# It is very important to 1. avoid bloat and 2. that this remains backwards
+# compatible across versions.
 class SendSearchQueryRequest(BaseModel):
     search_query: str
     filters: BaseFilters | None = None
     num_docs_fed_to_llm_selection: int | None = None
     run_query_expansion: bool = False
     num_hits: int = 30
-
+    hybrid_alpha: float | None = None
     include_content: bool = False
     stream: bool = False
 

backend/ee/onyx/server/query_and_chat/search_backend.py

@@ -20,6 +20,7 @@ from ee.onyx.server.query_and_chat.models import SearchQueryResponse
 from ee.onyx.server.query_and_chat.models import SendSearchQueryRequest
 from ee.onyx.server.query_and_chat.streaming_models import SearchErrorPacket
 from onyx.auth.users import current_user
+from onyx.configs.app_configs import ONYX_SEARCH_UI_USES_OPENSEARCH_KEYWORD_SEARCH
 from onyx.db.engine.sql_engine import get_session
 from onyx.db.engine.sql_engine import get_session_with_current_tenant
 from onyx.db.models import User
@@ -67,8 +68,10 @@ def search_flow_classification(
     return SearchFlowClassificationResponse(is_search_flow=is_search_flow)
 
 
-# NOTE: This endpoint is used for the core flow of the Onyx application, any changes to it should be reviewed and approved by an
-# experienced team member. It is very important to 1. avoid bloat and 2. that this remains backwards compatible across versions.
+# NOTE: This endpoint is used for the core flow of the Onyx application, any
+# changes to it should be reviewed and approved by an experienced team member.
+# It is very important to 1. avoid bloat and 2. that this remains backwards
+# compatible across versions.
 @router.post(
     "/send-search-message",
     response_model=None,
@@ -80,13 +83,19 @@ def handle_send_search_message(
     db_session: Session = Depends(get_session),
 ) -> StreamingResponse | SearchFullResponse:
     """
-    Execute a search query with optional streaming.
+    Executes a search query with optional streaming.
 
-    When stream=True: Returns StreamingResponse with SSE
-    When stream=False: Returns SearchFullResponse
+    If hybrid_alpha is unset and ONYX_SEARCH_UI_USES_OPENSEARCH_KEYWORD_SEARCH
+    is True, executes pure keyword search.
+
+    Returns:
+        StreamingResponse with SSE if stream=True, otherwise SearchFullResponse.
     """
     logger.debug(f"Received search query: {request.search_query}")
 
+    if request.hybrid_alpha is None and ONYX_SEARCH_UI_USES_OPENSEARCH_KEYWORD_SEARCH:
+        request.hybrid_alpha = 0.0
+
     # Non-streaming path
     if not request.stream:
         try:

backend/ee/onyx/server/reporting/usage_export_generation.py

@@ -152,10 +152,7 @@ def create_new_usage_report(
         zip_buffer.seek(0)
 
         # store zip blob to file_store
-        report_name = (
-            f"{datetime.now(tz=timezone.utc).strftime('%Y-%m-%d')}"
-            f"_{report_id}_usage_report.zip"
-        )
+        report_name = f"{datetime.now(tz=timezone.utc).strftime('%Y-%m-%d')}_{report_id}_usage_report.zip"
         file_store.save_file(
             content=zip_buffer,
             display_name=report_name,

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

@@ -449,8 +449,7 @@ def _apply_group_remove(
     match = _MEMBER_FILTER_RE.match(op.path)
     if not match:
         raise ScimPatchError(
-            f"Unsupported remove path '{op.path}'. "
-            'Expected: members[value eq "user-id"]'
+            f"Unsupported remove path '{op.path}'. Expected: members[value eq \"user-id\"]"
         )
 
     target_id = match.group(1)

backend/ee/onyx/server/seeding.py

@@ -26,6 +26,7 @@ from onyx.db.models import Tool
 from onyx.db.persona import upsert_persona
 from onyx.server.features.persona.models import PersonaUpsertRequest
 from onyx.server.manage.llm.models import LLMProviderUpsertRequest
+from onyx.server.manage.llm.models import LLMProviderView
 from onyx.server.settings.models import Settings
 from onyx.server.settings.store import store_settings as store_base_settings
 from onyx.utils.logger import setup_logger
@@ -125,10 +126,16 @@ def _seed_llms(
         existing = fetch_existing_llm_provider(name=request.name, db_session=db_session)
         if existing:
             request.id = existing.id
-    seeded_providers = [
-        upsert_llm_provider(llm_upsert_request, db_session)
-        for llm_upsert_request in llm_upsert_requests
-    ]
+    seeded_providers: list[LLMProviderView] = []
+    for llm_upsert_request in llm_upsert_requests:
+        try:
+            seeded_providers.append(upsert_llm_provider(llm_upsert_request, db_session))
+        except ValueError as e:
+            logger.warning(
+                "Failed to upsert LLM provider '%s' during seeding: %s",
+                llm_upsert_request.name,
+                e,
+            )
 
     default_provider = next(
         (p for p in seeded_providers if p.model_configurations), None
@@ -171,7 +178,7 @@ def _seed_personas(db_session: Session, personas: list[PersonaUpsertRequest]) ->
                     system_prompt=persona.system_prompt,
                     task_prompt=persona.task_prompt,
                     datetime_aware=persona.datetime_aware,
-                    featured=persona.featured,
+                    is_featured=persona.is_featured,
                     commit=False,
                 )
             db_session.commit()

backend/ee/onyx/server/tenants/provisioning.py

@@ -29,7 +29,6 @@ from onyx.configs.app_configs import OPENAI_DEFAULT_API_KEY
 from onyx.configs.app_configs import OPENROUTER_DEFAULT_API_KEY
 from onyx.configs.app_configs import VERTEXAI_DEFAULT_CREDENTIALS
 from onyx.configs.app_configs import VERTEXAI_DEFAULT_LOCATION
-from onyx.configs.constants import MilestoneRecordType
 from onyx.db.engine.sql_engine import get_session_with_shared_schema
 from onyx.db.engine.sql_engine import get_session_with_tenant
 from onyx.db.image_generation import create_default_image_gen_config_from_api_key
@@ -59,7 +58,6 @@ from onyx.server.manage.llm.models import LLMProviderUpsertRequest
 from onyx.server.manage.llm.models import ModelConfigurationUpsertRequest
 from onyx.setup import setup_onyx
 from onyx.utils.logger import setup_logger
-from onyx.utils.telemetry import mt_cloud_telemetry
 from shared_configs.configs import MULTI_TENANT
 from shared_configs.configs import POSTGRES_DEFAULT_SCHEMA
 from shared_configs.configs import TENANT_ID_PREFIX
@@ -71,7 +69,9 @@ logger = setup_logger()
 
 
 async def get_or_provision_tenant(
-    email: str, referral_source: str | None = None, request: Request | None = None
+    email: str,
+    referral_source: str | None = None,
+    request: Request | None = None,
 ) -> str:
     """
     Get existing tenant ID for an email or create a new tenant if none exists.
@@ -123,7 +123,8 @@ async def get_or_provision_tenant(
 
 
 async def create_tenant(
-    email: str, referral_source: str | None = None  # noqa: ARG001
+    email: str,
+    referral_source: str | None = None,  # noqa: ARG001
 ) -> str:
     """
     Create a new tenant on-demand when no pre-provisioned tenants are available.
@@ -679,7 +680,9 @@ async def setup_tenant(tenant_id: str) -> None:
 
 
 async def assign_tenant_to_user(
-    tenant_id: str, email: str, referral_source: str | None = None  # noqa: ARG001
+    tenant_id: str,
+    email: str,
+    referral_source: str | None = None,  # noqa: ARG001
 ) -> None:
     """
     Assign a tenant to a user and perform necessary operations.
@@ -690,12 +693,6 @@ async def assign_tenant_to_user(
 
     try:
         add_users_to_tenant([email], tenant_id)
-
-        mt_cloud_telemetry(
-            tenant_id=tenant_id,
-            distinct_id=email,
-            event=MilestoneRecordType.TENANT_CREATED,
-        )
     except Exception:
         logger.exception(f"Failed to assign tenant {tenant_id} to user {email}")
         raise Exception("Failed to assign tenant to user")

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

@@ -4,6 +4,7 @@ from fastapi import HTTPException
 from sqlalchemy.exc import IntegrityError
 from sqlalchemy.orm import Session
 
+from ee.onyx.db.persona import update_persona_access
 from ee.onyx.db.user_group import add_users_to_user_group
 from ee.onyx.db.user_group import delete_user_group as db_delete_user_group
 from ee.onyx.db.user_group import fetch_user_group
@@ -11,13 +12,16 @@ from ee.onyx.db.user_group import fetch_user_groups
 from ee.onyx.db.user_group import fetch_user_groups_for_user
 from ee.onyx.db.user_group import insert_user_group
 from ee.onyx.db.user_group import prepare_user_group_for_deletion
+from ee.onyx.db.user_group import rename_user_group
 from ee.onyx.db.user_group import update_user_curator_relationship
 from ee.onyx.db.user_group import update_user_group
 from ee.onyx.server.user_group.models import AddUsersToUserGroupRequest
 from ee.onyx.server.user_group.models import MinimalUserGroupSnapshot
 from ee.onyx.server.user_group.models import SetCuratorRequest
+from ee.onyx.server.user_group.models import UpdateGroupAgentsRequest
 from ee.onyx.server.user_group.models import UserGroup
 from ee.onyx.server.user_group.models import UserGroupCreate
+from ee.onyx.server.user_group.models import UserGroupRename
 from ee.onyx.server.user_group.models import UserGroupUpdate
 from onyx.auth.users import current_admin_user
 from onyx.auth.users import current_curator_or_admin_user
@@ -27,6 +31,9 @@ from onyx.configs.constants import PUBLIC_API_TAGS
 from onyx.db.engine.sql_engine import get_session
 from onyx.db.models import User
 from onyx.db.models import UserRole
+from onyx.db.persona import get_persona_by_id
+from onyx.error_handling.error_codes import OnyxErrorCode
+from onyx.error_handling.exceptions import OnyxError
 from onyx.utils.logger import setup_logger
 
 logger = setup_logger()
@@ -87,6 +94,32 @@ def create_user_group(
     return UserGroup.from_model(db_user_group)
 
 
+@router.patch("/admin/user-group/rename")
+def rename_user_group_endpoint(
+    rename_request: UserGroupRename,
+    _: User = Depends(current_admin_user),
+    db_session: Session = Depends(get_session),
+) -> UserGroup:
+    try:
+        return UserGroup.from_model(
+            rename_user_group(
+                db_session=db_session,
+                user_group_id=rename_request.id,
+                new_name=rename_request.name,
+            )
+        )
+    except IntegrityError:
+        raise OnyxError(
+            OnyxErrorCode.DUPLICATE_RESOURCE,
+            f"User group with name '{rename_request.name}' already exists.",
+        )
+    except ValueError as e:
+        msg = str(e)
+        if "not found" in msg.lower():
+            raise OnyxError(OnyxErrorCode.NOT_FOUND, msg)
+        raise OnyxError(OnyxErrorCode.CONFLICT, msg)
+
+
 @router.patch("/admin/user-group/{user_group_id}")
 def patch_user_group(
     user_group_id: int,
@@ -161,3 +194,38 @@ def delete_user_group(
         user_group = fetch_user_group(db_session, user_group_id)
         if user_group:
             db_delete_user_group(db_session, user_group)
+
+
+@router.patch("/admin/user-group/{user_group_id}/agents")
+def update_group_agents(
+    user_group_id: int,
+    request: UpdateGroupAgentsRequest,
+    user: User = Depends(current_admin_user),
+    db_session: Session = Depends(get_session),
+) -> None:
+    for agent_id in request.added_agent_ids:
+        persona = get_persona_by_id(
+            persona_id=agent_id, user=user, db_session=db_session
+        )
+        current_group_ids = [g.id for g in persona.groups]
+        if user_group_id not in current_group_ids:
+            update_persona_access(
+                persona_id=agent_id,
+                creator_user_id=user.id,
+                db_session=db_session,
+                group_ids=current_group_ids + [user_group_id],
+            )
+
+    for agent_id in request.removed_agent_ids:
+        persona = get_persona_by_id(
+            persona_id=agent_id, user=user, db_session=db_session
+        )
+        current_group_ids = [g.id for g in persona.groups]
+        update_persona_access(
+            persona_id=agent_id,
+            creator_user_id=user.id,
+            db_session=db_session,
+            group_ids=[gid for gid in current_group_ids if gid != user_group_id],
+        )
+
+    db_session.commit()

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

@@ -104,6 +104,16 @@ class AddUsersToUserGroupRequest(BaseModel):
     user_ids: list[UUID]
 
 
+class UserGroupRename(BaseModel):
+    id: int
+    name: str
+
+
 class SetCuratorRequest(BaseModel):
     user_id: UUID
     is_curator: bool
+
+
+class UpdateGroupAgentsRequest(BaseModel):
+    added_agent_ids: list[int]
+    removed_agent_ids: list[int]

backend/ee/onyx/utils/encryption.py

@@ -14,67 +14,90 @@ from onyx.utils.variable_functionality import fetch_versioned_implementation
 logger = setup_logger()
 
 
-@lru_cache(maxsize=1)
+@lru_cache(maxsize=2)
 def _get_trimmed_key(key: str) -> bytes:
     encoded_key = key.encode()
     key_length = len(encoded_key)
     if key_length < 16:
         raise RuntimeError("Invalid ENCRYPTION_KEY_SECRET - too short")
-    elif key_length > 32:
-        key = key[:32]
-    elif key_length not in (16, 24, 32):
-        valid_lengths = [16, 24, 32]
-        key = key[: min(valid_lengths, key=lambda x: abs(x - key_length))]
 
-    return encoded_key
+    # Trim to the largest valid AES key size that fits
+    valid_lengths = [32, 24, 16]
+    for size in valid_lengths:
+        if key_length >= size:
+            return encoded_key[:size]
+
+    raise AssertionError("unreachable")
 
 
-def _encrypt_string(input_str: str) -> bytes:
-    if not ENCRYPTION_KEY_SECRET:
+def _encrypt_string(input_str: str, key: str | None = None) -> bytes:
+    effective_key = key if key is not None else ENCRYPTION_KEY_SECRET
+    if not effective_key:
         return input_str.encode()
 
-    key = _get_trimmed_key(ENCRYPTION_KEY_SECRET)
+    trimmed = _get_trimmed_key(effective_key)
     iv = urandom(16)
     padder = padding.PKCS7(algorithms.AES.block_size).padder()
     padded_data = padder.update(input_str.encode()) + padder.finalize()
 
-    cipher = Cipher(algorithms.AES(key), modes.CBC(iv), backend=default_backend())
+    cipher = Cipher(algorithms.AES(trimmed), modes.CBC(iv), backend=default_backend())
     encryptor = cipher.encryptor()
     encrypted_data = encryptor.update(padded_data) + encryptor.finalize()
 
     return iv + encrypted_data
 
 
-def _decrypt_bytes(input_bytes: bytes) -> str:
-    if not ENCRYPTION_KEY_SECRET:
+def _decrypt_bytes(input_bytes: bytes, key: str | None = None) -> str:
+    effective_key = key if key is not None else ENCRYPTION_KEY_SECRET
+    if not effective_key:
         return input_bytes.decode()
 
-    key = _get_trimmed_key(ENCRYPTION_KEY_SECRET)
-    iv = input_bytes[:16]
-    encrypted_data = input_bytes[16:]
+    trimmed = _get_trimmed_key(effective_key)
+    try:
+        iv = input_bytes[:16]
+        encrypted_data = input_bytes[16:]
 
-    cipher = Cipher(algorithms.AES(key), modes.CBC(iv), backend=default_backend())
-    decryptor = cipher.decryptor()
-    decrypted_padded_data = decryptor.update(encrypted_data) + decryptor.finalize()
+        cipher = Cipher(
+            algorithms.AES(trimmed), modes.CBC(iv), backend=default_backend()
+        )
+        decryptor = cipher.decryptor()
+        decrypted_padded_data = decryptor.update(encrypted_data) + decryptor.finalize()
 
-    unpadder = padding.PKCS7(algorithms.AES.block_size).unpadder()
-    decrypted_data = unpadder.update(decrypted_padded_data) + unpadder.finalize()
+        unpadder = padding.PKCS7(algorithms.AES.block_size).unpadder()
+        decrypted_data = unpadder.update(decrypted_padded_data) + unpadder.finalize()
 
-    return decrypted_data.decode()
+        return decrypted_data.decode()
+    except (ValueError, UnicodeDecodeError):
+        if key is not None:
+            # Explicit key was provided — don't fall back silently
+            raise
+        # Read path: attempt raw UTF-8 decode as a fallback for legacy data.
+        # Does NOT handle data encrypted with a different key — that
+        # ciphertext is not valid UTF-8 and will raise below.
+        logger.warning(
+            "AES decryption failed — falling back to raw decode. Run the re-encrypt secrets script to rotate to the current key."
+        )
+        try:
+            return input_bytes.decode()
+        except UnicodeDecodeError:
+            raise ValueError(
+                "Data is not valid UTF-8 — likely encrypted with a different key. "
+                "Run the re-encrypt secrets script to rotate to the current key."
+            ) from None
 
 
-def encrypt_string_to_bytes(input_str: str) -> bytes:
+def encrypt_string_to_bytes(input_str: str, key: str | None = None) -> bytes:
     versioned_encryption_fn = fetch_versioned_implementation(
         "onyx.utils.encryption", "_encrypt_string"
     )
-    return versioned_encryption_fn(input_str)
+    return versioned_encryption_fn(input_str, key=key)
 
 
-def decrypt_bytes_to_string(input_bytes: bytes) -> str:
+def decrypt_bytes_to_string(input_bytes: bytes, key: str | None = None) -> str:
     versioned_decryption_fn = fetch_versioned_implementation(
         "onyx.utils.encryption", "_decrypt_bytes"
     )
-    return versioned_decryption_fn(input_bytes)
+    return versioned_decryption_fn(input_bytes, key=key)
 
 
 def test_encryption() -> None:

backend/ee/onyx/utils/posthog_client.py

@@ -9,6 +9,7 @@ from ee.onyx.configs.app_configs import POSTHOG_API_KEY
 from ee.onyx.configs.app_configs import POSTHOG_DEBUG_LOGS_ENABLED
 from ee.onyx.configs.app_configs import POSTHOG_HOST
 from onyx.utils.logger import setup_logger
+from shared_configs.configs import MULTI_TENANT
 
 logger = setup_logger()
 
@@ -18,12 +19,19 @@ def posthog_on_error(error: Any, items: Any) -> None:
     logger.error(f"PostHog error: {error}, items: {items}")
 
 
-posthog = Posthog(
-    project_api_key=POSTHOG_API_KEY,
-    host=POSTHOG_HOST,
-    debug=POSTHOG_DEBUG_LOGS_ENABLED,
-    on_error=posthog_on_error,
-)
+posthog: Posthog | None = None
+if POSTHOG_API_KEY:
+    posthog = Posthog(
+        project_api_key=POSTHOG_API_KEY,
+        host=POSTHOG_HOST,
+        debug=POSTHOG_DEBUG_LOGS_ENABLED,
+        on_error=posthog_on_error,
+    )
+elif MULTI_TENANT:
+    logger.warning(
+        "POSTHOG_API_KEY is not set but MULTI_TENANT is enabled — "
+        "PostHog telemetry and feature flags will be disabled"
+    )
 
 # For cross referencing between cloud and www Onyx sites
 # NOTE: These clients are separate because they are separate posthog projects.
@@ -60,7 +68,7 @@ def capture_and_sync_with_alternate_posthog(
         logger.error(f"Error capturing marketing posthog event: {e}")
 
     try:
-        if cloud_user_id := props.get("onyx_cloud_user_id"):
+        if posthog and (cloud_user_id := props.get("onyx_cloud_user_id")):
             cloud_props = props.copy()
             cloud_props.pop("onyx_cloud_user_id", None)
 
@@ -72,15 +80,45 @@ def capture_and_sync_with_alternate_posthog(
         logger.error(f"Error identifying cloud posthog user: {e}")
 
 
+def alias_user(distinct_id: str, anonymous_id: str) -> None:
+    """Link an anonymous distinct_id to an identified user, merging person profiles.
+
+    No-ops when the IDs match (e.g. returning users whose PostHog cookie
+    already contains their identified user ID).
+    """
+    if not posthog or anonymous_id == distinct_id:
+        return
+
+    try:
+        posthog.alias(previous_id=anonymous_id, distinct_id=distinct_id)
+        posthog.flush()
+    except Exception as e:
+        logger.error(f"Error aliasing PostHog user: {e}")
+
+
+def get_anon_id_from_request(request: Any) -> str | None:
+    """Extract the anonymous distinct_id from the app PostHog cookie on a request."""
+    if not POSTHOG_API_KEY:
+        return None
+
+    cookie_name = f"ph_{POSTHOG_API_KEY}_posthog"
+    if (cookie_value := request.cookies.get(cookie_name)) and (
+        parsed := parse_posthog_cookie(cookie_value)
+    ):
+        return parsed.get("distinct_id")
+
+    return None
+
+
 def get_marketing_posthog_cookie_name() -> str | None:
     if not MARKETING_POSTHOG_API_KEY:
         return None
     return f"onyx_custom_ph_{MARKETING_POSTHOG_API_KEY}_posthog"
 
 
-def parse_marketing_cookie(cookie_value: str) -> dict[str, Any] | None:
+def parse_posthog_cookie(cookie_value: str) -> dict[str, Any] | None:
     """
-    Parse the URL-encoded JSON marketing cookie.
+    Parse a URL-encoded JSON PostHog cookie
 
     Expected format (URL-encoded):
     {"distinct_id":"...", "featureFlags":{"landing_page_variant":"..."}, ...}
@@ -94,7 +132,7 @@ def parse_marketing_cookie(cookie_value: str) -> dict[str, Any] | None:
         cookie_data = json.loads(decoded_cookie)
 
         distinct_id = cookie_data.get("distinct_id")
-        if not distinct_id:
+        if not distinct_id or not isinstance(distinct_id, str):
             return None
 
         return cookie_data

backend/ee/onyx/utils/telemetry.py

@@ -1,3 +1,5 @@
+from typing import Any
+
 from ee.onyx.utils.posthog_client import posthog
 from onyx.utils.logger import setup_logger
 
@@ -5,12 +7,27 @@ logger = setup_logger()
 
 
 def event_telemetry(
-    distinct_id: str, event: str, properties: dict | None = None
+    distinct_id: str, event: str, properties: dict[str, Any] | None = None
 ) -> None:
     """Capture and send an event to PostHog, flushing immediately."""
+    if not posthog:
+        return
+
     logger.info(f"Capturing PostHog event: {distinct_id} {event} {properties}")
     try:
         posthog.capture(distinct_id, event, properties)
         posthog.flush()
     except Exception as e:
         logger.error(f"Error capturing PostHog event: {e}")
+
+
+def identify_user(distinct_id: str, properties: dict[str, Any] | None = None) -> None:
+    """Create/update a PostHog person profile, flushing immediately."""
+    if not posthog:
+        return
+
+    try:
+        posthog.identify(distinct_id, properties)
+        posthog.flush()
+    except Exception as e:
+        logger.error(f"Error identifying PostHog user: {e}")

backend/onyx/access/access.py

@@ -1,7 +1,6 @@
 from collections.abc import Callable
 from typing import cast
 
-from sqlalchemy.orm import joinedload
 from sqlalchemy.orm import Session
 
 from onyx.access.models import DocumentAccess
@@ -12,6 +11,7 @@ from onyx.db.document import get_access_info_for_document
 from onyx.db.document import get_access_info_for_documents
 from onyx.db.models import User
 from onyx.db.models import UserFile
+from onyx.db.user_file import fetch_user_files_with_access_relationships
 from onyx.utils.variable_functionality import fetch_ee_implementation_or_noop
 from onyx.utils.variable_functionality import fetch_versioned_implementation
 
@@ -96,7 +96,9 @@ def get_access_for_documents(
     return versioned_get_access_for_documents_fn(document_ids, db_session)
 
 
-def _get_acl_for_user(user: User, db_session: Session) -> set[str]:  # noqa: ARG001
+def _get_acl_for_user(
+    user: User, db_session: Session  # noqa: ARG001
+) -> set[str]:  # noqa: ARG001
     """Returns a list of ACL entries that the user has access to. This is meant to be
     used downstream to filter out documents that the user does not have access to. The
     user should have access to a document if at least one entry in the document's ACL
@@ -132,19 +134,61 @@ def get_access_for_user_files(
     user_file_ids: list[str],
     db_session: Session,
 ) -> dict[str, DocumentAccess]:
-    user_files = (
-        db_session.query(UserFile)
-        .options(joinedload(UserFile.user))  # Eager load the user relationship
-        .filter(UserFile.id.in_(user_file_ids))
-        .all()
+    versioned_fn = fetch_versioned_implementation(
+        "onyx.access.access", "get_access_for_user_files_impl"
     )
-    return {
-        str(user_file.id): DocumentAccess.build(
-            user_emails=[user_file.user.email] if user_file.user else [],
+    return versioned_fn(user_file_ids, db_session)
+
+
+def get_access_for_user_files_impl(
+    user_file_ids: list[str],
+    db_session: Session,
+) -> dict[str, DocumentAccess]:
+    user_files = fetch_user_files_with_access_relationships(user_file_ids, db_session)
+    return build_access_for_user_files_impl(user_files)
+
+
+def build_access_for_user_files(
+    user_files: list[UserFile],
+) -> dict[str, DocumentAccess]:
+    """Compute access from pre-loaded UserFile objects (with relationships).
+    Callers must ensure UserFile.user, Persona.users, and Persona.user are
+    eagerly loaded (and Persona.groups for the EE path)."""
+    versioned_fn = fetch_versioned_implementation(
+        "onyx.access.access", "build_access_for_user_files_impl"
+    )
+    return versioned_fn(user_files)
+
+
+def build_access_for_user_files_impl(
+    user_files: list[UserFile],
+) -> dict[str, DocumentAccess]:
+    result: dict[str, DocumentAccess] = {}
+    for user_file in user_files:
+        emails, is_public = collect_user_file_access(user_file)
+        result[str(user_file.id)] = DocumentAccess.build(
+            user_emails=list(emails),
             user_groups=[],
-            is_public=True if user_file.user is None else False,
+            is_public=is_public,
             external_user_emails=[],
             external_user_group_ids=[],
         )
-        for user_file in user_files
-    }
+    return result
+
+
+def collect_user_file_access(user_file: UserFile) -> tuple[set[str], bool]:
+    """Collect all user emails that should have access to this user file.
+    Includes the owner plus any users who have access via shared personas.
+    Returns (emails, is_public)."""
+    emails: set[str] = {user_file.user.email}
+    is_public = False
+    for persona in user_file.assistants:
+        if persona.deleted:
+            continue
+        if persona.is_public:
+            is_public = True
+        if persona.user_id is not None and persona.user:
+            emails.add(persona.user.email)
+        for shared_user in persona.users:
+            emails.add(shared_user.email)
+    return emails, is_public

backend/onyx/access/hierarchy_access.py

@@ -5,7 +5,8 @@ from onyx.utils.variable_functionality import fetch_versioned_implementation
 
 
 def _get_user_external_group_ids(
-    db_session: Session, user: User  # noqa: ARG001
+    db_session: Session,  # noqa: ARG001
+    user: User,  # noqa: ARG001
 ) -> list[str]:
     return []
 

backend/onyx/access/models.py

@@ -8,7 +8,6 @@ from onyx.configs.constants import PUBLIC_DOC_PAT
 
 @dataclass(frozen=True)
 class ExternalAccess:
-
     # arbitrary limit to prevent excessively large permissions sets
     # not internally enforced ... the caller can check this before using the instance
     MAX_NUM_ENTRIES = 5000

backend/onyx/auth/captcha.py

@@ -96,8 +96,7 @@ async def verify_captcha_token(
                     )
 
             logger.debug(
-                f"Captcha verification passed: score={result.score}, "
-                f"action={result.action}"
+                f"Captcha verification passed: score={result.score}, action={result.action}"
             )
 
     except httpx.HTTPError as e:

backend/onyx/auth/email_utils.py

@@ -353,20 +353,11 @@ def build_user_email_invite(
             "or login with Google and complete your registration.</p>"
         )
     elif auth_type == AuthType.BASIC:
-        message += (
-            "<p>To join the organization, please click the button below to set a password "
-            "and complete your registration.</p>"
-        )
+        message += "<p>To join the organization, please click the button below to set a password and complete your registration.</p>"
     elif auth_type == AuthType.GOOGLE_OAUTH:
-        message += (
-            "<p>To join the organization, please click the button below to login with Google "
-            "and complete your registration.</p>"
-        )
+        message += "<p>To join the organization, please click the button below to login with Google and complete your registration.</p>"
     elif auth_type == AuthType.OIDC or auth_type == AuthType.SAML:
-        message += (
-            "<p>To join the organization, please click the button below to"
-            " complete your registration.</p>"
-        )
+        message += "<p>To join the organization, please click the button below to complete your registration.</p>"
     else:
         raise ValueError(f"Invalid auth type: {auth_type}")
 

backend/onyx/auth/users.py

@@ -1,4 +1,7 @@
+import base64
+import hashlib
 import json
+import os
 import random
 import secrets
 import string
@@ -16,6 +19,7 @@ from typing import Optional
 from typing import Protocol
 from typing import Tuple
 from typing import TypeVar
+from urllib.parse import urlparse
 
 import jwt
 from email_validator import EmailNotValidError
@@ -28,6 +32,8 @@ from fastapi import Query
 from fastapi import Request
 from fastapi import Response
 from fastapi import status
+from fastapi import WebSocket
+from fastapi.responses import JSONResponse
 from fastapi.responses import RedirectResponse
 from fastapi.security import OAuth2PasswordRequestForm
 from fastapi_users import BaseUserManager
@@ -54,6 +60,7 @@ from fastapi_users.router.common import ErrorModel
 from fastapi_users_db_sqlalchemy import SQLAlchemyUserDatabase
 from httpx_oauth.integrations.fastapi import OAuth2AuthorizeCallback
 from httpx_oauth.oauth2 import BaseOAuth2
+from httpx_oauth.oauth2 import GetAccessTokenError
 from httpx_oauth.oauth2 import OAuth2Token
 from pydantic import BaseModel
 from sqlalchemy import nulls_last
@@ -119,10 +126,18 @@ from onyx.db.models import Persona
 from onyx.db.models import User
 from onyx.db.pat import fetch_user_for_pat
 from onyx.db.users import get_user_by_email
+from onyx.error_handling.error_codes import OnyxErrorCode
+from onyx.error_handling.exceptions import log_onyx_error
+from onyx.error_handling.exceptions import onyx_error_to_json_response
+from onyx.error_handling.exceptions import OnyxError
 from onyx.redis.redis_pool import get_async_redis_connection
+from onyx.redis.redis_pool import retrieve_ws_token_data
 from onyx.server.settings.store import load_settings
 from onyx.server.utils import BasicAuthenticationError
 from onyx.utils.logger import setup_logger
+from onyx.utils.telemetry import mt_cloud_alias
+from onyx.utils.telemetry import mt_cloud_get_anon_id
+from onyx.utils.telemetry import mt_cloud_identify
 from onyx.utils.telemetry import mt_cloud_telemetry
 from onyx.utils.telemetry import optional_telemetry
 from onyx.utils.telemetry import RecordType
@@ -145,10 +160,21 @@ def is_user_admin(user: User) -> bool:
 
 
 def verify_auth_setting() -> None:
-    if AUTH_TYPE == AuthType.CLOUD:
+    """Log warnings for AUTH_TYPE issues.
+
+    This only runs on app startup not during migrations/scripts.
+    """
+    raw_auth_type = (os.environ.get("AUTH_TYPE") or "").lower()
+
+    if raw_auth_type == "cloud":
         raise ValueError(
-            f"{AUTH_TYPE.value} is not a valid auth type for self-hosted deployments."
+            "'cloud' is not a valid auth type for self-hosted deployments."
         )
+    if raw_auth_type == "disabled":
+        logger.warning(
+            "AUTH_TYPE='disabled' is no longer supported. Using 'basic' instead. Please update your configuration."
+        )
+
     logger.notice(f"Using Auth Type: {AUTH_TYPE.value}")
 
 
@@ -227,18 +253,12 @@ def verify_email_is_invited(email: str) -> None:
     whitelist = get_invited_users()
 
     if not email:
-        raise HTTPException(
-            status_code=status.HTTP_400_BAD_REQUEST,
-            detail={"reason": "Email must be specified"},
-        )
+        raise OnyxError(OnyxErrorCode.INVALID_INPUT, "Email must be specified")
 
     try:
         email_info = validate_email(email, check_deliverability=False)
     except EmailUndeliverableError:
-        raise HTTPException(
-            status_code=status.HTTP_400_BAD_REQUEST,
-            detail={"reason": "Email is not valid"},
-        )
+        raise OnyxError(OnyxErrorCode.INVALID_INPUT, "Email is not valid")
 
     for email_whitelist in whitelist:
         try:
@@ -255,12 +275,9 @@ def verify_email_is_invited(email: str) -> None:
         if email_info.normalized.lower() == email_info_whitelist.normalized.lower():
             return
 
-    raise HTTPException(
-        status_code=status.HTTP_403_FORBIDDEN,
-        detail={
-            "code": REGISTER_INVITE_ONLY_CODE,
-            "reason": "This workspace is invite-only. Please ask your admin to invite you.",
-        },
+    raise OnyxError(
+        OnyxErrorCode.UNAUTHORIZED,
+        "This workspace is invite-only. Please ask your admin to invite you.",
     )
 
 
@@ -270,48 +287,47 @@ def verify_email_in_whitelist(email: str, tenant_id: str) -> None:
             verify_email_is_invited(email)
 
 
-def verify_email_domain(email: str) -> None:
+def verify_email_domain(email: str, *, is_registration: bool = False) -> None:
     if email.count("@") != 1:
-        raise HTTPException(
-            status_code=status.HTTP_400_BAD_REQUEST,
-            detail="Email is not valid",
-        )
+        raise OnyxError(OnyxErrorCode.INVALID_INPUT, "Email is not valid")
 
     local_part, domain = email.split("@")
     domain = domain.lower()
+    local_part = local_part.lower()
 
     if AUTH_TYPE == AuthType.CLOUD:
         # Normalize googlemail.com to gmail.com (they deliver to the same inbox)
         if domain == "googlemail.com":
-            raise HTTPException(
-                status_code=status.HTTP_400_BAD_REQUEST,
-                detail={"reason": "Please use @gmail.com instead of @googlemail.com."},
+            raise OnyxError(
+                OnyxErrorCode.INVALID_INPUT,
+                "Please use @gmail.com instead of @googlemail.com.",
+            )
+
+        # Only block dotted Gmail on new signups — existing users must still be
+        # able to sign in with the address they originally registered with.
+        if is_registration and domain == "gmail.com" and "." in local_part:
+            raise OnyxError(
+                OnyxErrorCode.INVALID_INPUT,
+                "Gmail addresses with '.' are not allowed. Please use your base email address.",
             )
 
         if "+" in local_part and domain != "onyx.app":
-            raise HTTPException(
-                status_code=status.HTTP_400_BAD_REQUEST,
-                detail={
-                    "reason": "Email addresses with '+' are not allowed. Please use your base email address."
-                },
+            raise OnyxError(
+                OnyxErrorCode.INVALID_INPUT,
+                "Email addresses with '+' are not allowed. Please use your base email address.",
             )
 
     # Check if email uses a disposable/temporary domain
     if is_disposable_email(email):
-        raise HTTPException(
-            status_code=status.HTTP_400_BAD_REQUEST,
-            detail={
-                "reason": "Disposable email addresses are not allowed. Please use a permanent email address."
-            },
+        raise OnyxError(
+            OnyxErrorCode.INVALID_INPUT,
+            "Disposable email addresses are not allowed. Please use a permanent email address.",
         )
 
     # Check domain whitelist if configured
     if VALID_EMAIL_DOMAINS:
         if domain not in VALID_EMAIL_DOMAINS:
-            raise HTTPException(
-                status_code=status.HTTP_400_BAD_REQUEST,
-                detail="Email domain is not valid",
-            )
+            raise OnyxError(OnyxErrorCode.INVALID_INPUT, "Email domain is not valid")
 
 
 def enforce_seat_limit(db_session: Session, seats_needed: int = 1) -> None:
@@ -327,7 +343,7 @@ def enforce_seat_limit(db_session: Session, seats_needed: int = 1) -> None:
     )(db_session, seats_needed=seats_needed)
 
     if result is not None and not result.available:
-        raise HTTPException(status_code=402, detail=result.error_message)
+        raise OnyxError(OnyxErrorCode.SEAT_LIMIT_EXCEEDED, result.error_message)
 
 
 class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
@@ -380,10 +396,7 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
                     captcha_token or "", expected_action="signup"
                 )
             except CaptchaVerificationError as e:
-                raise HTTPException(
-                    status_code=status.HTTP_400_BAD_REQUEST,
-                    detail={"reason": str(e)},
-                )
+                raise OnyxError(OnyxErrorCode.INVALID_INPUT, str(e))
 
         # We verify the password here to make sure it's valid before we proceed
         await self.validate_password(
@@ -393,13 +406,10 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
         # Check for disposable emails BEFORE provisioning tenant
         # This prevents creating tenants for throwaway email addresses
         try:
-            verify_email_domain(user_create.email)
-        except HTTPException as e:
+            verify_email_domain(user_create.email, is_registration=True)
+        except OnyxError as e:
             # Log blocked disposable email attempts
-            if (
-                e.status_code == status.HTTP_400_BAD_REQUEST
-                and "Disposable email" in str(e.detail)
-            ):
+            if "Disposable email" in e.detail:
                 domain = (
                     user_create.email.split("@")[-1]
                     if "@" in user_create.email
@@ -543,9 +553,9 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
         result = await db_session.execute(
             select(Persona.id)
             .where(
-                Persona.featured.is_(True),
+                Persona.is_featured.is_(True),
                 Persona.is_public.is_(True),
-                Persona.is_visible.is_(True),
+                Persona.is_listed.is_(True),
                 Persona.deleted.is_(False),
             )
             .order_by(
@@ -589,8 +599,7 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
             char in PASSWORD_SPECIAL_CHARS for char in password
         ):
             raise exceptions.InvalidPasswordException(
-                reason="Password must contain at least one special character from the following set: "
-                f"{PASSWORD_SPECIAL_CHARS}."
+                reason=f"Password must contain at least one special character from the following set: {PASSWORD_SPECIAL_CHARS}."
             )
         return
 
@@ -674,6 +683,8 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
                         raise exceptions.UserNotExists()
 
                 except exceptions.UserNotExists:
+                    verify_email_domain(account_email, is_registration=True)
+
                     # Check seat availability before creating (single-tenant only)
                     with get_session_with_current_tenant() as sync_db:
                         enforce_seat_limit(sync_db)
@@ -771,6 +782,18 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
         except Exception:
             logger.exception("Error deleting anonymous user cookie")
 
+        tenant_id = CURRENT_TENANT_ID_CONTEXTVAR.get()
+
+        # Link the anonymous PostHog session to the identified user so that
+        # pre-login session recordings and events merge into one person profile.
+        if anon_id := mt_cloud_get_anon_id(request):
+            mt_cloud_alias(distinct_id=str(user.id), anonymous_id=anon_id)
+
+        mt_cloud_identify(
+            distinct_id=str(user.id),
+            properties={"email": user.email, "tenant_id": tenant_id},
+        )
+
     async def on_after_register(
         self, user: User, request: Optional[Request] = None
     ) -> None:
@@ -789,12 +812,30 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
             user_count = await get_user_count()
             logger.debug(f"Current tenant user count: {user_count}")
 
+            # Link the anonymous PostHog session to the identified user so
+            # that pre-signup session recordings merge into one person profile.
+            if anon_id := mt_cloud_get_anon_id(request):
+                mt_cloud_alias(distinct_id=str(user.id), anonymous_id=anon_id)
+
+            # Ensure a PostHog person profile exists for this user.
+            mt_cloud_identify(
+                distinct_id=str(user.id),
+                properties={"email": user.email, "tenant_id": tenant_id},
+            )
+
             mt_cloud_telemetry(
                 tenant_id=tenant_id,
-                distinct_id=user.email,
+                distinct_id=str(user.id),
                 event=MilestoneRecordType.USER_SIGNED_UP,
             )
 
+            if user_count == 1:
+                mt_cloud_telemetry(
+                    tenant_id=tenant_id,
+                    distinct_id=str(user.id),
+                    event=MilestoneRecordType.TENANT_CREATED,
+                )
+
         finally:
             CURRENT_TENANT_ID_CONTEXTVAR.reset(token)
 
@@ -804,9 +845,9 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
             attribute="get_marketing_posthog_cookie_name",
             noop_return_value=None,
         )
-        parse_marketing_cookie = fetch_ee_implementation_or_noop(
+        parse_posthog_cookie = fetch_ee_implementation_or_noop(
             module="onyx.utils.posthog_client",
-            attribute="parse_marketing_cookie",
+            attribute="parse_posthog_cookie",
             noop_return_value=None,
         )
         capture_and_sync_with_alternate_posthog = fetch_ee_implementation_or_noop(
@@ -820,7 +861,7 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
             and user_count is not None
             and (marketing_cookie_name := get_marketing_posthog_cookie_name())
             and (marketing_cookie_value := request.cookies.get(marketing_cookie_name))
-            and (parsed_cookie := parse_marketing_cookie(marketing_cookie_value))
+            and (parsed_cookie := parse_posthog_cookie(marketing_cookie_value))
         ):
             marketing_anonymous_id = parsed_cookie["distinct_id"]
 
@@ -857,7 +898,10 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
         )
 
     async def on_after_forgot_password(
-        self, user: User, token: str, request: Optional[Request] = None  # noqa: ARG002
+        self,
+        user: User,
+        token: str,
+        request: Optional[Request] = None,  # noqa: ARG002
     ) -> None:
         if not EMAIL_CONFIGURED:
             logger.error(
@@ -876,7 +920,10 @@ class UserManager(UUIDIDMixin, BaseUserManager[User, uuid.UUID]):
         send_forgot_password_email(user.email, tenant_id=tenant_id, token=token)
 
     async def on_after_request_verify(
-        self, user: User, token: str, request: Optional[Request] = None  # noqa: ARG002
+        self,
+        user: User,
+        token: str,
+        request: Optional[Request] = None,  # noqa: ARG002
     ) -> None:
         verify_email_domain(user.email)
 
@@ -1172,7 +1219,9 @@ class SingleTenantJWTStrategy(JWTStrategy[User, uuid.UUID]):
         return
 
     async def refresh_token(
-        self, token: Optional[str], user: User  # noqa: ARG002
+        self,
+        token: Optional[str],  # noqa: ARG002
+        user: User,  # noqa: ARG002
     ) -> str:
         """Issue a fresh JWT with a new expiry."""
         return await self.write_token(user)
@@ -1200,8 +1249,7 @@ def get_jwt_strategy() -> SingleTenantJWTStrategy:
 if AUTH_BACKEND == AuthBackend.JWT:
     if MULTI_TENANT or AUTH_TYPE == AuthType.CLOUD:
         raise ValueError(
-            "JWT auth backend is only supported for single-tenant, self-hosted deployments. "
-            "Use 'redis' or 'postgres' instead."
+            "JWT auth backend is only supported for single-tenant, self-hosted deployments. Use 'redis' or 'postgres' instead."
         )
     if not USER_AUTH_SECRET:
         raise ValueError("USER_AUTH_SECRET is required for JWT auth backend.")
@@ -1599,6 +1647,125 @@ async def current_admin_user(user: User = Depends(current_user)) -> User:
     return user
 
 
+async def _get_user_from_token_data(token_data: dict) -> User | None:
+    """Shared logic: token data dict → User object.
+
+    Args:
+        token_data: Decoded token data containing 'sub' (user ID).
+
+    Returns:
+        User object if found and active, None otherwise.
+    """
+    user_id = token_data.get("sub")
+    if not user_id:
+        return None
+
+    try:
+        user_uuid = uuid.UUID(user_id)
+    except ValueError:
+        return None
+
+    async with get_async_session_context_manager() as async_db_session:
+        user = await async_db_session.get(User, user_uuid)
+        if user is None or not user.is_active:
+            return None
+        return user
+
+
+_LOOPBACK_HOSTNAMES = frozenset({"localhost", "127.0.0.1", "::1"})
+
+
+def _is_same_origin(actual: str, expected: str) -> bool:
+    """Compare two origins for the WebSocket CSWSH check.
+
+    Scheme and hostname must match exactly.  Port must also match, except
+    when the hostname is a loopback address (localhost / 127.0.0.1 / ::1),
+    where port is ignored.  On loopback, all ports belong to the same
+    operator, so port differences carry no security significance — the
+    CSWSH threat is remote origins, not local ones.
+    """
+    a = urlparse(actual.rstrip("/"))
+    e = urlparse(expected.rstrip("/"))
+
+    if a.scheme != e.scheme or a.hostname != e.hostname:
+        return False
+
+    if a.hostname in _LOOPBACK_HOSTNAMES:
+        return True
+
+    actual_port = a.port or (443 if a.scheme == "https" else 80)
+    expected_port = e.port or (443 if e.scheme == "https" else 80)
+
+    return actual_port == expected_port
+
+
+async def current_user_from_websocket(
+    websocket: WebSocket,
+    token: str = Query(..., description="WebSocket authentication token"),
+) -> User:
+    """
+    WebSocket authentication dependency using query parameter.
+
+    Validates the WS token from query param and returns the User.
+    Raises BasicAuthenticationError if authentication fails.
+
+    The token must be obtained from POST /voice/ws-token before connecting.
+    Tokens are single-use and expire after 60 seconds.
+
+    Usage:
+        1. POST /voice/ws-token -> {"token": "xxx"}
+        2. Connect to ws://host/path?token=xxx
+
+    This applies the same auth checks as current_user() for HTTP endpoints.
+    """
+    # Check Origin header to prevent Cross-Site WebSocket Hijacking (CSWSH).
+    # Browsers always send Origin on WebSocket connections.
+    origin = websocket.headers.get("origin")
+    if not origin:
+        logger.warning("WS auth: missing Origin header")
+        raise BasicAuthenticationError(detail="Access denied. Missing origin.")
+
+    if not _is_same_origin(origin, WEB_DOMAIN):
+        logger.warning(f"WS auth: origin mismatch. Expected {WEB_DOMAIN}, got {origin}")
+        raise BasicAuthenticationError(detail="Access denied. Invalid origin.")
+
+    # Validate WS token in Redis (single-use, deleted after retrieval)
+    try:
+        token_data = await retrieve_ws_token_data(token)
+        if token_data is None:
+            raise BasicAuthenticationError(
+                detail="Access denied. Invalid or expired authentication token."
+            )
+    except BasicAuthenticationError:
+        raise
+    except Exception as e:
+        logger.error(f"WS auth: error during token validation: {e}")
+        raise BasicAuthenticationError(
+            detail="Authentication verification failed."
+        ) from e
+
+    # Get user from token data
+    user = await _get_user_from_token_data(token_data)
+    if user is None:
+        logger.warning(f"WS auth: user not found for id={token_data.get('sub')}")
+        raise BasicAuthenticationError(
+            detail="Access denied. User not found or inactive."
+        )
+
+    # Apply same checks as HTTP auth (verification, OIDC expiry, role)
+    user = await double_check_user(user)
+
+    # Block LIMITED users (same as current_user)
+    if user.role == UserRole.LIMITED:
+        logger.warning(f"WS auth: user {user.email} has LIMITED role")
+        raise BasicAuthenticationError(
+            detail="Access denied. User role is LIMITED. BASIC or higher permissions are required.",
+        )
+
+    logger.debug(f"WS auth: authenticated {user.email}")
+    return user
+
+
 def get_default_admin_user_emails_() -> list[str]:
     # No default seeding available for Onyx MIT
     return []
@@ -1608,6 +1775,7 @@ STATE_TOKEN_AUDIENCE = "fastapi-users:oauth-state"
 STATE_TOKEN_LIFETIME_SECONDS = 3600
 CSRF_TOKEN_KEY = "csrftoken"
 CSRF_TOKEN_COOKIE_NAME = "fastapiusersoauthcsrf"
+PKCE_COOKIE_NAME_PREFIX = "fastapiusersoauthpkce"
 
 
 class OAuth2AuthorizeResponse(BaseModel):
@@ -1628,6 +1796,21 @@ def generate_csrf_token() -> str:
     return secrets.token_urlsafe(32)
 
 
+def _base64url_encode(data: bytes) -> str:
+    return base64.urlsafe_b64encode(data).rstrip(b"=").decode("ascii")
+
+
+def generate_pkce_pair() -> tuple[str, str]:
+    verifier = secrets.token_urlsafe(64)
+    challenge = _base64url_encode(hashlib.sha256(verifier.encode("ascii")).digest())
+    return verifier, challenge
+
+
+def get_pkce_cookie_name(state: str) -> str:
+    state_hash = hashlib.sha256(state.encode("utf-8")).hexdigest()
+    return f"{PKCE_COOKIE_NAME_PREFIX}_{state_hash}"
+
+
 # refer to https://github.com/fastapi-users/fastapi-users/blob/42ddc241b965475390e2bce887b084152ae1a2cd/fastapi_users/fastapi_users.py#L91
 def create_onyx_oauth_router(
     oauth_client: BaseOAuth2,
@@ -1636,6 +1819,7 @@ def create_onyx_oauth_router(
     redirect_url: Optional[str] = None,
     associate_by_email: bool = False,
     is_verified_by_default: bool = False,
+    enable_pkce: bool = False,
 ) -> APIRouter:
     return get_oauth_router(
         oauth_client,
@@ -1645,6 +1829,7 @@ def create_onyx_oauth_router(
         redirect_url,
         associate_by_email,
         is_verified_by_default,
+        enable_pkce=enable_pkce,
     )
 
 
@@ -1663,6 +1848,7 @@ def get_oauth_router(
     csrf_token_cookie_secure: Optional[bool] = None,
     csrf_token_cookie_httponly: bool = True,
     csrf_token_cookie_samesite: Optional[Literal["lax", "strict", "none"]] = "lax",
+    enable_pkce: bool = False,
 ) -> APIRouter:
     """Generate a router with the OAuth routes."""
     router = APIRouter()
@@ -1679,6 +1865,13 @@ def get_oauth_router(
             route_name=callback_route_name,
         )
 
+    async def null_access_token_state() -> tuple[OAuth2Token, Optional[str]] | None:
+        return None
+
+    access_token_state_dependency = (
+        oauth2_authorize_callback if not enable_pkce else null_access_token_state
+    )
+
     if csrf_token_cookie_secure is None:
         csrf_token_cookie_secure = WEB_DOMAIN.startswith("https")
 
@@ -1712,13 +1905,26 @@ def get_oauth_router(
             CSRF_TOKEN_KEY: csrf_token,
         }
         state = generate_state_token(state_data, state_secret)
+        pkce_cookie: tuple[str, str] | None = None
 
-        # Get the basic authorization URL
-        authorization_url = await oauth_client.get_authorization_url(
-            authorize_redirect_url,
-            state,
-            scopes,
-        )
+        if enable_pkce:
+            code_verifier, code_challenge = generate_pkce_pair()
+            pkce_cookie_name = get_pkce_cookie_name(state)
+            pkce_cookie = (pkce_cookie_name, code_verifier)
+            authorization_url = await oauth_client.get_authorization_url(
+                authorize_redirect_url,
+                state,
+                scopes,
+                code_challenge=code_challenge,
+                code_challenge_method="S256",
+            )
+        else:
+            # Get the basic authorization URL
+            authorization_url = await oauth_client.get_authorization_url(
+                authorize_redirect_url,
+                state,
+                scopes,
+            )
 
         # For Google OAuth, add parameters to request refresh tokens
         if oauth_client.name == "google":
@@ -1726,11 +1932,15 @@ def get_oauth_router(
                 authorization_url, {"access_type": "offline", "prompt": "consent"}
             )
 
-        if redirect:
-            redirect_response = RedirectResponse(authorization_url, status_code=302)
-            redirect_response.set_cookie(
-                key=csrf_token_cookie_name,
-                value=csrf_token,
+        def set_oauth_cookie(
+            target_response: Response,
+            *,
+            key: str,
+            value: str,
+        ) -> None:
+            target_response.set_cookie(
+                key=key,
+                value=value,
                 max_age=STATE_TOKEN_LIFETIME_SECONDS,
                 path=csrf_token_cookie_path,
                 domain=csrf_token_cookie_domain,
@@ -1738,18 +1948,28 @@ def get_oauth_router(
                 httponly=csrf_token_cookie_httponly,
                 samesite=csrf_token_cookie_samesite,
             )
-            return redirect_response
 
-        response.set_cookie(
+        response_with_cookies: Response
+        if redirect:
+            response_with_cookies = RedirectResponse(authorization_url, status_code=302)
+        else:
+            response_with_cookies = response
+
+        set_oauth_cookie(
+            response_with_cookies,
             key=csrf_token_cookie_name,
             value=csrf_token,
-            max_age=STATE_TOKEN_LIFETIME_SECONDS,
-            path=csrf_token_cookie_path,
-            domain=csrf_token_cookie_domain,
-            secure=csrf_token_cookie_secure,
-            httponly=csrf_token_cookie_httponly,
-            samesite=csrf_token_cookie_samesite,
         )
+        if pkce_cookie is not None:
+            pkce_cookie_name, code_verifier = pkce_cookie
+            set_oauth_cookie(
+                response_with_cookies,
+                key=pkce_cookie_name,
+                value=code_verifier,
+            )
+
+        if redirect:
+            return response_with_cookies
 
         return OAuth2AuthorizeResponse(authorization_url=authorization_url)
 
@@ -1780,119 +2000,242 @@ def get_oauth_router(
     )
     async def callback(
         request: Request,
-        access_token_state: Tuple[OAuth2Token, str] = Depends(
-            oauth2_authorize_callback
+        access_token_state: Tuple[OAuth2Token, Optional[str]] | None = Depends(
+            access_token_state_dependency
         ),
+        code: Optional[str] = None,
+        state: Optional[str] = None,
+        error: Optional[str] = None,
         user_manager: BaseUserManager[models.UP, models.ID] = Depends(get_user_manager),
         strategy: Strategy[models.UP, models.ID] = Depends(backend.get_strategy),
-    ) -> RedirectResponse:
-        token, state = access_token_state
-        account_id, account_email = await oauth_client.get_id_email(
-            token["access_token"]
-        )
+    ) -> Response:
+        pkce_cookie_name: str | None = None
 
-        if account_email is None:
-            raise HTTPException(
-                status_code=status.HTTP_400_BAD_REQUEST,
-                detail=ErrorCode.OAUTH_NOT_AVAILABLE_EMAIL,
-            )
+        def delete_pkce_cookie(response: Response) -> None:
+            if enable_pkce and pkce_cookie_name:
+                response.delete_cookie(
+                    key=pkce_cookie_name,
+                    path=csrf_token_cookie_path,
+                    domain=csrf_token_cookie_domain,
+                    secure=csrf_token_cookie_secure,
+                    httponly=csrf_token_cookie_httponly,
+                    samesite=csrf_token_cookie_samesite,
+                )
 
-        try:
-            state_data = decode_jwt(state, state_secret, [STATE_TOKEN_AUDIENCE])
-        except jwt.DecodeError:
-            raise HTTPException(
-                status_code=status.HTTP_400_BAD_REQUEST,
-                detail=getattr(
-                    ErrorCode, "ACCESS_TOKEN_DECODE_ERROR", "ACCESS_TOKEN_DECODE_ERROR"
-                ),
-            )
-        except jwt.ExpiredSignatureError:
-            raise HTTPException(
-                status_code=status.HTTP_400_BAD_REQUEST,
-                detail=getattr(
-                    ErrorCode,
-                    "ACCESS_TOKEN_ALREADY_EXPIRED",
-                    "ACCESS_TOKEN_ALREADY_EXPIRED",
-                ),
-            )
+        def build_error_response(exc: OnyxError) -> JSONResponse:
+            log_onyx_error(exc)
+            error_response = onyx_error_to_json_response(exc)
+            delete_pkce_cookie(error_response)
+            return error_response
 
-        cookie_csrf_token = request.cookies.get(csrf_token_cookie_name)
-        state_csrf_token = state_data.get(CSRF_TOKEN_KEY)
-        if (
-            not cookie_csrf_token
-            or not state_csrf_token
-            or not secrets.compare_digest(cookie_csrf_token, state_csrf_token)
-        ):
-            raise HTTPException(
-                status_code=status.HTTP_400_BAD_REQUEST,
-                detail=getattr(ErrorCode, "OAUTH_INVALID_STATE", "OAUTH_INVALID_STATE"),
-            )
+        def decode_and_validate_state(state_value: str) -> Dict[str, str]:
+            try:
+                state_data = decode_jwt(
+                    state_value, state_secret, [STATE_TOKEN_AUDIENCE]
+                )
+            except jwt.DecodeError:
+                raise OnyxError(
+                    OnyxErrorCode.VALIDATION_ERROR,
+                    getattr(
+                        ErrorCode,
+                        "ACCESS_TOKEN_DECODE_ERROR",
+                        "ACCESS_TOKEN_DECODE_ERROR",
+                    ),
+                )
+            except jwt.ExpiredSignatureError:
+                raise OnyxError(
+                    OnyxErrorCode.VALIDATION_ERROR,
+                    getattr(
+                        ErrorCode,
+                        "ACCESS_TOKEN_ALREADY_EXPIRED",
+                        "ACCESS_TOKEN_ALREADY_EXPIRED",
+                    ),
+                )
+            except jwt.PyJWTError:
+                raise OnyxError(
+                    OnyxErrorCode.VALIDATION_ERROR,
+                    getattr(
+                        ErrorCode,
+                        "ACCESS_TOKEN_DECODE_ERROR",
+                        "ACCESS_TOKEN_DECODE_ERROR",
+                    ),
+                )
 
-        next_url = state_data.get("next_url", "/")
-        referral_source = state_data.get("referral_source", None)
-        try:
-            tenant_id = fetch_ee_implementation_or_noop(
-                "onyx.server.tenants.user_mapping", "get_tenant_id_for_email", None
-            )(account_email)
-        except exceptions.UserNotExists:
-            tenant_id = None
+            cookie_csrf_token = request.cookies.get(csrf_token_cookie_name)
+            state_csrf_token = state_data.get(CSRF_TOKEN_KEY)
+            if (
+                not cookie_csrf_token
+                or not state_csrf_token
+                or not secrets.compare_digest(cookie_csrf_token, state_csrf_token)
+            ):
+                raise OnyxError(
+                    OnyxErrorCode.VALIDATION_ERROR,
+                    getattr(ErrorCode, "OAUTH_INVALID_STATE", "OAUTH_INVALID_STATE"),
+                )
 
-        request.state.referral_source = referral_source
+            return state_data
 
-        # Proceed to authenticate or create the user
-        try:
-            user = await user_manager.oauth_callback(
-                oauth_client.name,
-                token["access_token"],
-                account_id,
-                account_email,
-                token.get("expires_at"),
-                token.get("refresh_token"),
-                request,
-                associate_by_email=associate_by_email,
-                is_verified_by_default=is_verified_by_default,
-            )
-        except UserAlreadyExists:
-            raise HTTPException(
-                status_code=status.HTTP_400_BAD_REQUEST,
-                detail=ErrorCode.OAUTH_USER_ALREADY_EXISTS,
-            )
+        token: OAuth2Token
+        state_data: Dict[str, str]
 
-        if not user.is_active:
-            raise HTTPException(
-                status_code=status.HTTP_400_BAD_REQUEST,
-                detail=ErrorCode.LOGIN_BAD_CREDENTIALS,
-            )
+        # `code`, `state`, and `error` are read directly only in the PKCE path.
+        # In the non-PKCE path, `oauth2_authorize_callback` consumes them.
+        if enable_pkce:
+            if state is not None:
+                pkce_cookie_name = get_pkce_cookie_name(state)
 
-        # Login user
-        response = await backend.login(strategy, user)
-        await user_manager.on_after_login(user, request, response)
+            if error is not None:
+                return build_error_response(
+                    OnyxError(
+                        OnyxErrorCode.VALIDATION_ERROR,
+                        "Authorization request failed or was denied",
+                    )
+                )
+            if code is None:
+                return build_error_response(
+                    OnyxError(
+                        OnyxErrorCode.VALIDATION_ERROR,
+                        "Missing authorization code in OAuth callback",
+                    )
+                )
+            if state is None:
+                return build_error_response(
+                    OnyxError(
+                        OnyxErrorCode.VALIDATION_ERROR,
+                        "Missing state parameter in OAuth callback",
+                    )
+                )
 
-        # Prepare redirect response
-        if tenant_id is None:
-            # Use URL utility to add parameters
-            redirect_url = add_url_params(next_url, {"new_team": "true"})
-            redirect_response = RedirectResponse(redirect_url, status_code=302)
-        else:
-            # No parameters to add
-            redirect_response = RedirectResponse(next_url, status_code=302)
+            state_value = state
 
-        # Copy headers from auth response to redirect response, with special handling for Set-Cookie
-        for header_name, header_value in response.headers.items():
-            # FastAPI can have multiple Set-Cookie headers as a list
-            if header_name.lower() == "set-cookie" and isinstance(header_value, list):
-                for cookie_value in header_value:
-                    redirect_response.headers.append(header_name, cookie_value)
+            if redirect_url is not None:
+                callback_redirect_url = redirect_url
             else:
+                callback_path = request.app.url_path_for(callback_route_name)
+                callback_redirect_url = f"{WEB_DOMAIN}{callback_path}"
+
+            code_verifier = request.cookies.get(cast(str, pkce_cookie_name))
+            if not code_verifier:
+                return build_error_response(
+                    OnyxError(
+                        OnyxErrorCode.VALIDATION_ERROR,
+                        "Missing PKCE verifier cookie in OAuth callback",
+                    )
+                )
+
+            try:
+                state_data = decode_and_validate_state(state_value)
+            except OnyxError as e:
+                return build_error_response(e)
+
+            try:
+                token = await oauth_client.get_access_token(
+                    code, callback_redirect_url, code_verifier
+                )
+            except GetAccessTokenError:
+                return build_error_response(
+                    OnyxError(
+                        OnyxErrorCode.VALIDATION_ERROR,
+                        "Authorization code exchange failed",
+                    )
+                )
+        else:
+            if access_token_state is None:
+                raise OnyxError(
+                    OnyxErrorCode.INTERNAL_ERROR, "Missing OAuth callback state"
+                )
+            token, callback_state = access_token_state
+            if callback_state is None:
+                raise OnyxError(
+                    OnyxErrorCode.VALIDATION_ERROR,
+                    "Missing state parameter in OAuth callback",
+                )
+            state_data = decode_and_validate_state(callback_state)
+
+        async def complete_login_flow(
+            token: OAuth2Token, state_data: Dict[str, str]
+        ) -> RedirectResponse:
+            account_id, account_email = await oauth_client.get_id_email(
+                token["access_token"]
+            )
+
+            if account_email is None:
+                raise OnyxError(
+                    OnyxErrorCode.VALIDATION_ERROR,
+                    ErrorCode.OAUTH_NOT_AVAILABLE_EMAIL,
+                )
+
+            next_url = state_data.get("next_url", "/")
+            referral_source = state_data.get("referral_source", None)
+            try:
+                tenant_id = fetch_ee_implementation_or_noop(
+                    "onyx.server.tenants.user_mapping", "get_tenant_id_for_email", None
+                )(account_email)
+            except exceptions.UserNotExists:
+                tenant_id = None
+
+            request.state.referral_source = referral_source
+
+            # Proceed to authenticate or create the user
+            try:
+                user = await user_manager.oauth_callback(
+                    oauth_client.name,
+                    token["access_token"],
+                    account_id,
+                    account_email,
+                    token.get("expires_at"),
+                    token.get("refresh_token"),
+                    request,
+                    associate_by_email=associate_by_email,
+                    is_verified_by_default=is_verified_by_default,
+                )
+            except UserAlreadyExists:
+                raise OnyxError(
+                    OnyxErrorCode.VALIDATION_ERROR,
+                    ErrorCode.OAUTH_USER_ALREADY_EXISTS,
+                )
+
+            if not user.is_active:
+                raise OnyxError(
+                    OnyxErrorCode.VALIDATION_ERROR,
+                    ErrorCode.LOGIN_BAD_CREDENTIALS,
+                )
+
+            # Login user
+            response = await backend.login(strategy, user)
+            await user_manager.on_after_login(user, request, response)
+
+            # Prepare redirect response
+            if tenant_id is None:
+                # Use URL utility to add parameters
+                redirect_destination = add_url_params(next_url, {"new_team": "true"})
+                redirect_response = RedirectResponse(
+                    redirect_destination, status_code=302
+                )
+            else:
+                # No parameters to add
+                redirect_response = RedirectResponse(next_url, status_code=302)
+
+            # Copy headers from auth response to redirect response, with special handling for Set-Cookie
+            for header_name, header_value in response.headers.items():
+                header_name_lower = header_name.lower()
+                if header_name_lower == "set-cookie":
+                    redirect_response.headers.append(header_name, header_value)
+                    continue
+                if header_name_lower in {"location", "content-length"}:
+                    continue
                 redirect_response.headers[header_name] = header_value
 
-        if hasattr(response, "body"):
-            redirect_response.body = response.body
-        if hasattr(response, "status_code"):
-            redirect_response.status_code = response.status_code
-        if hasattr(response, "media_type"):
-            redirect_response.media_type = response.media_type
+            return redirect_response
 
-        return redirect_response
+        if enable_pkce:
+            try:
+                redirect_response = await complete_login_flow(token, state_data)
+            except OnyxError as e:
+                return build_error_response(e)
+            delete_pkce_cookie(redirect_response)
+            return redirect_response
+
+        return await complete_login_flow(token, state_data)
 
     return router

backend/onyx/background/celery/apps/app_base.py

@@ -154,8 +154,7 @@ def on_task_postrun(
         tenant_id = cast(str, kwargs.get("tenant_id", POSTGRES_DEFAULT_SCHEMA))
 
     task_logger.debug(
-        f"Task {task.name} (ID: {task_id}) completed with state: {state} "
-        f"{f'for tenant_id={tenant_id}' if tenant_id else ''}"
+        f"Task {task.name} (ID: {task_id}) completed with state: {state} {f'for tenant_id={tenant_id}' if tenant_id else ''}"
     )
 
     r = get_redis_client(tenant_id=tenant_id)
@@ -211,7 +210,9 @@ def on_task_postrun(
 
 
 def on_celeryd_init(
-    sender: str, conf: Any = None, **kwargs: Any  # noqa: ARG001
+    sender: str,  # noqa: ARG001
+    conf: Any = None,  # noqa: ARG001
+    **kwargs: Any,  # noqa: ARG001
 ) -> None:
     """The first signal sent on celery worker startup"""
 
@@ -277,10 +278,7 @@ def wait_for_redis(sender: Any, **kwargs: Any) -> None:  # noqa: ARG001
         time.sleep(WAIT_INTERVAL)
 
     if not ready:
-        msg = (
-            f"Redis: Readiness probe did not succeed within the timeout "
-            f"({WAIT_LIMIT} seconds). Exiting..."
-        )
+        msg = f"Redis: Readiness probe did not succeed within the timeout ({WAIT_LIMIT} seconds). Exiting..."
         logger.error(msg)
         raise WorkerShutdown(msg)
 
@@ -319,10 +317,7 @@ def wait_for_db(sender: Any, **kwargs: Any) -> None:  # noqa: ARG001
         time.sleep(WAIT_INTERVAL)
 
     if not ready:
-        msg = (
-            f"Database: Readiness probe did not succeed within the timeout "
-            f"({WAIT_LIMIT} seconds). Exiting..."
-        )
+        msg = f"Database: Readiness probe did not succeed within the timeout ({WAIT_LIMIT} seconds). Exiting..."
         logger.error(msg)
         raise WorkerShutdown(msg)
 
@@ -349,10 +344,7 @@ def on_secondary_worker_init(sender: Any, **kwargs: Any) -> None:  # noqa: ARG00
             f"Primary worker is not ready yet. elapsed={time_elapsed:.1f} timeout={WAIT_LIMIT:.1f}"
         )
         if time_elapsed > WAIT_LIMIT:
-            msg = (
-                f"Primary worker was not ready within the timeout. "
-                f"({WAIT_LIMIT} seconds). Exiting..."
-            )
+            msg = f"Primary worker was not ready within the timeout. ({WAIT_LIMIT} seconds). Exiting..."
             logger.error(msg)
             raise WorkerShutdown(msg)
 
@@ -522,7 +514,9 @@ def reset_tenant_id(
     CURRENT_TENANT_ID_CONTEXTVAR.set(POSTGRES_DEFAULT_SCHEMA)
 
 
-def wait_for_vespa_or_shutdown(sender: Any, **kwargs: Any) -> None:  # noqa: ARG001
+def wait_for_vespa_or_shutdown(
+    sender: Any, **kwargs: Any  # noqa: ARG001
+) -> None:  # noqa: ARG001
     """Waits for Vespa to become ready subject to a timeout.
     Raises WorkerShutdown if the timeout is reached."""