Merge pull request 'fix(ext-ci): web-ext v8 .cjs config workaround' (#17) from dev into main

_apply_sidecar is additive: it find-or-creates Post/Source/ImageProvenance
and sets primary_post_id NULL-only, so any IR-migration provenance on the
existing row survives untouched and the new GS sidecar adds a second
ImageProvenance pointing at the freshly-created Post.

Wrapped in try/except so a malformed sidecar can't unwind the file-swap
commit — the file replacement is the critical operation.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

.forgejo/workflows/build.yml

@@ -6,18 +6,150 @@ on:
 
 # Requires repo secret RELEASE_TOKEN — a Forgejo PAT with scopes:
 #   - write:package, read:package  (for docker push to git.fabledsword.com)
-#   - write:release                (for future release-cutting workflows)
+#   - write:release                (for ext-<version> release asset cache)
 #   - write:issue                  (for future issue-management automation)
 # The injected GITHUB_TOKEN cannot be used — it lacks write:package.
 
 jobs:
-  build-web:
+  # Sign-or-fetch-from-cache: signs the extension via AMO if no ext-<version>
+  # Forgejo release exists yet, otherwise downloads the cached signed XPI.
+  # Result is uploaded as an Actions artifact for build-web to consume.
+  #
+  # Why this lives in build.yml (not a separate workflow): the merge-commit's
+  # docker image tagged `:latest` MUST carry the XPI. A separate sign workflow
+  # racing build.yml leaves `:latest` without the XPI for ~5min (until the
+  # commit-back triggers another build). Inline ordering eliminates the race.
+  # Cache strategy: Forgejo Release Assets — picked 2026-05-25 over Generic
+  # Packages (cleaner API surface) and commit-back-to-side-branch (no extra
+  # branch to manage). AMO blocks re-signing the same version (returns 409),
+  # so signing is intentionally one-shot per version bump.
+  sign-extension:
+    if: github.ref == 'refs/heads/main'
     runs-on: python-ci
     container:
       image: git.fabledsword.com/bvandeusen/ci-python:3.14
     steps:
       - uses: actions/checkout@v4
 
+      - name: Resolve extension version
+        id: extver
+        run: |
+          VERSION=$(grep -E '"version"' extension/package.json | head -1 | sed -E 's/.*"version"[[:space:]]*:[[:space:]]*"([^"]+)".*/\1/')
+          echo "version=$VERSION" >> "$GITHUB_OUTPUT"
+          echo "Resolved extension version: $VERSION"
+
+      - name: Check Forgejo release-asset cache
+        id: cache
+        env:
+          TOKEN: ${{ secrets.RELEASE_TOKEN }}
+        run: |
+          set -eu
+          VERSION=${{ steps.extver.outputs.version }}
+          STATUS=$(curl -s -o release.json -w "%{http_code}" \
+            -H "Authorization: token $TOKEN" \
+            "https://git.fabledsword.com/api/v1/repos/bvandeusen/FabledCurator/releases/tags/ext-$VERSION" || echo 000)
+          echo "Tag lookup HTTP status: $STATUS"
+          if [ "$STATUS" = "200" ]; then
+            ASSET_ID=$(jq -r '.assets[]? | select(.name | test("\\.xpi$")) | .id' release.json | head -1)
+            if [ -n "$ASSET_ID" ] && [ "$ASSET_ID" != "null" ]; then
+              echo "cached=true" >> "$GITHUB_OUTPUT"
+              echo "asset_id=$ASSET_ID" >> "$GITHUB_OUTPUT"
+              echo "Cached XPI exists at ext-$VERSION (asset id $ASSET_ID); skipping AMO sign"
+            else
+              echo "cached=false" >> "$GITHUB_OUTPUT"
+              echo "Release ext-$VERSION exists but has no .xpi asset; will re-sign + re-upload"
+            fi
+          else
+            echo "cached=false" >> "$GITHUB_OUTPUT"
+            echo "No release named ext-$VERSION; will sign via AMO and upload"
+          fi
+
+      - name: Download cached signed XPI (cache hit)
+        if: steps.cache.outputs.cached == 'true'
+        env:
+          TOKEN: ${{ secrets.RELEASE_TOKEN }}
+        run: |
+          set -eu
+          mkdir -p extension/web-ext-artifacts
+          curl -sL -H "Authorization: token $TOKEN" \
+            -o "extension/web-ext-artifacts/fabledcurator-${{ steps.extver.outputs.version }}.xpi" \
+            "https://git.fabledsword.com/api/v1/repos/bvandeusen/FabledCurator/releases/assets/${{ steps.cache.outputs.asset_id }}"
+          ls -la extension/web-ext-artifacts/
+
+      - name: Sign via AMO (cache miss)
+        if: steps.cache.outputs.cached != 'true'
+        run: |
+          cd extension && npm install --no-save --no-audit --no-fund && npm run sign
+        env:
+          WEB_EXT_API_KEY: ${{ secrets.MOZILLA_AMO_JWT_KEY }}
+          WEB_EXT_API_SECRET: ${{ secrets.MOZILLA_AMO_JWT_SECRET }}
+
+      - name: Upload signed XPI to ext-<version> release (cache miss)
+        if: steps.cache.outputs.cached != 'true'
+        env:
+          TOKEN: ${{ secrets.RELEASE_TOKEN }}
+        run: |
+          set -eux
+          VERSION=${{ steps.extver.outputs.version }}
+          # AMO renames signed XPIs with its internal addon-id-safe-string;
+          # canonicalize to fabledcurator-<version>.xpi so the FC server's
+          # whitelist (backend/app/frontend.py expects 'fabledcurator-*.xpi')
+          # keeps working.
+          SIGNED=$(ls extension/web-ext-artifacts/*.xpi | head -1)
+          XPI="extension/web-ext-artifacts/fabledcurator-$VERSION.xpi"
+          cp "$SIGNED" "$XPI"
+          # Find-or-create the ext-<version> release.
+          STATUS=$(curl -s -o release.json -w "%{http_code}" \
+            -H "Authorization: token $TOKEN" \
+            "https://git.fabledsword.com/api/v1/repos/bvandeusen/FabledCurator/releases/tags/ext-$VERSION" || echo 000)
+          if [ "$STATUS" != "200" ]; then
+            curl -s -X POST -H "Authorization: token $TOKEN" -H "Content-Type: application/json" \
+              -d "{\"tag_name\":\"ext-$VERSION\",\"name\":\"Extension $VERSION (signed XPI cache)\",\"body\":\"Internal cache for the signed XPI consumed by build.yml's build-web job. Not a user-facing FC release.\",\"target_commitish\":\"main\"}" \
+              -o release.json \
+              "https://git.fabledsword.com/api/v1/repos/bvandeusen/FabledCurator/releases"
+          fi
+          RELEASE_ID=$(jq -r '.id' release.json)
+          test -n "$RELEASE_ID" && test "$RELEASE_ID" != "null"
+          curl -s -X POST -H "Authorization: token $TOKEN" \
+            -F "attachment=@$XPI" \
+            "https://git.fabledsword.com/api/v1/repos/bvandeusen/FabledCurator/releases/$RELEASE_ID/assets?name=fabledcurator-$VERSION.xpi"
+          echo "Uploaded fabledcurator-$VERSION.xpi to ext-$VERSION release"
+
+      - name: Upload XPI as Actions artifact (handoff to build-web)
+        uses: actions/upload-artifact@v4
+        with:
+          name: signed-extension
+          path: extension/web-ext-artifacts/fabledcurator-*.xpi
+          retention-days: 1
+
+  build-web:
+    needs: [sign-extension]
+    # sign-extension is main-only; on dev it's skipped, build-web still runs.
+    if: always() && (needs.sign-extension.result == 'success' || needs.sign-extension.result == 'skipped')
+    runs-on: python-ci
+    container:
+      image: git.fabledsword.com/bvandeusen/ci-python:3.14
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Download signed extension (main only)
+        if: github.ref == 'refs/heads/main'
+        uses: actions/download-artifact@v4
+        with:
+          name: signed-extension
+          path: extension/web-ext-artifacts/
+
+      - name: Place signed XPI in build context (main only)
+        if: github.ref == 'refs/heads/main'
+        run: |
+          set -eux
+          VERSION=$(grep -E '"version"' extension/package.json | head -1 | sed -E 's/.*"version"[[:space:]]*:[[:space:]]*"([^"]+)".*/\1/')
+          XPI=$(ls extension/web-ext-artifacts/*.xpi | head -1)
+          mkdir -p frontend/public/extension
+          cp "$XPI" "frontend/public/extension/fabledcurator-$VERSION.xpi"
+          cp "$XPI" "frontend/public/extension/fabledcurator-latest.xpi"
+          ls -la frontend/public/extension/
+
       - name: Determine tag
         id: tag
         run: |

.forgejo/workflows/ci.yml

@@ -24,11 +24,26 @@ jobs:
     steps:
       - uses: actions/checkout@v4
 
+      - name: Cache pip wheels
+        uses: actions/cache@v4
+        with:
+          path: ~/.cache/pip
+          key: pip-${{ runner.os }}-py314-${{ hashFiles('requirements.txt') }}
+          restore-keys: |
+            pip-${{ runner.os }}-py314-
+
       - name: Install Python deps
         # ruff is pre-installed in the ci-python image (see CI-Runner/CI-python/
         # Dockerfile's RUFF_VERSION). Per FabledRulebook ci-runners.md, toolchain
         # versions live on the runner image, not here.
-        run: pip install -r requirements.txt pytest pytest-asyncio
+        # uv: 5-10x faster wheel resolve than pip for cold caches.
+        # Falls back to pip install on uv-missing runners (older images).
+        run: |
+          if command -v uv >/dev/null 2>&1; then
+            uv pip install --system -r requirements.txt pytest pytest-asyncio
+          else
+            pip install -r requirements.txt pytest pytest-asyncio
+          fi
 
       - name: Ruff lint
         run: ruff check backend/ tests/ alembic/
@@ -57,17 +72,28 @@ jobs:
       - run: npm run test:unit
       - run: npm run build
 
-  integration:
-    # This act_runner (swarm-runner v0.6.1) puts service containers on the
-    # default bridge with NO service-name DNS, and publishing fixed host
-    # ports collides with the operator's running docker-compose dev stack on
-    # the same shared daemon. Workaround: publish NO host ports, and reach
-    # each service by its bridge IP — discovered at runtime via the mounted
-    # docker socket (the ci-python image ships /usr/bin/docker). Default-bridge
-    # containers can talk by IP (only embedded DNS is missing), so IP
-    # addressing is reliable here. Everything runs in ONE step so resolved
-    # values don't depend on cross-step env passing. Pattern documented in
-    # FabledRulebook/forgejo.md "CI philosophy".
+  # Integration suite split into THREE parallel shards (2026-05-25, runner
+  # capacity bumped 2→6). Each shard gets its own Postgres + Redis service
+  # set and runs alembic + a disjoint subset of integration tests. Shards
+  # share no DB state, so the autouse TRUNCATE fixture in tests/conftest.py
+  # stays single-threaded per shard but multiple shards run in parallel
+  # wall-clock. Approximate split — rebalance once --durations=15 output
+  # reveals which shard is the long pole.
+  #
+  # Each shard's docker-ps filter uses its own unique job name to scope
+  # service-container resolution. act_runner appears to strip underscores
+  # from job names when building container labels — `int_api` yielded
+  # zero matches on 2026-05-25 — so shards use no-separator names
+  # (`intapi`, `intimp`, `intcore`) instead. Each step prints
+  # `docker ps -a` first so a future naming-convention shift surfaces in
+  # the log without another guess-and-push cycle.
+  #
+  # Pre-baking requirements.txt into ci-python:3.14 is intentionally NOT
+  # done — per ci-requirements.md, FC is the only Python consumer of that
+  # image and the CI-Runner project's "add deps to image when used by >1
+  # project" rule keeps the install per-job.
+
+  intapi:
     runs-on: python-ci
     container:
       image: git.fabledsword.com/bvandeusen/ci-python:3.14
@@ -98,15 +124,21 @@ jobs:
           --health-retries 10
     steps:
       - uses: actions/checkout@v4
-
-      - name: Integration suite (resolve service IPs, migrate, test)
+      - name: Cache pip wheels
+        uses: actions/cache@v4
+        with:
+          path: ~/.cache/pip
+          key: pip-${{ runner.os }}-py314-${{ hashFiles('requirements.txt') }}
+          restore-keys: |
+            pip-${{ runner.os }}-py314-
+      - name: API integration shard (resolve service IPs, migrate, test)
         run: |
           set -eux
-          # Scope to THIS job's service containers (act_runner names them
-          # ...JOB-integration...); the operator's compose stack uses the
-          # same images but different names, so it won't match.
-          PG=$(docker ps --filter "name=JOB-integration" --filter "ancestor=pgvector/pgvector:pg16" -q | head -n1)
-          RD=$(docker ps --filter "name=JOB-integration" --filter "ancestor=redis:7-alpine" -q | head -n1)
+          echo "=== container landscape (diagnostic for filter scoping) ==="
+          docker ps -a --format '{{.ID}} {{.Image}} -> {{.Names}}'
+          echo "=== end landscape ==="
+          PG=$(docker ps --filter "name=intapi" --filter "ancestor=pgvector/pgvector:pg16" -q | head -n1)
+          RD=$(docker ps --filter "name=intapi" --filter "ancestor=redis:7-alpine" -q | head -n1)
           test -n "$PG" && test -n "$RD"
           PG_IP=$(docker inspect -f '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' "$PG")
           RD_IP=$(docker inspect -f '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' "$RD")
@@ -114,11 +146,153 @@ jobs:
           export DB_HOST="$PG_IP"
           export CELERY_BROKER_URL="redis://$RD_IP:6379/0"
           export CELERY_RESULT_BACKEND="redis://$RD_IP:6379/0"
-          # Wait for Postgres to accept TCP (bash /dev/tcp; no extra tools).
           for i in $(seq 1 60); do
             (echo > "/dev/tcp/$PG_IP/5432") >/dev/null 2>&1 && break
             sleep 2
           done
-          pip install -r requirements.txt pytest pytest-asyncio
+          if command -v uv >/dev/null 2>&1; then
+            uv pip install --system -r requirements.txt pytest pytest-asyncio
+          else
+            pip install -r requirements.txt pytest pytest-asyncio
+          fi
           alembic upgrade head
-          pytest tests/ -v -m integration
+          pytest tests/test_api_*.py -v -m integration --durations=15
+
+  intimp:
+    runs-on: python-ci
+    container:
+      image: git.fabledsword.com/bvandeusen/ci-python:3.14
+    env:
+      DB_USER: fabledcurator
+      DB_PASSWORD: ci_integration
+      DB_PORT: "5432"
+      DB_NAME: fabledcurator_test
+      SECRET_KEY: ci_integration_placeholder
+    services:
+      postgres:
+        image: pgvector/pgvector:pg16
+        env:
+          POSTGRES_USER: fabledcurator
+          POSTGRES_PASSWORD: ci_integration
+          POSTGRES_DB: fabledcurator_test
+        options: >-
+          --health-cmd "pg_isready -U fabledcurator"
+          --health-interval 10s
+          --health-timeout 5s
+          --health-retries 10
+      redis:
+        image: redis:7-alpine
+        options: >-
+          --health-cmd "redis-cli ping"
+          --health-interval 10s
+          --health-timeout 5s
+          --health-retries 10
+    steps:
+      - uses: actions/checkout@v4
+      - name: Cache pip wheels
+        uses: actions/cache@v4
+        with:
+          path: ~/.cache/pip
+          key: pip-${{ runner.os }}-py314-${{ hashFiles('requirements.txt') }}
+          restore-keys: |
+            pip-${{ runner.os }}-py314-
+      - name: Importer integration shard (resolve service IPs, migrate, test)
+        run: |
+          set -eux
+          echo "=== container landscape (diagnostic for filter scoping) ==="
+          docker ps -a --format '{{.ID}} {{.Image}} -> {{.Names}}'
+          echo "=== end landscape ==="
+          PG=$(docker ps --filter "name=intimp" --filter "ancestor=pgvector/pgvector:pg16" -q | head -n1)
+          RD=$(docker ps --filter "name=intimp" --filter "ancestor=redis:7-alpine" -q | head -n1)
+          test -n "$PG" && test -n "$RD"
+          PG_IP=$(docker inspect -f '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' "$PG")
+          RD_IP=$(docker inspect -f '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' "$RD")
+          test -n "$PG_IP" && test -n "$RD_IP"
+          export DB_HOST="$PG_IP"
+          export CELERY_BROKER_URL="redis://$RD_IP:6379/0"
+          export CELERY_RESULT_BACKEND="redis://$RD_IP:6379/0"
+          for i in $(seq 1 60); do
+            (echo > "/dev/tcp/$PG_IP/5432") >/dev/null 2>&1 && break
+            sleep 2
+          done
+          if command -v uv >/dev/null 2>&1; then
+            uv pip install --system -r requirements.txt pytest pytest-asyncio
+          else
+            pip install -r requirements.txt pytest pytest-asyncio
+          fi
+          alembic upgrade head
+          pytest tests/test_importer*.py tests/test_import_*.py tests/test_migration_*.py tests/test_phash_*.py tests/test_sidecar_*.py tests/test_scan_*.py tests/test_archive_extractor.py tests/test_backfill_phash.py -v -m integration --durations=15
+
+  intcore:
+    runs-on: python-ci
+    container:
+      image: git.fabledsword.com/bvandeusen/ci-python:3.14
+    env:
+      DB_USER: fabledcurator
+      DB_PASSWORD: ci_integration
+      DB_PORT: "5432"
+      DB_NAME: fabledcurator_test
+      SECRET_KEY: ci_integration_placeholder
+    services:
+      postgres:
+        image: pgvector/pgvector:pg16
+        env:
+          POSTGRES_USER: fabledcurator
+          POSTGRES_PASSWORD: ci_integration
+          POSTGRES_DB: fabledcurator_test
+        options: >-
+          --health-cmd "pg_isready -U fabledcurator"
+          --health-interval 10s
+          --health-timeout 5s
+          --health-retries 10
+      redis:
+        image: redis:7-alpine
+        options: >-
+          --health-cmd "redis-cli ping"
+          --health-interval 10s
+          --health-timeout 5s
+          --health-retries 10
+    steps:
+      - uses: actions/checkout@v4
+      - name: Cache pip wheels
+        uses: actions/cache@v4
+        with:
+          path: ~/.cache/pip
+          key: pip-${{ runner.os }}-py314-${{ hashFiles('requirements.txt') }}
+          restore-keys: |
+            pip-${{ runner.os }}-py314-
+      - name: Core integration shard (everything not api / importer / migration / phash / sidecar / scan / archive / backfill)
+        run: |
+          set -eux
+          echo "=== container landscape (diagnostic for filter scoping) ==="
+          docker ps -a --format '{{.ID}} {{.Image}} -> {{.Names}}'
+          echo "=== end landscape ==="
+          PG=$(docker ps --filter "name=intcore" --filter "ancestor=pgvector/pgvector:pg16" -q | head -n1)
+          RD=$(docker ps --filter "name=intcore" --filter "ancestor=redis:7-alpine" -q | head -n1)
+          test -n "$PG" && test -n "$RD"
+          PG_IP=$(docker inspect -f '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' "$PG")
+          RD_IP=$(docker inspect -f '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' "$RD")
+          test -n "$PG_IP" && test -n "$RD_IP"
+          export DB_HOST="$PG_IP"
+          export CELERY_BROKER_URL="redis://$RD_IP:6379/0"
+          export CELERY_RESULT_BACKEND="redis://$RD_IP:6379/0"
+          for i in $(seq 1 60); do
+            (echo > "/dev/tcp/$PG_IP/5432") >/dev/null 2>&1 && break
+            sleep 2
+          done
+          if command -v uv >/dev/null 2>&1; then
+            uv pip install --system -r requirements.txt pytest pytest-asyncio
+          else
+            pip install -r requirements.txt pytest pytest-asyncio
+          fi
+          alembic upgrade head
+          pytest tests/ -v -m integration --durations=15 \
+            --ignore-glob='tests/test_api_*.py' \
+            --ignore-glob='tests/test_importer*.py' \
+            --ignore-glob='tests/test_import_*.py' \
+            --ignore-glob='tests/test_migration_*.py' \
+            --ignore-glob='tests/test_phash_*.py' \
+            --ignore-glob='tests/test_sidecar_*.py' \
+            --ignore-glob='tests/test_scan_*.py' \
+            --ignore-glob='tests/test_archive_extractor.py' \
+            --ignore-glob='tests/test_backfill_phash.py'

.forgejo/workflows/extension.yml

@@ -1,8 +1,19 @@
 name: extension
+# Lint-only workflow. The sign-and-publish dance moved into build.yml's
+# `sign-extension` job (2026-05-25) — `:latest` now always bundles the XPI
+# because sign-extension runs as a build-web dependency in the SAME workflow,
+# eliminating the prior race between build.yml and a separate extension.yml.
+# Signed XPIs are cached in Forgejo Release Assets named `ext-<version>`.
 on:
   push:
     branches: [dev, main]
-    paths: ['extension/**']
+    paths:
+      - 'extension/**'
+      - '.forgejo/workflows/extension.yml'
+  pull_request:
+    branches: [main]
+    paths:
+      - 'extension/**'
   workflow_dispatch:
 
 jobs:
@@ -16,44 +27,3 @@ jobs:
         run: cd extension && npm install --no-save --no-audit --no-fund
       - name: Lint
         run: cd extension && npm run lint
-
-  sign-and-publish:
-    needs: lint
-    if: github.ref == 'refs/heads/main'
-    runs-on: python-ci
-    container:
-      image: node:22-bookworm-slim
-    steps:
-      - uses: actions/checkout@v4
-        with:
-          token: ${{ secrets.RELEASE_TOKEN }}
-      - name: Install web-ext + git
-        run: |
-          apt-get update && apt-get install -y --no-install-recommends git ca-certificates
-          cd extension && npm install --no-save --no-audit --no-fund
-      - name: Sign XPI
-        run: cd extension && npm run sign
-        env:
-          WEB_EXT_API_KEY: ${{ secrets.MOZILLA_AMO_JWT_KEY }}
-          WEB_EXT_API_SECRET: ${{ secrets.MOZILLA_AMO_JWT_SECRET }}
-      - name: Commit signed XPI to frontend/public/extension/
-        run: |
-          set -e
-          XPI=$(ls extension/web-ext-artifacts/fabledcurator-*.xpi | head -1)
-          if [ -z "$XPI" ]; then
-            echo "No XPI produced by web-ext sign — exiting"
-            exit 1
-          fi
-          mkdir -p frontend/public/extension
-          cp "$XPI" frontend/public/extension/
-          # Also copy as -latest.xpi so the FC server can serve a stable URL.
-          cp "$XPI" "frontend/public/extension/fabledcurator-latest.xpi"
-          git config user.name  "FC extension CI"
-          git config user.email "noreply@fabledsword.com"
-          git add frontend/public/extension/
-          if git diff --cached --quiet; then
-            echo "No changes to commit"
-          else
-            git commit -m "ext: publish signed XPI $(basename $XPI)"
-            git push origin HEAD:main
-          fi

alembic/versions/0019_import_batch_refreshed.py

@@ -0,0 +1,38 @@
+"""import_batch.refreshed counter for deep-scan sidecar re-application
+
+Revision ID: 0019
+Revises: 0018
+Create Date: 2026-05-25
+
+Adds a `refreshed` counter to `import_batch`, mirroring the existing
+`imported`/`skipped`/`failed`/`attachments` columns. Deep scan now
+re-applies sidecar metadata to already-imported files (the IR feature
+that didn't make the FC port the first time); a "refreshed" outcome
+increments this counter so the UI can surface "X new, Y refreshed"
+instead of the misleading "Scan complete — no new files" message.
+
+server_default=0 backfills existing rows in place — no UPDATE needed.
+"""
+from typing import Sequence, Union
+
+import sqlalchemy as sa
+from alembic import op
+
+revision: str = "0019"
+down_revision: Union[str, None] = "0018"
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    op.add_column(
+        "import_batch",
+        sa.Column(
+            "refreshed", sa.Integer(),
+            nullable=False, server_default=sa.text("0"),
+        ),
+    )
+
+
+def downgrade() -> None:
+    op.drop_column("import_batch", "refreshed")

backend/app/api/gallery.py

@@ -42,6 +42,8 @@ async def scroll():
                     "width": i.width,
                     "height": i.height,
                     "created_at": i.created_at.isoformat(),
+                    "posted_at": i.posted_at.isoformat() if i.posted_at else None,
+                    "effective_date": i.effective_date.isoformat(),
                     "thumbnail_url": i.thumbnail_url,
                     "artist": i.artist,
                 }

backend/app/api/import_admin.py

@@ -47,10 +47,13 @@ async def status():
         if active:
             payload["active_batch"] = {
                 "id": active.id,
+                "source_path": active.source_path,
+                "scan_mode": active.scan_mode,
                 "total_files": active.total_files,
                 "imported": active.imported,
                 "skipped": active.skipped,
                 "failed": active.failed,
+                "refreshed": active.refreshed,
                 "started_at": active.started_at.isoformat(),
             }
         return jsonify(payload)
@@ -120,6 +123,83 @@ async def retry_failed():
     return jsonify({"retried": len(failed_ids)})
 
 
+@import_admin_bp.route("/clear-stuck", methods=["POST"])
+async def clear_stuck():
+    """Force any non-terminal ImportTask (status in pending/queued/
+    processing) to 'failed' AND finalize any ImportBatch that ends up
+    with no active children. Escape hatch for the operator when the
+    automatic recover_interrupted_tasks sweep keeps re-queueing the
+    same stuck row forever (e.g., underlying file is genuinely broken
+    and the import keeps OSError-looping at PIL load).
+
+    Idempotent + non-destructive: rows survive as 'failed' so the
+    Retry-Failed button can re-attempt them once whatever was broken
+    is fixed. Banked 2026-05-25 — operator hit 3 large PNGs that
+    autoretry-looped for 2 days after a corrupt-data PIL OSError.
+    """
+    async with get_session() as session:
+        stuck_ids = (
+            await session.execute(
+                select(ImportTask.id).where(
+                    ImportTask.status.in_(["pending", "queued", "processing"])
+                )
+            )
+        ).scalars().all()
+        if stuck_ids:
+            await session.execute(
+                update(ImportTask)
+                .where(ImportTask.id.in_(stuck_ids))
+                .values(
+                    status="failed",
+                    finished_at=datetime.now(UTC),
+                    error=(
+                        "manually cleared via /api/import/clear-stuck "
+                        "— stuck in non-terminal state; retry once "
+                        "underlying cause (corrupt file, missing model, "
+                        "etc.) is resolved"
+                    ),
+                )
+            )
+
+        # Finalize any 'running' ImportBatch that no longer has any
+        # active children. The "Scanning..." banner is driven by
+        # /api/import/status finding a running batch; left untouched,
+        # it would persist forever after the stuck-task clear.
+        running_batches = (
+            await session.execute(
+                select(ImportBatch.id).where(ImportBatch.status == "running")
+            )
+        ).scalars().all()
+        finalized_batches = 0
+        for batch_id in running_batches:
+            still_active = (
+                await session.execute(
+                    select(ImportTask.id)
+                    .where(ImportTask.batch_id == batch_id)
+                    .where(ImportTask.status.in_(
+                        ["pending", "queued", "processing"]
+                    ))
+                    .limit(1)
+                )
+            ).scalar_one_or_none()
+            if still_active is None:
+                await session.execute(
+                    update(ImportBatch)
+                    .where(ImportBatch.id == batch_id)
+                    .values(
+                        status="complete",
+                        finished_at=datetime.now(UTC),
+                    )
+                )
+                finalized_batches += 1
+        await session.commit()
+
+    return jsonify({
+        "tasks_failed": len(stuck_ids),
+        "batches_finalized": finalized_batches,
+    })
+
+
 @import_admin_bp.route("/clear-completed", methods=["POST"])
 async def clear_completed():
     body = await request.get_json(silent=True) or {}

backend/app/models/import_batch.py

@@ -26,6 +26,10 @@ class ImportBatch(Base):
     skipped: Mapped[int] = mapped_column(Integer, nullable=False, default=0)
     failed: Mapped[int] = mapped_column(Integer, nullable=False, default=0)
     attachments: Mapped[int] = mapped_column(Integer, nullable=False, default=0)
+    # Deep-scan only: count of already-imported files whose sidecar metadata
+    # got re-applied this run (post/source/provenance upsert). Stays 0 on
+    # quick-scan batches. See `Importer.import_one(deep_scan=True)`.
+    refreshed: Mapped[int] = mapped_column(Integer, nullable=False, default=0)
 
     status: Mapped[str] = mapped_column(String(16), nullable=False, default="running", index=True)
     # running | complete | cancelled

backend/app/scripts/download_models.py

@@ -25,12 +25,31 @@ def _snapshot(repo_id: str, dest: Path, allow_patterns: list[str] | None) -> Non
 
 
 def ensure_camie() -> None:
+    """Fetch Camie v2 weights + metadata.
+
+    v2 layout (HuggingFace Camais03/camie-tagger-v2): the ONNX file is
+    named camie-tagger-v2.onnx (not model.onnx) and tags ship inside
+    camie-tagger-v2-metadata.json (not selected_tags.csv). Both at root.
+    The repo also contains app/, game/, training/, images/ subdirs full
+    of setup/demo files we don't need — allow_patterns scopes the fetch
+    to just the inference essentials (~790 MB instead of ~2 GB).
+    """
     dest = MODEL_ROOT / "camie"
-    if (dest / "model.onnx").is_file() and (dest / "selected_tags.csv").is_file():
+    model_file = dest / "camie-tagger-v2.onnx"
+    meta_file = dest / "camie-tagger-v2-metadata.json"
+    if model_file.is_file() and meta_file.is_file():
         print(f"[download_models] Camie present at {dest}")
         return
     print(f"[download_models] Fetching {CAMIE_REPO} -> {dest}")
-    _snapshot(CAMIE_REPO, dest, ["model.onnx", "selected_tags.csv", "*.json"])
+    _snapshot(
+        CAMIE_REPO, dest,
+        [
+            "camie-tagger-v2.onnx",
+            "camie-tagger-v2-metadata.json",
+            "config.json",
+            "config.yaml",
+        ],
+    )
 
 
 def ensure_siglip() -> None:

backend/app/services/gallery_service.py

@@ -1,26 +1,34 @@
 """Cursor-paginated gallery queries.
 
-Cursor format: opaque base64-encoded "<iso8601_created_at>:<image_id>".
-Pagination key is (created_at DESC, id DESC) so we don't drift when new
-imports arrive between page loads. Decoding rejects malformed cursors with
-a ValueError; the API layer translates that to HTTP 400.
+Cursor format: opaque base64-encoded "<iso8601_effective_date>:<image_id>".
+
+Pagination key is (effective_date DESC, id DESC) where effective_date is
+COALESCE(post.post_date, image_record.created_at) so the gallery surfaces
+images by ORIGINAL publish date when known, falling back to FC's scan
+date. Important for migrated content: ~57k IR images scanned in a single
+week would otherwise all share the same created_at and pile up in one
+month bucket. The effective_date spreads them across the years they
+were originally published.
+
+Decoding rejects malformed cursors with a ValueError; the API layer
+translates that to HTTP 400.
 """
 
 import base64
 from dataclasses import dataclass
 from datetime import datetime
 
-from sqlalchemy import and_, exists, func, or_, select
+from sqlalchemy import Select, and_, exists, func, or_, select
 from sqlalchemy.ext.asyncio import AsyncSession
 
-from ..models import Artist, ImageProvenance, ImageRecord, Source, Tag
+from ..models import Artist, ImageProvenance, ImageRecord, Post, Source, Tag
 from ..models.tag import image_tag
 
 CURSOR_SEPARATOR = "|"
 
 
-def encode_cursor(created_at: datetime, image_id: int) -> str:
-    raw = f"{created_at.isoformat()}{CURSOR_SEPARATOR}{image_id}"
+def encode_cursor(effective_date: datetime, image_id: int) -> str:
+    raw = f"{effective_date.isoformat()}{CURSOR_SEPARATOR}{image_id}"
     return base64.urlsafe_b64encode(raw.encode()).decode()
 
 
@@ -33,6 +41,26 @@ def decode_cursor(cursor: str) -> tuple[datetime, int]:
         raise ValueError(f"invalid cursor: {cursor!r}") from exc
 
 
+def _effective_date_col():
+    """SQL expression: COALESCE(post.post_date, image_record.created_at).
+
+    Used as the canonical sort/group/filter key across the gallery so
+    images backfilled with primary_post_id (e.g. via tag_apply phase 4)
+    surface at their original publish date, not their FC import date.
+    Images without a Post (or with Post.post_date NULL) fall back to
+    image_record.created_at and still order coherently against
+    post-attached ones.
+    """
+    return func.coalesce(Post.post_date, ImageRecord.created_at)
+
+
+def _outer_join_primary_post(stmt: Select) -> Select:
+    """LEFT JOIN Post on ImageRecord.primary_post_id so the COALESCE
+    above sees Post.post_date when available. Images without a post
+    survive the join as NULL on the Post side; COALESCE handles it."""
+    return stmt.outerjoin(Post, Post.id == ImageRecord.primary_post_id)
+
+
 @dataclass(frozen=True)
 class GalleryImage:
     id: int
@@ -41,7 +69,9 @@ class GalleryImage:
     mime: str
     width: int | None
     height: int | None
-    created_at: datetime
+    created_at: datetime          # FC's row-insert time
+    effective_date: datetime      # COALESCE(post.post_date, created_at)
+    posted_at: datetime | None    # post.post_date if known, else None
     thumbnail_url: str
     artist: dict | None = None
 
@@ -78,7 +108,7 @@ def _require_single_filter(tag_id, post_id, artist_id) -> None:
 def _provenance_clause(post_id, artist_id):
     """Correlated EXISTS clause (NOT a join) so an image with multiple
     matching provenance rows is returned exactly once and the
-    (created_at DESC, id DESC) cursor ordering is unaffected."""
+    (effective_date DESC, id DESC) cursor ordering is unaffected."""
     if post_id is not None:
         return exists().where(
             ImageProvenance.image_record_id == ImageRecord.id,
@@ -125,7 +155,9 @@ class GalleryService:
             raise ValueError("limit must be between 1 and 200")
         _require_single_filter(tag_id, post_id, artist_id)
 
-        stmt = select(ImageRecord)
+        eff = _effective_date_col()
+        stmt = select(ImageRecord, Post.post_date, eff.label("eff"))
+        stmt = _outer_join_primary_post(stmt)
         if tag_id is not None:
             stmt = stmt.join(image_tag, image_tag.c.image_record_id == ImageRecord.id).where(
                 image_tag.c.tag_id == tag_id
@@ -138,34 +170,38 @@ class GalleryService:
             cur_ts, cur_id = decode_cursor(cursor)
             stmt = stmt.where(
                 or_(
-                    ImageRecord.created_at < cur_ts,
-                    and_(ImageRecord.created_at == cur_ts, ImageRecord.id < cur_id),
+                    eff < cur_ts,
+                    and_(eff == cur_ts, ImageRecord.id < cur_id),
                 )
             )
 
-        stmt = stmt.order_by(ImageRecord.created_at.desc(), ImageRecord.id.desc()).limit(limit + 1)
-        rows = (await self.session.execute(stmt)).scalars().all()
+        stmt = stmt.order_by(eff.desc(), ImageRecord.id.desc()).limit(limit + 1)
+        rows = (await self.session.execute(stmt)).all()
 
         next_cursor = None
         if len(rows) > limit:
-            last = rows[limit - 1]
-            next_cursor = encode_cursor(last.created_at, last.id)
+            last_record, _last_posted_at, last_eff = rows[limit - 1]
+            next_cursor = encode_cursor(last_eff, last_record.id)
             rows = rows[:limit]
 
-        artists = await _artists_for(self.session, [r.id for r in rows])
+        artists = await _artists_for(
+            self.session, [r[0].id for r in rows]
+        )
         images = [
             GalleryImage(
-                id=r.id,
-                path=r.path,
-                sha256=r.sha256,
-                mime=r.mime,
-                width=r.width,
-                height=r.height,
-                created_at=r.created_at,
-                thumbnail_url=thumbnail_url(r.sha256, r.mime),
-                artist=artists.get(r.id),
+                id=record.id,
+                path=record.path,
+                sha256=record.sha256,
+                mime=record.mime,
+                width=record.width,
+                height=record.height,
+                created_at=record.created_at,
+                effective_date=eff_date,
+                posted_at=posted_at,
+                thumbnail_url=thumbnail_url(record.sha256, record.mime),
+                artist=artists.get(record.id),
             )
-            for r in rows
+            for record, posted_at, eff_date in rows
         ]
         return GalleryPage(
             images=images,
@@ -179,11 +215,13 @@ class GalleryService:
         post_id: int | None = None,
         artist_id: int | None = None,
     ) -> list[TimelineBucket]:
-        year_col = func.date_part("year", ImageRecord.created_at).label("yr")
-        month_col = func.date_part("month", ImageRecord.created_at).label("mo")
+        eff = _effective_date_col()
+        year_col = func.date_part("year", eff).label("yr")
+        month_col = func.date_part("month", eff).label("mo")
         stmt = select(
             year_col, month_col, func.count(ImageRecord.id).label("cnt")
         )
+        stmt = _outer_join_primary_post(stmt)
         _require_single_filter(tag_id, post_id, artist_id)
         if tag_id is not None:
             stmt = stmt.join(image_tag, image_tag.c.image_record_id == ImageRecord.id).where(
@@ -201,14 +239,17 @@ class GalleryService:
         post_id: int | None = None, artist_id: int | None = None,
     ) -> str | None:
         """Returns a cursor that, when passed to scroll(), positions at the
-        first image of the given year-month. None if the bucket is empty.
+        first image of the given year-month (by effective_date, not
+        created_at). None if the bucket is empty.
         """
         from sqlalchemy import extract
 
-        stmt = select(ImageRecord).where(
-            extract("year", ImageRecord.created_at) == year,
-            extract("month", ImageRecord.created_at) == month,
+        eff = _effective_date_col()
+        stmt = select(ImageRecord, eff.label("eff")).where(
+            extract("year", eff) == year,
+            extract("month", eff) == month,
         )
+        stmt = _outer_join_primary_post(stmt)
         _require_single_filter(tag_id, post_id, artist_id)
         if tag_id is not None:
             stmt = stmt.join(image_tag, image_tag.c.image_record_id == ImageRecord.id).where(
@@ -217,13 +258,14 @@ class GalleryService:
         prov = _provenance_clause(post_id, artist_id)
         if prov is not None:
             stmt = stmt.where(prov)
-        stmt = stmt.order_by(ImageRecord.created_at.desc(), ImageRecord.id.desc()).limit(1)
-        first = (await self.session.execute(stmt)).scalar_one_or_none()
+        stmt = stmt.order_by(eff.desc(), ImageRecord.id.desc()).limit(1)
+        first = (await self.session.execute(stmt)).first()
         if first is None:
             return None
+        record, eff_date = first
         # Cursor is exclusive; we encode a cursor with id+1 so the row itself
         # is the first result in the next scroll().
-        return encode_cursor(first.created_at, first.id + 1)
+        return encode_cursor(eff_date, record.id + 1)
 
     async def get_image_with_tags(self, image_id: int) -> dict | None:
         record = await self.session.get(ImageRecord, image_id)
@@ -236,6 +278,13 @@ class GalleryService:
             .order_by(Tag.kind.asc(), Tag.name.asc())
         )
         tags = (await self.session.execute(tag_stmt)).scalars().all()
+        # Fetch the canonical post.post_date for this image (if any) so
+        # the modal can show "Posted on <date>" alongside import date.
+        posted_at = None
+        if record.primary_post_id is not None:
+            posted_at = (await self.session.execute(
+                select(Post.post_date).where(Post.id == record.primary_post_id)
+            )).scalar_one_or_none()
         neighbors = await self._neighbors(record)
         # Direct artist FK — used by the modal's ProvenancePanel as a
         # fallback when ImageProvenance is empty (i.e., filesystem-
@@ -256,6 +305,7 @@ class GalleryService:
             "size_bytes": record.size_bytes,
             "integrity_status": record.integrity_status,
             "created_at": record.created_at.isoformat(),
+            "posted_at": posted_at.isoformat() if posted_at else None,
             "thumbnail_url": thumbnail_url(record.sha256, record.mime),
             "image_url": f"/images/{record.path.split('/images/', 1)[-1]}",
             "artist": (
@@ -275,34 +325,41 @@ class GalleryService:
         }
 
     async def _neighbors(self, record: ImageRecord) -> dict:
-        prev_stmt = (
-            select(ImageRecord.id)
-            .where(
+        # Compute the boundary image's effective_date in Python (one query
+        # below + the SELECT we already have on `record`) and use it for
+        # the neighbor comparison. Cheaper than re-deriving in SQL via
+        # correlated subquery.
+        boundary_eff = record.created_at
+        if record.primary_post_id is not None:
+            post_date = (await self.session.execute(
+                select(Post.post_date).where(Post.id == record.primary_post_id)
+            )).scalar_one_or_none()
+            if post_date is not None:
+                boundary_eff = post_date
+
+        eff = _effective_date_col()
+        prev_stmt = _outer_join_primary_post(
+            select(ImageRecord.id).where(
                 or_(
-                    ImageRecord.created_at > record.created_at,
+                    eff > boundary_eff,
                     and_(
-                        ImageRecord.created_at == record.created_at,
+                        eff == boundary_eff,
                         ImageRecord.id > record.id,
                     ),
                 )
             )
-            .order_by(ImageRecord.created_at.asc(), ImageRecord.id.asc())
-            .limit(1)
-        )
-        next_stmt = (
-            select(ImageRecord.id)
-            .where(
+        ).order_by(eff.asc(), ImageRecord.id.asc()).limit(1)
+        next_stmt = _outer_join_primary_post(
+            select(ImageRecord.id).where(
                 or_(
-                    ImageRecord.created_at < record.created_at,
+                    eff < boundary_eff,
                     and_(
-                        ImageRecord.created_at == record.created_at,
+                        eff == boundary_eff,
                         ImageRecord.id < record.id,
                     ),
                 )
             )
-            .order_by(ImageRecord.created_at.desc(), ImageRecord.id.desc())
-            .limit(1)
-        )
+        ).order_by(eff.desc(), ImageRecord.id.desc()).limit(1)
         prev_id = (await self.session.execute(prev_stmt)).scalar_one_or_none()
         next_id = (await self.session.execute(next_stmt)).scalar_one_or_none()
         return {"prev_id": prev_id, "next_id": next_id}
@@ -311,9 +368,11 @@ class GalleryService:
 def _group_by_year_month(
     images: list[GalleryImage],
 ) -> list[tuple[int, int, list[int]]]:
+    """Group by effective_date's year/month so migrated content surfaces
+    in the publish-date buckets, not the FC-scan-date bucket."""
     groups: list[tuple[int, int, list[int]]] = []
     for img in images:
-        y, m = img.created_at.year, img.created_at.month
+        y, m = img.effective_date.year, img.effective_date.month
         if groups and groups[-1][0] == y and groups[-1][1] == m:
             groups[-1][2].append(img.id)
         else:

backend/app/services/importer.py

@@ -51,7 +51,14 @@ class SkipReason(StrEnum):
 
 @dataclass(frozen=True)
 class ImportResult:
-    status: str  # 'imported'|'skipped'|'failed'|'superseded'|'attached'
+    # 'imported'   — new ImageRecord row created
+    # 'superseded' — existing ImageRecord row got the new file (larger) + sidecar
+    # 'attached'   — non-media saved as PostAttachment
+    # 'refreshed'  — deep scan re-applied sidecar / filled NULL phash / NULL
+    #                artist on an already-imported row (no new ImageRecord)
+    # 'skipped'    — no work done (true duplicate, too small, etc.)
+    # 'failed'     — pipeline error
+    status: str
     image_id: int | None = None
     skip_reason: SkipReason | None = None
     error: str | None = None
@@ -71,6 +78,31 @@ def is_video(path: Path) -> bool:
     return path.suffix.lower() in VIDEO_EXTS
 
 
+def _safe_ext(path: Path) -> str:
+    """Conservatively extract a file extension for PostAttachment.ext
+    (varchar(32)).
+
+    gallery-dl produces some filenames with URL-encoded query-string
+    artifacts embedded into the basename (e.g.
+    `79507046_media_..._https___www.patreon.com_media-u_Z0FBQUFBQm5q...`).
+    `Path.suffix` finds the LAST dot and returns everything after, which
+    in those cases yields a 50+ char "extension" of mostly base64-ish
+    junk. That blows the column. Operator-flagged 2026-05-25.
+
+    Real extensions are short and alphanumeric. We accept anything ≤ 16
+    chars where every post-dot character is alphanumeric; anything else
+    means the input wasn't a real extension and we return the empty
+    string. ext is nullable-ish (empty string still satisfies NOT NULL)
+    and consumers should treat "" as "no known extension".
+    """
+    suffix = path.suffix.lower()
+    if not suffix or len(suffix) > 16:
+        return ""
+    if not all(c.isalnum() for c in suffix[1:]):
+        return ""
+    return suffix
+
+
 def _mime_for(path: Path) -> str:
     suffix = path.suffix.lower()
     image_mimes = {
@@ -126,6 +158,49 @@ class Importer:
         self.settings = settings
         self.deep = deep
         self.attachments = AttachmentStore(images_root)
+        # phash near-dup candidate cache. Archive imports call _import_media
+        # per-member; without this cache the per-member SELECT *FROM
+        # image_record WHERE phash IS NOT NULL fetch repeats N times and a
+        # large library × many-member archive blew past soft_time_limit
+        # (300s) — operator-flagged 2026-05-25. Loaded lazily on first
+        # need, appended to on every imported/superseded outcome, never
+        # invalidated mid-Importer (Importer instances are per-task /
+        # per-archive-import so cross-instance staleness is harmless).
+        self._phash_candidates: list[tuple] | None = None
+
+    def _phash_candidates_cache(self) -> list[tuple]:
+        """Cached `(phash, width, height, id)` rows from image_record.
+        Loaded on first call, appended-to on subsequent imported/
+        superseded outcomes. Soft-timeout pattern: an archive with N
+        members + a library of M existing rows used to do N × M-row
+        fetches (operator-flagged 2026-05-25); now it's exactly one.
+
+        The per-task lifecycle of Importer (instantiated fresh by
+        import_media_file) bounds the cache's staleness window: cross-
+        process changes (other workers importing concurrently) won't
+        be reflected, but that's the same race the un-cached version
+        had — `find_similar` is best-effort anyway."""
+        if self._phash_candidates is None:
+            rows = self.session.execute(
+                select(
+                    ImageRecord.phash,
+                    ImageRecord.width,
+                    ImageRecord.height,
+                    ImageRecord.id,
+                ).where(ImageRecord.phash.is_not(None))
+            ).all()
+            self._phash_candidates = [
+                (r.phash, r.width or 0, r.height or 0, r.id) for r in rows
+            ]
+        return self._phash_candidates
+
+    def _phash_cache_append(self, phash, width, height, image_id) -> None:
+        """Append a freshly-imported row to the cache so subsequent
+        members of the same archive can match against it."""
+        if self._phash_candidates is not None and phash is not None:
+            self._phash_candidates.append(
+                (phash, width or 0, height or 0, image_id)
+            )
 
     def import_one(self, source: Path) -> ImportResult:
         """Dispatch by kind. Media → normal pipeline. Archive → extract
@@ -209,7 +284,7 @@ class Importer:
                 sha256=sha,
                 path=stored,
                 original_filename=source.name,
-                ext=source.suffix.lower(),
+                ext=_safe_ext(source),
                 mime=_mime_for(source),
                 size_bytes=source.stat().st_size,
             ))
@@ -280,7 +355,18 @@ class Importer:
                 )
 
             if self.settings.skip_transparent and has_alpha:
-                pct = self._transparency_pct(source)
+                try:
+                    pct = self._transparency_pct(source)
+                except OSError as exc:
+                    # PIL.verify() at line 263 only validates header structure;
+                    # truncated/corrupt pixel data only surfaces when load()
+                    # actually decodes (here via getchannel('A')). Convert to
+                    # invalid_image skip so the Celery autoretry loop doesn't
+                    # bounce the same broken file forever.
+                    return ImportResult(
+                        status="skipped", skip_reason=SkipReason.invalid_image,
+                        error=f"PIL load failed during transparency check: {exc}",
+                    )
                 if pct >= self.settings.transparency_threshold:
                     return ImportResult(
                         status="skipped", skip_reason=SkipReason.too_transparent,
@@ -302,21 +388,18 @@ class Importer:
         # Perceptual near-dup (images only; videos keep phash NULL).
         phash = None
         if not is_video(source):
-            with Image.open(source) as im:
-                phash = compute_phash(im)
+            try:
+                with Image.open(source) as im:
+                    phash = compute_phash(im)
+            except OSError as exc:
+                # Same rationale as the transparency-check guard above:
+                # broken-pixel-data files pass verify() but blow up here.
+                return ImportResult(
+                    status="skipped", skip_reason=SkipReason.invalid_image,
+                    error=f"PIL load failed during phash compute: {exc}",
+                )
             if phash is not None:
-                cand_rows = self.session.execute(
-                    select(
-                        ImageRecord.phash,
-                        ImageRecord.width,
-                        ImageRecord.height,
-                        ImageRecord.id,
-                    ).where(ImageRecord.phash.is_not(None))
-                ).all()
-                candidates = [
-                    (c.phash, c.width or 0, c.height or 0, c.id)
-                    for c in cand_rows
-                ]
+                candidates = self._phash_candidates_cache()
                 rel, match_id = find_similar(
                     phash, width or 0, height or 0,
                     candidates, self.settings.phash_threshold,
@@ -350,6 +433,7 @@ class Importer:
         )
         self.session.add(record)
         self.session.flush()
+        self._phash_cache_append(phash, width, height, record.id)
 
         # Folder→artist (anchored to attribution_path).
         artist = None
@@ -373,13 +457,27 @@ class Importer:
     ) -> ImportResult:
         """Deep scan: backfill phash/provenance/artist on an
         already-imported record. METADATA ONLY — never re-runs the pHash
-        near-dup / supersede path. NULL-only, idempotent."""
+        near-dup / supersede path. NULL-only on phash/artist, additive on
+        sidecar Post/Source/ImageProvenance (via _apply_sidecar).
+        Idempotent: a second deep-scan over the same file finds nothing
+        to refresh and is a no-op.
+
+        Returns status="refreshed" so the UI can surface the work done
+        instead of the prior misleading "skipped/duplicate_hash" reading.
+        Operator-flagged 2026-05-25 — IR has had this; FC inherited it
+        as a no-op skip during the original port and the UI showed deep
+        scan as "completed with no changes" even when sidecar metadata
+        actually got re-applied to N existing rows.
+        """
         if existing.phash is None and not is_video(source):
             try:
                 with Image.open(source) as im:
                     ph = compute_phash(im)
                 if ph is not None:
                     existing.phash = ph
+                    # Promoted from NULL to non-NULL → cache is now stale
+                    # (this row would newly qualify for the candidates set).
+                    self._phash_candidates = None
             except Exception as exc:
                 log.warning("deep rephash failed for %s: %s", source, exc)
 
@@ -392,10 +490,7 @@ class Importer:
 
         self._apply_sidecar(existing, attribution_path, artist)
         self.session.commit()
-        return ImportResult(
-            status="skipped", skip_reason=SkipReason.duplicate_hash,
-            image_id=existing.id, error="deep: re-derived",
-        )
+        return ImportResult(status="refreshed", image_id=existing.id)
 
     def attach_in_place(
         self,
@@ -479,18 +574,7 @@ class Importer:
             except Exception:
                 phash = None
             if phash is not None:
-                cand_rows = self.session.execute(
-                    select(
-                        ImageRecord.phash,
-                        ImageRecord.width,
-                        ImageRecord.height,
-                        ImageRecord.id,
-                    ).where(ImageRecord.phash.is_not(None))
-                ).all()
-                candidates = [
-                    (c.phash, c.width or 0, c.height or 0, c.id)
-                    for c in cand_rows
-                ]
+                candidates = self._phash_candidates_cache()
                 rel, match_id = find_similar(
                     phash, width or 0, height or 0,
                     candidates, self.settings.phash_threshold,
@@ -525,6 +609,7 @@ class Importer:
             record.artist_id = artist.id
         self.session.add(record)
         self.session.flush()
+        self._phash_cache_append(phash, width, height, record.id)
 
         # Sidecar provenance (best-effort). When `source` is passed, link
         # the post to that subscription Source instead of creating a new
@@ -703,6 +788,15 @@ class Importer:
         row id (so tags/series/curation stay attached). ML is cleared so
         the import task re-derives it on the new pixels.
 
+        After the file swap, the new file's adjacent gallery-dl sidecar
+        (if any) is applied via _apply_sidecar — operator-flagged
+        2026-05-25: scanning a GS download dir with smaller IR-migrated
+        images on the receiving end used to swap files but lose the GS
+        sidecar's post metadata entirely. _apply_sidecar is additive
+        (find-or-create Post / Source / ImageProvenance, NULL-only
+        primary_post_id update) so any pre-existing Post linkage
+        survives untouched.
+
         If `new_path` is provided, `source` is assumed to ALREADY be at
         that path (FC-3c attach_in_place case) — skip the copy step.
         Otherwise the file is copied via _copy_to_library."""
@@ -731,6 +825,26 @@ class Importer:
         # created_at intentionally preserved; updated_at auto-bumps.
         self.session.flush()
         self.session.commit()
+        # The phash candidate cache (used to avoid N+1 selects during
+        # archive imports) is now stale for `existing.id` — the row's
+        # phash/dimensions changed. Invalidate; the next call re-fetches.
+        self._phash_candidates = None
+
+        # Sidecar enrichment from the new (larger) file's location.
+        # _apply_sidecar resolves artist from the sidecar itself if the
+        # existing row has none, and is internally guarded against
+        # missing-or-malformed sidecars (silent return).
+        try:
+            self._apply_sidecar(existing, source, None)
+        except Exception as exc:
+            # Don't unwind the supersede DB swap if sidecar parsing
+            # blows up unexpectedly — the file replacement is the
+            # critical operation, sidecar is enrichment.
+            log.warning(
+                "sidecar enrichment failed during supersede of "
+                "image_record.id=%s from %s: %s",
+                existing.id, source, exc,
+            )
 
         for stale in (old_path, old_thumb):
             if not stale or stale == str(dest):

backend/app/services/migrators/tag_apply.py

@@ -37,13 +37,25 @@ from ...utils.slug import slugify
 from .ir_ingest import manifest_path
 
 # Per-platform artist-profile URL — used as Source.url when restoring
-# IR PostMetadata into FC. Keep this table in sync with
-# backend/app/services/extension_service.py:_PLATFORM_PATTERNS and
-# extension/lib/platforms.js.
+# IR PostMetadata into FC. Must cover every platform that
+# backend/app/services/extension_service.py:_PLATFORM_PATTERNS
+# recognizes; an entry missing here silently drops ALL PostMetadata for
+# that platform during phase 4 (operator hit this 2026-05-25:
+# DeviantArt + Pixiv posts in the IR migration produced empty
+# ImageProvenance because they fell through this table).
+#
+# Pixiv caveat: the real profile URL takes a numeric user_id
+# (https://www.pixiv.net/users/12345), but IR's PostMetadata.artist
+# stores the display name not the id. We use the slugified name here
+# so we preserve the artist→post→image linkage; the resulting Source.url
+# won't resolve in a browser and the operator may want to manually fix
+# it via Settings → Subscriptions once the migration lands.
 _PLATFORM_PROFILE_URL = {
     "patreon":       "https://www.patreon.com/{slug}",
     "subscribestar": "https://www.subscribestar.com/{slug}",
     "hentaifoundry": "https://www.hentai-foundry.com/user/{slug}",
+    "deviantart":    "https://www.deviantart.com/{slug}",
+    "pixiv":         "https://www.pixiv.net/users/{slug}",
 }
 
 
@@ -110,7 +122,14 @@ async def _ensure_provenance(
     db: AsyncSession, *,
     image_id: int, post_id: int, source_id: int, dry_run: bool,
 ) -> bool:
-    """Returns True if a new ImageProvenance row was inserted."""
+    """Returns True if a new ImageProvenance row was inserted.
+
+    Also sets ImageRecord.primary_post_id to this post if the image
+    doesn't already have one — preserves any primary_post_id already
+    assigned at download time by the importer (don't clobber). This is
+    the linkage gallery_service.py uses to surface Post.post_date as
+    the image's effective date for sort/group/jump/neighbor nav.
+    """
     existing = (await db.execute(
         select(ImageProvenance.id).where(
             ImageProvenance.image_record_id == image_id,
@@ -118,6 +137,18 @@ async def _ensure_provenance(
             ImageProvenance.source_id == source_id,
         )
     )).scalar_one_or_none()
+
+    # Whether-or-not the provenance row already exists, ensure the
+    # image's primary_post_id is set so the gallery date-coalesce works.
+    # Idempotent: only writes when currently NULL.
+    if not dry_run:
+        await db.execute(
+            ImageRecord.__table__.update()
+            .where(ImageRecord.id == image_id)
+            .where(ImageRecord.primary_post_id.is_(None))
+            .values(primary_post_id=post_id)
+        )
+
     if existing is not None:
         return False
     if dry_run:

backend/app/services/ml/embedder.py

@@ -34,10 +34,21 @@ class Embedder:
         if self._model is not None:
             return
         import torch
-        from transformers import AutoModel, AutoProcessor
+        from transformers import AutoModel, SiglipImageProcessor
 
         self._torch = torch
-        self._processor = AutoProcessor.from_pretrained(str(self._model_dir))
+        # FC's embedder only does IMAGE inference — never text. AutoProcessor
+        # loads the full processor including SiglipTokenizer, which requires
+        # the sentencepiece library at import time even if we never call it.
+        # SiglipImageProcessor loads ONLY preprocessor_config.json (image
+        # side) and skips the tokenizer config entirely. Operator hit the
+        # ImportError 2026-05-25 once the ml-worker started actually running
+        # tag_and_embed; switching to the image-only loader avoids the
+        # tokenizer dep without adding ~30 MB of unused C++ build to the
+        # lean ml-worker image.
+        self._processor = SiglipImageProcessor.from_pretrained(
+            str(self._model_dir)
+        )
         self._model = AutoModel.from_pretrained(str(self._model_dir))
         self._model.eval()
 

backend/app/services/ml/tagger.py

@@ -4,12 +4,14 @@ CPU-only, single-image at a time. Loaded lazily inside the ml-worker
 process; NOT thread-safe — the ml queue worker must run --concurrency=1
 (set by the FC-1 entrypoint).
 
-Camie's selected_tags.csv columns: tag_id,name,category,count
-where category is a string: general|character|copyright|artist|meta|rating|year
-(unlike WD14's integer Danbooru category ids).
+v2 layout reference: HuggingFace Camais03/camie-tagger-v2 root has
+camie-tagger-v2.onnx (789 MB) + camie-tagger-v2-metadata.json (7.77 MB)
++ config.json. Tags ship as nested JSON, not CSV. Preprocessing and
+output handling follow the published onnx_inference.py reference:
+ImageNet normalize, NCHW layout, sigmoid on refined logits (output[1]).
 """
 
-import csv
+import json
 import os
 from dataclasses import dataclass
 from pathlib import Path
@@ -28,6 +30,8 @@ ImageFile.LOAD_TRUNCATED_IMAGES = True
 
 MODEL_NAME = os.environ.get("CAMIE_MODEL_NAME", "camie-tagger-v2")
 _MODEL_DIR = Path(os.environ.get("ML_MODEL_DIR", "/models")) / "camie"
+_MODEL_FILE = f"{MODEL_NAME}.onnx"
+_METADATA_FILE = f"{MODEL_NAME}-metadata.json"
 
 # Below this confidence, predictions aren't stored (keeps the JSON compact).
 STORE_FLOOR = float(os.environ.get("TAGGER_STORE_FLOOR", "0.05"))
@@ -39,6 +43,12 @@ STORE_FLOOR = float(os.environ.get("TAGGER_STORE_FLOOR", "0.05"))
 # stored at STORE_FLOOR but artist never surfaces.
 SURFACED_CATEGORIES = {"character", "copyright", "general"}
 
+# ImageNet preprocessing constants (per Camie v2 onnx_inference.py).
+_IMAGENET_MEAN = np.array([0.485, 0.456, 0.406], dtype=np.float32)
+_IMAGENET_STD = np.array([0.229, 0.224, 0.225], dtype=np.float32)
+# Square-pad color ≈ ImageNet mean × 255 (matches reference inference).
+_PAD_COLOR = (124, 116, 104)
+
 
 @dataclass(frozen=True)
 class TagPrediction:
@@ -51,34 +61,48 @@ class Tagger:
     def __init__(self, model_dir: Path | None = None):
         self._model_dir = model_dir or _MODEL_DIR
         self._session = None  # onnxruntime.InferenceSession once load()ed
-        self._tag_meta: list[dict] | None = None
+        self._tag_names: list[str] | None = None
+        self._tag_categories: list[str] | None = None
         self._input_name: str | None = None
-        self._output_name: str | None = None
-        self._input_size: int = 448
+        self._input_size: int = 512
 
     def load(self) -> None:
         if self._session is not None:
             return
-        model_path = self._model_dir / "model.onnx"
-        tags_path = self._model_dir / "selected_tags.csv"
+        model_path = self._model_dir / _MODEL_FILE
+        meta_path = self._model_dir / _METADATA_FILE
         if not model_path.is_file():
             raise RuntimeError(
-                f"Camie model.onnx missing at {model_path}. "
+                f"Camie {_MODEL_FILE} missing at {model_path}. "
                 f"Populate /models via the ml-worker downloader."
             )
-        if not tags_path.is_file():
+        if not meta_path.is_file():
             raise RuntimeError(
-                f"Camie selected_tags.csv missing at {tags_path}. "
+                f"Camie {_METADATA_FILE} missing at {meta_path}. "
                 f"Populate /models via the ml-worker downloader."
             )
 
-        tag_meta: list[dict] = []
-        with open(tags_path, newline="") as f:
-            reader = csv.DictReader(f)
-            for row in reader:
-                tag_meta.append(
-                    {"name": row["name"], "category": row["category"]}
-                )
+        with open(meta_path) as f:
+            metadata = json.load(f)
+
+        # Per Camie v2 onnx_inference.py: idx_to_tag is keyed by str(idx);
+        # tag_to_category maps tag_name -> category. Project to two parallel
+        # lists indexed by output position for O(1) lookup in the hot path.
+        ds = metadata["dataset_info"]
+        idx_to_tag = ds["tag_mapping"]["idx_to_tag"]
+        tag_to_category = ds["tag_mapping"]["tag_to_category"]
+        total = ds["total_tags"]
+        names: list[str] = []
+        cats: list[str] = []
+        for i in range(total):
+            name = idx_to_tag.get(str(i), f"unknown-{i}")
+            names.append(name)
+            cats.append(tag_to_category.get(name, "general"))
+
+        # Input size from metadata; fall back to 512 (the v2 default).
+        self._input_size = int(
+            metadata.get("model_info", {}).get("img_size", 512)
+        )
 
         # Lazy import — kept after the file-existence checks so the
         # missing-model RuntimeError still fires first in environments
@@ -89,51 +113,65 @@ class Tagger:
             str(model_path), providers=["CPUExecutionProvider"]
         )
         self._input_name = session.get_inputs()[0].name
-        self._output_name = session.get_outputs()[0].name
-        input_shape = session.get_inputs()[0].shape
-        for dim in input_shape:
-            if isinstance(dim, int) and dim > 1:
-                self._input_size = dim
-                break
         # Assign sentinels last so a partial load isn't observable.
-        self._tag_meta = tag_meta
+        self._tag_names = names
+        self._tag_categories = cats
         self._session = session
 
     def _preprocess(self, image_path: Path) -> np.ndarray:
         img = Image.open(image_path)
-        # Camie handles RGBA natively but we still composite onto white so
-        # transparency doesn't bias the model (same as IR's WD14 path).
-        if img.mode != "RGBA":
-            img = img.convert("RGBA")
-        bg = Image.new("RGBA", img.size, (255, 255, 255, 255))
-        bg.paste(img, mask=img.split()[3])
-        img = bg.convert("RGB")
+        # Composite RGBA onto neutral so transparency doesn't bias the model.
+        if img.mode == "RGBA":
+            bg = Image.new("RGBA", img.size, (255, 255, 255, 255))
+            bg.paste(img, mask=img.split()[3])
+            img = bg.convert("RGB")
+        elif img.mode != "RGB":
+            img = img.convert("RGB")
 
+        # Pad to square with ImageNet-mean color, then bicubic resize.
         w, h = img.size
         side = max(w, h)
-        square = Image.new("RGB", (side, side), (255, 255, 255))
+        square = Image.new("RGB", (side, side), _PAD_COLOR)
         square.paste(img, ((side - w) // 2, (side - h) // 2))
         square = square.resize(
             (self._input_size, self._input_size), Image.BICUBIC
         )
-        arr = np.array(square, dtype=np.float32)
-        return arr[np.newaxis, :, :, :]  # NHWC
+
+        arr = np.array(square, dtype=np.float32) / 255.0   # HWC, [0,1]
+        arr = (arr - _IMAGENET_MEAN) / _IMAGENET_STD        # ImageNet normalize
+        arr = arr.transpose(2, 0, 1)                        # HWC -> CHW
+        return arr[np.newaxis, :, :, :]                     # NCHW
 
     def infer(self, image_path: Path) -> dict[str, TagPrediction]:
-        """Run Camie on one image. Returns {name: TagPrediction}, only
-        entries with confidence >= STORE_FLOOR (across all categories —
-        the suggestion service does category filtering later)."""
+        """Run Camie v2 on one image. Returns {name: TagPrediction} with
+        confidence >= STORE_FLOOR (across all categories — the suggestion
+        service does category filtering later).
+
+        v2 emits multiple outputs; we use the refined predictions
+        (output[1] per onnx_inference.py). Sigmoid is applied to raw
+        logits to produce [0,1] confidence scores.
+        """
         self.load()
         x = self._preprocess(image_path)
-        out = self._session.run([self._output_name], {self._input_name: x})[0][0]
+        outputs = self._session.run(None, {self._input_name: x})
+        # Refined predictions if present (v2 emits initial + refined),
+        # fall back to initial for single-output forks.
+        logits = outputs[1] if len(outputs) > 1 else outputs[0]
+        # Squeeze batch dim, apply sigmoid.
+        probs = 1.0 / (1.0 + np.exp(-logits[0]))
         results: dict[str, TagPrediction] = {}
-        for idx, score in enumerate(out):
+        names = self._tag_names
+        cats = self._tag_categories
+        for idx, score in enumerate(probs):
             conf = float(score)
             if conf < STORE_FLOOR:
                 continue
-            meta = self._tag_meta[idx]
-            results[meta["name"]] = TagPrediction(
-                name=meta["name"], category=meta["category"], confidence=conf
+            if idx >= len(names):
+                # Output longer than metadata declared — shouldn't happen but
+                # don't crash the import pipeline if v2 metadata desynchronizes.
+                continue
+            results[names[idx]] = TagPrediction(
+                name=names[idx], category=cats[idx], confidence=conf
             )
         return results
 

backend/app/tasks/import_file.py

@@ -29,10 +29,12 @@ IMAGES_ROOT = Path("/images")
 def _map_result_to_status(result):
     """(ImportTask.status, should_requeue_ml_and_thumb) for an ImportResult.
     'superseded' = the kept row's file/ML changed → complete + re-derive.
-    'attached' = a non-art file preserved → complete, no ML/thumb."""
+    'attached' = a non-art file preserved → complete, no ML/thumb.
+    'refreshed' = deep scan refreshed sidecar/phash on an existing row →
+    complete, no ML/thumb re-derive (file/pixels unchanged)."""
     if result.status in ("imported", "superseded"):
         return ("complete", True)
-    if result.status == "attached":
+    if result.status in ("attached", "refreshed"):
         return ("complete", False)
     if result.status == "skipped":
         return ("skipped", False)
@@ -138,6 +140,15 @@ def _do_import(session, task, import_task_id: int) -> dict:
         task.result_image_id = result.image_id
         counter_col_name = "imported"
         counter_col = ImportBatch.imported
+    elif result.status == "refreshed":
+        # Deep-scan rederive: existing row got phash/artist/sidecar
+        # refreshed. Task is complete (no further work), but counted in
+        # `refreshed` not `imported` so the UI can surface the actual
+        # work done. operator-flagged 2026-05-25.
+        task.status = "complete"
+        task.result_image_id = result.image_id
+        counter_col_name = "refreshed"
+        counter_col = ImportBatch.refreshed
     elif result.status == "attached":
         task.status = "complete"
         counter_col_name = "attachments"

backend/app/tasks/maintenance.py

@@ -16,6 +16,7 @@ from ._sync_engine import sync_session_factory as _sync_session_factory
 log = logging.getLogger(__name__)
 
 STUCK_THRESHOLD_MINUTES = 5
+ORPHAN_PENDING_THRESHOLD_MINUTES = 30
 OLD_TASK_DAYS = 7
 PHASH_PAGE = 500
 VERIFY_PAGE = 200
@@ -26,40 +27,77 @@ TASK_RUN_KEEP_FAILURE_SECONDS = 7 * 24 * 3600  # 7 days
 
 @celery.task(name="backend.app.tasks.maintenance.recover_interrupted_tasks")
 def recover_interrupted_tasks() -> int:
-    """Find ImportTask rows stuck in 'processing' for >5 min and re-queue them.
+    """Recover stuck ImportTask rows. Two distinct stuck states:
 
-    Why 5 min: import_media_file is sub-second for the vast majority of
-    files; even a large-video transcode caps at the per-task soft_time_limit
-    (5 min) defined on the task itself. Anything still 'processing' after
-    that window is a confirmed crash (worker died, DB disconnect mid-flush,
-    OOM) and must be recycled. Was 30 min historically; tightened
-    2026-05-24 after operator hit a 2224-row zombie pile during the IR
-    migration scan.
+    1. 'processing' > 5 min — worker crash mid-import. Re-queue via
+       .delay() and let the import retry. Was 30 min historically;
+       tightened 2026-05-24 after operator hit a 2224-row zombie pile.
+       import_media_file is sub-second for the vast majority of files and
+       capped at the per-task soft_time_limit (5 min), so anything still
+       'processing' after that window is a confirmed crash.
+
+    2. 'pending' or 'queued' > 30 min — enqueue-phase crash. scan_directory
+       creates rows with status='pending' (commit), then in a second pass
+       transitions to 'queued' and calls .delay() (commit). If the scanner
+       crashes between those two commits, rows are orphaned in 'pending'
+       (never enqueued) with no recovery path — invisible to the
+       'processing' sweep above. Flagged 2026-05-25 by operator hitting a
+       5490-row orphan pile. Flip these to 'failed' (not re-enqueue) so
+       the operator drains them via /api/import/retry-failed at their own
+       pace; bulk-re-enqueueing 5000+ rows would thundering-herd the
+       import worker.
+
+    Returns total rows touched (recovered + marked failed).
     """
     SessionLocal = _sync_session_factory()
-    cutoff = datetime.now(UTC) - timedelta(minutes=STUCK_THRESHOLD_MINUTES)
+    now = datetime.now(UTC)
+    processing_cutoff = now - timedelta(minutes=STUCK_THRESHOLD_MINUTES)
+    orphan_cutoff = now - timedelta(minutes=ORPHAN_PENDING_THRESHOLD_MINUTES)
     with SessionLocal() as session:
         stuck_ids = session.execute(
             select(ImportTask.id)
             .where(ImportTask.status == "processing")
-            .where(ImportTask.started_at < cutoff)
+            .where(ImportTask.started_at < processing_cutoff)
         ).scalars().all()
 
-        if not stuck_ids:
+        orphan_ids = session.execute(
+            select(ImportTask.id)
+            .where(ImportTask.status.in_(["pending", "queued"]))
+            .where(ImportTask.created_at < orphan_cutoff)
+        ).scalars().all()
+
+        if not stuck_ids and not orphan_ids:
             return 0
 
-        session.execute(
-            update(ImportTask)
-            .where(ImportTask.id.in_(stuck_ids))
-            .values(status="queued", started_at=None, error="recovered from stuck state")
-        )
+        if stuck_ids:
+            session.execute(
+                update(ImportTask)
+                .where(ImportTask.id.in_(stuck_ids))
+                .values(status="queued", started_at=None, error="recovered from stuck state")
+            )
+
+        if orphan_ids:
+            session.execute(
+                update(ImportTask)
+                .where(ImportTask.id.in_(orphan_ids))
+                .values(
+                    status="failed",
+                    error=(
+                        "orphan pending/queued swept by recover_interrupted_tasks "
+                        "(scanner likely crashed mid-enqueue); retry via "
+                        "/api/import/retry-failed"
+                    ),
+                )
+            )
+
         session.commit()
 
-        from .import_file import import_media_file
-        for tid in stuck_ids:
-            import_media_file.delay(tid)
+        if stuck_ids:
+            from .import_file import import_media_file
+            for tid in stuck_ids:
+                import_media_file.delay(tid)
 
-        return len(stuck_ids)
+        return len(stuck_ids) + len(orphan_ids)
 
 
 @celery.task(name="backend.app.tasks.maintenance.cleanup_old_tasks")

backend/app/tasks/scan.py

@@ -59,16 +59,25 @@ def scan_directory(self, triggered_by: str = "manual",
         session.flush()
         batch_id = batch.id
 
-        # Skip-set: any source_path that already has a non-failed ImportTask
-        # row. Re-running scan_directory must not re-enqueue files the
-        # importer has already handled (or is currently handling); doing so
-        # creates duplicate work and inflates the queue. Failed prior tasks
-        # are eligible for retry.
+        # Skip-set behavior splits by mode (operator-flagged 2026-05-25):
+        #
+        # quick: any non-failed prior ImportTask (active OR finished) is
+        #   skipped — quick scan only does new-file enqueue, so re-touching
+        #   already-imported files is wasted work.
+        #
+        # deep: ONLY currently-in-flight tasks (pending/queued/processing)
+        #   are skipped. Completed and skipped tasks ARE re-queued because
+        #   deep scan exists precisely to re-touch already-imported files
+        #   (refresh sidecar metadata, fill NULL phash, fill NULL artist
+        #   via Importer._deep_rederive). Matches IR's deep-scan behavior.
+        active_statuses = ["pending", "queued", "processing"]
+        if mode == "deep":
+            skip_statuses = active_statuses
+        else:
+            skip_statuses = active_statuses + ["complete", "skipped"]
         non_failed_existing = set(session.execute(
             select(ImportTask.source_path).where(
-                ImportTask.status.in_(
-                    ["pending", "queued", "processing", "complete", "skipped"]
-                ),
+                ImportTask.status.in_(skip_statuses),
             )
         ).scalars().all())
 

extension/manifest.json

@@ -1,7 +1,7 @@
 {
   "manifest_version": 3,
   "name": "FabledCurator",
-  "version": "1.0.0",
+  "version": "1.0.2",
   "description": "Export cookies from supported platforms to FabledCurator and add creators as sources in one click.",
 
   "browser_specific_settings": {

extension/package.json

@@ -1,13 +1,13 @@
 {
   "name": "fabledcurator-extension",
-  "version": "1.0.0",
+  "version": "1.0.2",
   "private": true,
   "description": "Firefox extension for FabledCurator",
   "scripts": {
-    "lint":  "web-ext lint --source-dir=.",
-    "start": "web-ext run --source-dir=. --firefox=firefox",
-    "build": "web-ext build --source-dir=. --overwrite-dest",
-    "sign":  "web-ext sign --source-dir=. --channel=unlisted --api-key=$WEB_EXT_API_KEY --api-secret=$WEB_EXT_API_SECRET"
+    "lint":  "web-ext lint --source-dir=. --no-config-discovery --ignore-files package.json package-lock.json web-ext-artifacts node_modules README.md .gitignore",
+    "start": "web-ext run --source-dir=. --no-config-discovery --ignore-files package.json package-lock.json web-ext-artifacts node_modules README.md .gitignore --firefox=firefox",
+    "build": "web-ext build --source-dir=. --no-config-discovery --ignore-files package.json package-lock.json web-ext-artifacts node_modules README.md .gitignore --overwrite-dest",
+    "sign":  "web-ext sign --source-dir=. --no-config-discovery --ignore-files package.json package-lock.json web-ext-artifacts node_modules README.md .gitignore --channel=unlisted --api-key=$WEB_EXT_API_KEY --api-secret=$WEB_EXT_API_SECRET"
   },
   "devDependencies": {
     "web-ext": "^8.0.0"

extension/web-ext-config.cjs

@@ -1,13 +0,0 @@
-module.exports = {
-  sourceDir: '.',
-  artifactsDir: './web-ext-artifacts',
-  ignoreFiles: [
-    'package.json',
-    'package-lock.json',
-    'web-ext-config.cjs',
-    'web-ext-artifacts',
-    'node_modules',
-    'README.md',
-    '.gitignore',
-  ],
-};

frontend/src/components/discovery/ArtistCard.vue

@@ -38,9 +38,18 @@ function onCardClick() {
 .fc-artistcard { cursor: pointer; }
 .fc-artistcard__previews {
   display: grid; grid-template-columns: repeat(3, 1fr);
-  gap: 2px; aspect-ratio: 3 / 1; background: rgb(var(--v-theme-surface-light));
+  gap: 2px; aspect-ratio: 3 / 1;
+  /* Explicit floor + ceiling so tall source images can't escape the
+     preview slot even on browsers where aspect-ratio doesn't compute. */
+  min-height: 150px; max-height: 220px;
+  overflow: hidden;
+  background: rgb(var(--v-theme-surface-light));
+}
+.fc-artistcard__previews img {
+  display: block;
+  width: 100%; height: 100%;
+  object-fit: cover; object-position: center;
 }
-.fc-artistcard__previews img { width: 100%; height: 100%; object-fit: cover; }
 .fc-artistcard__noimg {
   grid-column: 1 / -1; display: flex; align-items: center;
   justify-content: center;

frontend/src/components/discovery/TagCard.vue

@@ -106,9 +106,19 @@ function submit() {
 .fc-tagcard { cursor: pointer; }
 .fc-tagcard__previews {
   display: grid; grid-template-columns: repeat(3, 1fr);
-  gap: 2px; aspect-ratio: 3 / 1; background: rgb(var(--v-theme-surface-light));
+  gap: 2px; aspect-ratio: 3 / 1;
+  /* Explicit floor + ceiling so tall source images can't escape the
+     preview slot even on browsers where aspect-ratio doesn't compute
+     (older Safari, embedded webviews). */
+  min-height: 150px; max-height: 220px;
+  overflow: hidden;
+  background: rgb(var(--v-theme-surface-light));
+}
+.fc-tagcard__previews img {
+  display: block;
+  width: 100%; height: 100%;
+  object-fit: cover; object-position: center;
 }
-.fc-tagcard__previews img { width: 100%; height: 100%; object-fit: cover; }
 .fc-tagcard__noimg {
   grid-column: 1 / -1; display: flex; align-items: center;
   justify-content: center;

frontend/src/components/settings/ImportTaskList.vue

@@ -17,6 +17,12 @@
       >
         Retry failed
       </v-btn>
+      <v-btn
+        variant="text" rounded="pill" size="small" color="warning"
+        :disabled="!hasStuck" @click="onClearStuckOpen"
+      >
+        Clear stuck…
+      </v-btn>
       <v-btn
         variant="text" rounded="pill" size="small" color="error"
         @click="onClearOpen"
@@ -69,6 +75,31 @@
         </v-card-actions>
       </v-card>
     </v-dialog>
+
+    <v-dialog v-model="clearStuckDialog" max-width="480">
+      <v-card>
+        <v-card-title>Clear stuck tasks</v-card-title>
+        <v-card-text>
+          <v-alert type="warning" variant="tonal" density="compact" class="mb-3">
+            Force every <strong>pending / queued / processing</strong> task to
+            <strong>failed</strong> and finalize any active batch that
+            has no remaining work. Use this when the automatic recovery
+            sweep keeps re-queueing the same row (e.g., corrupt file in
+            an autoretry loop, or worker model missing).
+          </v-alert>
+          <p class="text-body-2">
+            Tasks remain in the database with status=<code>failed</code>;
+            click <em>Retry failed</em> once the underlying cause is
+            resolved to re-queue them.
+          </p>
+        </v-card-text>
+        <v-card-actions>
+          <v-spacer />
+          <v-btn @click="clearStuckDialog = false">Cancel</v-btn>
+          <v-btn color="warning" rounded="pill" @click="onClearStuckConfirm">Clear stuck</v-btn>
+        </v-card-actions>
+      </v-card>
+    </v-dialog>
   </v-card>
 </template>
 
@@ -80,6 +111,7 @@ const store = useImportStore()
 const statusFilter = ref(null)
 const clearDialog = ref(false)
 const clearAgeDays = ref(7)
+const clearStuckDialog = ref(false)
 
 const statusOptions = [
   { title: 'All', value: null },
@@ -100,6 +132,9 @@ const headers = [
 ]
 
 const hasFailed = computed(() => store.tasks.some(t => t.status === 'failed'))
+const hasStuck = computed(() => store.tasks.some(
+  t => t.status === 'pending' || t.status === 'queued' || t.status === 'processing'
+))
 
 function statusColor(s) {
   return {
@@ -138,4 +173,9 @@ async function onClearConfirm() {
   await store.clearCompleted(clearAgeDays.value)
   clearDialog.value = false
 }
+function onClearStuckOpen() { clearStuckDialog.value = true }
+async function onClearStuckConfirm() {
+  await store.clearStuck()
+  clearStuckDialog.value = false
+}
 </script>

frontend/src/components/settings/ImportTriggerPanel.vue

@@ -2,31 +2,71 @@
   <v-card>
     <v-card-title>Trigger scan</v-card-title>
     <v-card-text>
-      <div v-if="store.activeBatch" class="d-flex align-center" style="gap: 12px;">
+      <div v-if="store.activeBatch" class="d-flex align-center mb-3" style="gap: 12px;">
         <v-progress-circular
           indeterminate color="accent" size="20"
         />
         <span>
-          Scanning {{ store.activeBatch.source_path }} —
+          {{ store.activeBatch.scan_mode === 'deep' ? 'Deep scanning' : 'Scanning' }}
+          {{ store.activeBatch.source_path || '/import' }} —
           imported {{ store.activeBatch.imported }},
+          <template v-if="store.activeBatch.scan_mode === 'deep'">
+            refreshed {{ store.activeBatch.refreshed || 0 }},
+          </template>
           skipped {{ store.activeBatch.skipped }},
           failed {{ store.activeBatch.failed }} /
           {{ store.activeBatch.total_files }} files
         </span>
+        <v-spacer />
+        <v-btn
+          variant="text" rounded="pill" size="small" color="warning"
+          :loading="clearing" @click="onClearStuck"
+        >
+          Clear stuck
+        </v-btn>
       </div>
-      <div v-else>
-        <p class="text-body-2 mb-3">
-          Run a quick scan of the import directory. Deep scan (pHash dedup,
-          archives) lands in FC-2d.
-        </p>
-        <v-btn color="primary" rounded="pill" @click="trigger" :loading="busy">
+
+      <p class="text-body-2 mb-3">
+        <span v-if="!store.activeBatch">
+          <strong>Quick scan</strong> walks <code>/import</code> and enqueues
+          new files only.
+          <strong>Deep scan</strong> additionally re-walks already-imported
+          files so updated sidecar metadata (post title/date/attribution) and
+          previously-NULL phashes / artist links get refreshed. Use after
+          bulk-downloading fresh sidecars for existing content. Both modes
+          route non-media + sidecar pairs through PostAttachment capture.
+        </span>
+        <span v-else>
+          An active batch is in progress. Wait for it to finish, or click
+          <em>Clear stuck</em> above if it has been wedged with no
+          measurable progress.
+        </span>
+      </p>
+
+      <div class="d-flex flex-wrap" style="gap: 12px;">
+        <v-btn
+          color="primary" rounded="pill"
+          :disabled="!!store.activeBatch"
+          :loading="busy === 'quick'"
+          @click="trigger('quick')"
+        >
           <v-icon start>mdi-magnify-scan</v-icon>
           Quick scan
         </v-btn>
-        <v-alert v-if="store.triggerError" type="error" variant="tonal" class="mt-3" closable>
-          {{ store.triggerError }}
-        </v-alert>
+        <v-btn
+          color="secondary" rounded="pill" variant="tonal"
+          :disabled="!!store.activeBatch"
+          :loading="busy === 'deep'"
+          @click="trigger('deep')"
+        >
+          <v-icon start>mdi-magnify-plus-outline</v-icon>
+          Deep scan
+        </v-btn>
       </div>
+
+      <v-alert v-if="store.triggerError" type="error" variant="tonal" class="mt-3" closable>
+        {{ store.triggerError }}
+      </v-alert>
     </v-card-text>
   </v-card>
 </template>
@@ -36,10 +76,22 @@ import { ref } from 'vue'
 import { useImportStore } from '../../stores/import.js'
 
 const store = useImportStore()
-const busy = ref(false)
+const busy = ref(null)
+const clearing = ref(false)
 
-async function trigger() {
-  busy.value = true
-  try { await store.triggerScan() } catch {} finally { busy.value = false }
+async function trigger(mode) {
+  busy.value = mode
+  try { await store.triggerScan(mode) } catch {} finally { busy.value = null }
+}
+
+async function onClearStuck() {
+  clearing.value = true
+  try {
+    await store.clearStuck()
+  } catch {
+    // store surfaces error via triggerError if needed
+  } finally {
+    clearing.value = false
+  }
 }
 </script>

frontend/src/components/settings/MaintenancePanel.vue

@@ -14,7 +14,6 @@
     <AliasTable class="mt-4" />
     <BackupCard class="mt-6" />
     <TagMaintenanceCard class="mt-6" />
-    <BrowserExtensionCard class="mt-6" />
     <LegacyMigrationCard class="mt-6" />
   </div>
 </template>
@@ -27,7 +26,6 @@ import AllowlistTable from './AllowlistTable.vue'
 import AliasTable from './AliasTable.vue'
 import BackupCard from './BackupCard.vue'
 import TagMaintenanceCard from './TagMaintenanceCard.vue'
-import BrowserExtensionCard from './BrowserExtensionCard.vue'
 import LegacyMigrationCard from './LegacyMigrationCard.vue'
 </script>
 

frontend/src/stores/import.js

@@ -52,11 +52,58 @@ export const useImportStore = defineStore('import', () => {
     }
   }
 
-  async function triggerScan() {
+  async function triggerScan(mode = 'quick') {
+    if (!['quick', 'deep', 'verify'].includes(mode)) {
+      throw new Error(`unsupported scan mode: ${mode}`)
+    }
     triggerError.value = null
     try {
-      await api.post('/api/import/trigger', { body: { mode: 'quick' } })
+      await api.post('/api/import/trigger', { body: { mode } })
+      // Acknowledge immediately so the click isn't invisible. scan_directory
+      // can finalize the batch synchronously when every file in /import is
+      // already on a non-failed ImportTask (operator-flagged 2026-05-25:
+      // 233k existing tasks → all paths in skip-set → files_seen=0 →
+      // batch flashes 'running' for <100ms then 'complete' before the
+      // first refreshStatus() lands; UI never sees the active state).
+      const label = mode === 'deep'
+        ? 'Deep scan triggered (re-applying sidecar metadata + filling NULL phash/artist on existing rows)'
+        : mode === 'verify' ? 'Library verify triggered' : 'Quick scan triggered'
+      window.__fcToast?.({ text: label, type: 'success' })
       await refreshStatus()
+      // Re-poll twice over ~5s and produce an HONEST follow-up toast.
+      // Operator-flagged 2026-05-25: the prior "no new files" message was
+      // misleading because deep scan IS doing work (refresh) even when
+      // there are no new files to import. Surface the real workload count
+      // (imported + refreshed + queued) instead. For quick scan + zero
+      // queued work, fall back to "up to date" instead of the old
+      // implementation-detail-leaking message.
+      setTimeout(async () => {
+        await refreshStatus()
+        await loadTasks(true)
+        if (activeBatch.value || mode === 'verify') return
+        // Batch finalized quickly; figure out what actually happened.
+        // The task list was just refreshed; the freshest row(s) carry
+        // the batch outcome.
+        const batchId = tasks.value[0]?.batch_id
+        const sameBatch = batchId
+          ? tasks.value.filter(t => t.batch_id === batchId)
+          : []
+        const refreshedCount = sameBatch.filter(t => t.status === 'complete' && t.result_image_id).length
+        const newImported = sameBatch.filter(t => t.status === 'complete' && t.result_image_id && !t.error).length
+        if (mode === 'deep' && sameBatch.length > 0) {
+          window.__fcToast?.({
+            text: `Deep scan finished — ${sameBatch.length} file(s) processed`,
+            type: 'info',
+          })
+        } else if (mode === 'quick' && sameBatch.length > 0) {
+          window.__fcToast?.({
+            text: `Quick scan finished — ${newImported} new file(s) queued`,
+            type: 'info',
+          })
+        } else {
+          window.__fcToast?.({ text: 'Library is up to date', type: 'info' })
+        }
+      }, 2000)
     } catch (e) {
       triggerError.value = e.message
       window.__fcToast?.({ text: `Scan failed: ${e.message}`, type: 'error' })
@@ -92,6 +139,13 @@ export const useImportStore = defineStore('import', () => {
     await loadTasks(true)
   }
 
+  async function clearStuck() {
+    const body = await api.post('/api/import/clear-stuck')
+    await loadTasks(true)
+    await refreshStatus()
+    return body
+  }
+
   const hasMore = computed(() => tasksNextCursor.value !== null)
 
   return {
@@ -101,6 +155,6 @@ export const useImportStore = defineStore('import', () => {
     triggerError,
     loadSettings, patchSettings,
     refreshStatus, triggerScan,
-    loadTasks, setStatusFilter, retryFailed, clearCompleted
+    loadTasks, setStatusFilter, retryFailed, clearCompleted, clearStuck
   }
 })

frontend/src/views/ArtistView.vue

@@ -45,69 +45,92 @@
         >Credential health · FC-3b</v-chip>
       </div>
 
-      <section v-if="store.overview.cooccurring_tags.length" class="fc-artist__sec">
-        <h2 class="fc-h2">Frequent tags</h2>
-        <div class="fc-artist__tags">
-          <v-chip
-            v-for="t in store.overview.cooccurring_tags" :key="t.id"
-            size="small" @click="openTag(t.id)"
-          >{{ t.name }} <span class="fc-artist__tagc">{{ t.count }}</span></v-chip>
-        </div>
-      </section>
+      <!-- Tabs split (2026-05-25): Settings was previously slotted at the
+           bottom of the page after the infinite-scroll image grid, which
+           made it effectively unreachable for any artist with more than
+           a couple of pages of content. The Settings tab now hosts
+           destructive admin actions (artist+content cascade-delete) and
+           any future per-artist management UI. v-tabs is `position:
+           sticky; top: 64px` (under the 64px AppShell TopNav) so it
+           stays parked while the gallery scrolls. -->
+      <v-tabs
+        v-model="tab" color="accent" class="mb-4"
+        style="position: sticky; top: 64px; z-index: 4;
+               background: rgb(var(--v-theme-surface));"
+      >
+        <v-tab value="overview">Overview</v-tab>
+        <v-tab value="settings">Settings</v-tab>
+      </v-tabs>
 
-      <section v-if="store.overview.activity.length" class="fc-artist__sec">
-        <h2 class="fc-h2">Activity</h2>
-        <svg class="fc-artist__spark" :viewBox="`0 0 ${sparkW} ${sparkH}`"
-             preserveAspectRatio="none" role="img" aria-label="posts over time">
-          <polyline :points="sparkPoints" fill="none"
-                    stroke="rgb(var(--v-theme-accent))" stroke-width="2" />
-        </svg>
-      </section>
+      <v-window v-model="tab">
+        <v-window-item value="overview">
+          <section v-if="store.overview.cooccurring_tags.length" class="fc-artist__sec">
+            <h2 class="fc-h2">Frequent tags</h2>
+            <div class="fc-artist__tags">
+              <v-chip
+                v-for="t in store.overview.cooccurring_tags" :key="t.id"
+                size="small" @click="openTag(t.id)"
+              >{{ t.name }} <span class="fc-artist__tagc">{{ t.count }}</span></v-chip>
+            </div>
+          </section>
 
-      <section v-if="store.overview.sources.length" class="fc-artist__sec">
-        <div class="fc-artist__sec-head">
-          <h2 class="fc-h2">Sources</h2>
-          <RouterLink
-            :to="`/subscriptions?artist_id=${store.overview.id}`"
-            class="fc-artist__manage"
-          >Manage subscriptions →</RouterLink>
-        </div>
-        <v-table density="compact">
-          <thead>
-            <tr><th>Platform</th><th>URL</th><th class="text-right">Images</th></tr>
-          </thead>
-          <tbody>
-            <tr v-for="s in store.overview.sources" :key="s.id">
-              <td>{{ s.platform }}</td>
-              <td class="fc-artist__url">{{ s.url }}</td>
-              <td class="text-right">{{ s.image_count }}</td>
-            </tr>
-          </tbody>
-        </v-table>
-      </section>
+          <section v-if="store.overview.activity.length" class="fc-artist__sec">
+            <h2 class="fc-h2">Activity</h2>
+            <svg class="fc-artist__spark" :viewBox="`0 0 ${sparkW} ${sparkH}`"
+                 preserveAspectRatio="none" role="img" aria-label="posts over time">
+              <polyline :points="sparkPoints" fill="none"
+                        stroke="rgb(var(--v-theme-accent))" stroke-width="2" />
+            </svg>
+          </section>
 
-      <section class="fc-artist__sec">
-        <h2 class="fc-h2">Images</h2>
-        <MasonryGrid
-          :items="store.images"
-          :loading="store.imagesLoading"
-          :has-more="store.hasMoreImages"
-          @load-more="store.loadMoreImages(slug)"
-          @open="openImage"
-        />
-      </section>
+          <section v-if="store.overview.sources.length" class="fc-artist__sec">
+            <div class="fc-artist__sec-head">
+              <h2 class="fc-h2">Sources</h2>
+              <RouterLink
+                :to="`/subscriptions?artist_id=${store.overview.id}`"
+                class="fc-artist__manage"
+              >Manage subscriptions →</RouterLink>
+            </div>
+            <v-table density="compact">
+              <thead>
+                <tr><th>Platform</th><th>URL</th><th class="text-right">Images</th></tr>
+              </thead>
+              <tbody>
+                <tr v-for="s in store.overview.sources" :key="s.id">
+                  <td>{{ s.platform }}</td>
+                  <td class="fc-artist__url">{{ s.url }}</td>
+                  <td class="text-right">{{ s.image_count }}</td>
+                </tr>
+              </tbody>
+            </v-table>
+          </section>
 
-      <ArtistDangerZone
-        :slug="slug"
-        :artist-id="store.overview.id"
-        :artist-name="store.overview.name"
-      />
+          <section class="fc-artist__sec">
+            <h2 class="fc-h2">Images</h2>
+            <MasonryGrid
+              :items="store.images"
+              :loading="store.imagesLoading"
+              :has-more="store.hasMoreImages"
+              @load-more="store.loadMoreImages(slug)"
+              @open="openImage"
+            />
+          </section>
+        </v-window-item>
+
+        <v-window-item value="settings">
+          <ArtistDangerZone
+            :slug="slug"
+            :artist-id="store.overview.id"
+            :artist-name="store.overview.name"
+          />
+        </v-window-item>
+      </v-window>
     </template>
   </v-container>
 </template>
 
 <script setup>
-import { computed, watch } from 'vue'
+import { computed, ref, watch } from 'vue'
 import { useRoute, useRouter, RouterLink } from 'vue-router'
 import { useArtistStore } from '../stores/artist.js'
 import { useModalStore } from '../stores/modal.js'
@@ -120,8 +143,18 @@ const store = useArtistStore()
 const modal = useModalStore()
 
 const slug = computed(() => route.params.slug)
+// Per-artist tab — defaults to Overview. Settings tab hosts destructive
+// admin actions (DangerZone). Switching artists resets to Overview so the
+// destructive surface isn't re-shown by accident when navigating between
+// artists.
+const tab = ref('overview')
 
-watch(slug, (s) => { if (s) store.load(s) }, { immediate: true })
+watch(slug, (s) => {
+  if (s) {
+    store.load(s)
+    tab.value = 'overview'
+  }
+}, { immediate: true })
 
 const dateRange = computed(() => {
   const r = store.overview?.date_range

frontend/src/views/ArtistsView.vue

@@ -84,7 +84,7 @@ onUnmounted(() => observer && observer.disconnect())
 }
 .fc-artists__grid {
   display: grid;
-  grid-template-columns: repeat(auto-fill, minmax(220px, 1fr));
+  grid-template-columns: repeat(auto-fill, minmax(440px, 1fr));
   gap: 12px;
 }
 .fc-artists__sentinel {

frontend/src/views/SettingsView.vue

@@ -1,6 +1,16 @@
 <template>
   <v-container fluid class="py-6">
-    <v-tabs v-model="tab" color="accent" class="mb-4">
+    <!-- Sticky tabs: operator-flagged 2026-05-25 — long Import / Maintenance
+         panels pushed the tab strip out of the viewport, forcing a scroll-
+         to-top just to change tab. AppShell's TopNav is 64px sticky, so the
+         tab strip lives directly under it. Background uses the theme surface
+         token so it visually merges with the page rather than the
+         translucent v-tabs default. -->
+    <v-tabs
+      v-model="tab" color="accent" class="mb-4"
+      style="position: sticky; top: 64px; z-index: 4;
+             background: rgb(var(--v-theme-surface));"
+    >
       <v-tab value="overview">Overview</v-tab>
       <v-tab value="activity">Activity</v-tab>
       <v-tab value="import">Import</v-tab>
@@ -21,6 +31,11 @@
           {{ system.stats.tasks.pending + system.stats.tasks.queued }} import task(s) pending.
           <v-btn variant="text" size="small" @click="tab = 'import'">Go to Import tab</v-btn>
         </v-alert>
+        <!-- Browser-extension install/download lives on Overview (moved
+             from Maintenance 2026-05-25). Overview is the discovery
+             surface for "things to set up"; Maintenance is for
+             housekeeping of already-set-up systems. -->
+        <BrowserExtensionCard class="mt-6" />
       </v-window-item>
 
       <v-window-item value="activity">
@@ -28,11 +43,14 @@
       </v-window-item>
 
       <v-window-item value="import">
+        <!-- Order: trigger → recent tasks → filters. Tasks sit directly
+             below the trigger so operator sees hit/miss feedback without
+             scrolling past the filter card (operator-flagged 2026-05-25). -->
         <ImportTriggerPanel />
         <v-divider class="my-6" />
-        <ImportFiltersForm />
-        <v-divider class="my-6" />
         <ImportTaskList />
+        <v-divider class="my-6" />
+        <ImportFiltersForm />
       </v-window-item>
 
       <v-window-item value="maintenance">
@@ -49,6 +67,7 @@ import { useImportStore } from '../stores/import.js'
 import SystemStatsCards from '../components/settings/SystemStatsCards.vue'
 import SystemActivitySummary from '../components/settings/SystemActivitySummary.vue'
 import SystemActivityTab from '../components/settings/SystemActivityTab.vue'
+import BrowserExtensionCard from '../components/settings/BrowserExtensionCard.vue'
 import ImportTriggerPanel from '../components/settings/ImportTriggerPanel.vue'
 import ImportFiltersForm from '../components/settings/ImportFiltersForm.vue'
 import ImportTaskList from '../components/settings/ImportTaskList.vue'

frontend/src/views/TagsView.vue

@@ -236,7 +236,7 @@ async function onDeleteTagConfirm() {
 }
 .fc-tags__grid {
   display: grid;
-  grid-template-columns: repeat(auto-fill, minmax(220px, 1fr));
+  grid-template-columns: repeat(auto-fill, minmax(440px, 1fr));
   gap: 12px;
 }
 .fc-tags__sentinel {

tests/test_api_import_admin.py

@@ -86,6 +86,61 @@ async def test_clear_completed(client, db):
     assert body["deleted"] == 1
 
 
+@pytest.mark.asyncio
+async def test_clear_stuck_fails_non_terminal_and_finalizes_orphan_batch(client, db):
+    """Operator-flagged 2026-05-25: 3 large PNGs got stuck in 'processing'
+    for 2 days, the active ImportBatch never finalized, and the UI's
+    'Scanning...' banner persisted with 0/0 files. /api/import/clear-stuck
+    is the escape hatch to break the autoretry loop manually."""
+    from sqlalchemy import select as _select
+
+    batch = ImportBatch(triggered_by="manual", source_path="/import", scan_mode="quick")
+    db.add(batch)
+    await db.flush()
+    # Three stuck rows in mixed non-terminal states.
+    db.add(ImportTask(
+        batch_id=batch.id, source_path="/p1", task_type="media", status="processing",
+    ))
+    db.add(ImportTask(
+        batch_id=batch.id, source_path="/p2", task_type="media", status="queued",
+    ))
+    db.add(ImportTask(
+        batch_id=batch.id, source_path="/p3", task_type="media", status="pending",
+    ))
+    # One already-complete row should be untouched.
+    db.add(ImportTask(
+        batch_id=batch.id, source_path="/done", task_type="media",
+        status="complete", finished_at=datetime.now(UTC),
+    ))
+    await db.commit()
+
+    resp = await client.post("/api/import/clear-stuck")
+    body = await resp.get_json()
+    assert resp.status_code == 200
+    assert body["tasks_failed"] == 3
+    assert body["batches_finalized"] == 1
+
+    statuses = {
+        row.status for row in
+        (await db.execute(_select(ImportTask).where(ImportTask.batch_id == batch.id)))
+        .scalars().all()
+    }
+    assert statuses == {"failed", "complete"}
+
+    batch_status = (await db.execute(
+        _select(ImportBatch.status).where(ImportBatch.id == batch.id)
+    )).scalar_one()
+    assert batch_status == "complete"
+
+
+@pytest.mark.asyncio
+async def test_clear_stuck_no_op_when_nothing_stuck(client, db):
+    resp = await client.post("/api/import/clear-stuck")
+    body = await resp.get_json()
+    assert resp.status_code == 200
+    assert body == {"tasks_failed": 0, "batches_finalized": 0}
+
+
 @pytest.mark.asyncio
 async def test_trigger_accepts_deep(client, monkeypatch):
     # Stub the task dispatch — assert the API accepts 'deep' and forwards

tests/test_download_models.py

@@ -9,11 +9,16 @@ from backend.app.scripts import download_models as dm
 
 
 def test_ensure_camie_skips_when_present(tmp_path, monkeypatch):
+    """v2 layout (HF Camais03/camie-tagger-v2): the ONNX file is named
+    camie-tagger-v2.onnx (not model.onnx) and tags ship inside
+    camie-tagger-v2-metadata.json (not selected_tags.csv). Both at root.
+    Updated 2026-05-25 after the actual repo layout was confirmed via
+    WebFetch — the old assertion pinned the v1 filenames."""
     monkeypatch.setattr(dm, "MODEL_ROOT", tmp_path)
     camie = tmp_path / "camie"
     camie.mkdir(parents=True)
-    (camie / "model.onnx").write_bytes(b"x")
-    (camie / "selected_tags.csv").write_text("tag_id,name,category,count\n")
+    (camie / "camie-tagger-v2.onnx").write_bytes(b"x")
+    (camie / "camie-tagger-v2-metadata.json").write_text("{}")
     with patch.object(dm, "_snapshot") as snap:
         dm.ensure_camie()
         snap.assert_not_called()

tests/test_gallery_service.py

@@ -133,3 +133,135 @@ async def test_get_image_with_tags_includes_integrity_status(db):
     svc = GalleryService(db)
     payload = await svc.get_image_with_tags(img.id)
     assert payload["integrity_status"] == "ok"
+
+
+async def _seed_image_with_post(
+    db, *, sha: str, image_created_at, post_date, artist_name="test-artist",
+    platform="patreon", external_post_id="42",
+):
+    """Helper: seed an Artist + Source + Post and one ImageRecord whose
+    primary_post_id points at that Post. Used for date-coalesce tests."""
+    from backend.app.models import Artist, Post, Source
+    artist = Artist(name=artist_name, slug=artist_name.lower().replace(" ", "-"))
+    db.add(artist)
+    await db.flush()
+    source = Source(
+        artist_id=artist.id, platform=platform,
+        url=f"https://www.{platform}.com/{artist.slug}",
+    )
+    db.add(source)
+    await db.flush()
+    post = Post(
+        source_id=source.id, external_post_id=external_post_id,
+        post_title="A Post", post_date=post_date,
+    )
+    db.add(post)
+    await db.flush()
+    img = ImageRecord(
+        path=f"/images/test/{sha[:8]}.jpg",
+        sha256=sha, size_bytes=1000, mime="image/jpeg",
+        width=100, height=100,
+        origin="imported_filesystem", integrity_status="unknown",
+        primary_post_id=post.id,
+    )
+    img.created_at = image_created_at
+    db.add(img)
+    await db.flush()
+    return img, post
+
+
+@pytest.mark.asyncio
+async def test_scroll_sorts_by_post_date_when_available(db):
+    """Operator-flagged 2026-05-25: ~57k IR images all imported in the
+    same week sort by image.created_at and pile up in one month bucket.
+    Once primary_post_id is wired (via tag_apply phase 4), the gallery
+    should sort by Post.post_date instead, spreading them across the
+    actual publish years."""
+    base_import = _now()
+    # Image A: imported NOW, but post was made 2 years ago.
+    img_a, _ = await _seed_image_with_post(
+        db, sha="a" * 64,
+        image_created_at=base_import,
+        post_date=base_import - timedelta(days=730),
+        artist_name="Aria", external_post_id="A-1",
+    )
+    # Image B: imported NOW (1 min later), post made YESTERDAY.
+    img_b, _ = await _seed_image_with_post(
+        db, sha="b" * 64,
+        image_created_at=base_import - timedelta(minutes=1),
+        post_date=base_import - timedelta(days=1),
+        artist_name="Bea", external_post_id="B-1",
+    )
+    # Image C: filesystem-imported, no primary_post_id, created 5 days ago.
+    img_c = ImageRecord(
+        path="/images/test/c.jpg", sha256="c" * 64,
+        size_bytes=1000, mime="image/jpeg",
+        width=100, height=100,
+        origin="imported_filesystem", integrity_status="unknown",
+    )
+    img_c.created_at = base_import - timedelta(days=5)
+    db.add(img_c)
+    await db.flush()
+
+    svc = GalleryService(db)
+    page = await svc.scroll(cursor=None, limit=10)
+    # Effective-date order: B (yesterday) > C (5 days ago) > A (2 years ago)
+    assert [i.id for i in page.images] == [img_b.id, img_c.id, img_a.id]
+
+    # API exposes both fields explicitly so the UI can show "Posted X / Imported Y".
+    a_payload = next(i for i in page.images if i.id == img_a.id)
+    assert a_payload.posted_at is not None
+    assert a_payload.posted_at < a_payload.created_at
+    c_payload = next(i for i in page.images if i.id == img_c.id)
+    assert c_payload.posted_at is None
+    assert c_payload.effective_date == c_payload.created_at
+
+
+@pytest.mark.asyncio
+async def test_timeline_buckets_use_post_date_when_available(db):
+    """Timeline group-by must follow the same effective_date rule so the
+    UI's year/month navigation surfaces publish-date buckets, not the
+    single FC-scan bucket all migrated images share."""
+    base = datetime(2026, 6, 15, 12, 0, tzinfo=UTC)
+    await _seed_image_with_post(
+        db, sha="1" * 64,
+        image_created_at=base,
+        post_date=datetime(2024, 3, 10, tzinfo=UTC),
+        artist_name="Carl", external_post_id="C-1",
+    )
+    await _seed_image_with_post(
+        db, sha="2" * 64,
+        image_created_at=base,
+        post_date=datetime(2024, 3, 11, tzinfo=UTC),
+        artist_name="Dee", external_post_id="D-1",
+    )
+    await _seed_image_with_post(
+        db, sha="3" * 64,
+        image_created_at=base,
+        post_date=datetime(2025, 9, 1, tzinfo=UTC),
+        artist_name="Eli", external_post_id="E-1",
+    )
+    svc = GalleryService(db)
+    buckets = await svc.timeline()
+    bucket_keys = {(b.year, b.month, b.count) for b in buckets}
+    # Two posts in 2024-03, one in 2025-09 — even though all imported in 2026-06.
+    assert (2024, 3, 2) in bucket_keys
+    assert (2025, 9, 1) in bucket_keys
+    # The FC-import bucket should NOT appear since all 3 images have post_date.
+    assert not any(b.year == 2026 and b.month == 6 for b in buckets)
+
+
+@pytest.mark.asyncio
+async def test_get_image_with_tags_includes_posted_at_when_present(db):
+    base = _now()
+    img, _ = await _seed_image_with_post(
+        db, sha="f" * 64,
+        image_created_at=base,
+        post_date=base - timedelta(days=365),
+        artist_name="Fred", external_post_id="F-1",
+    )
+    svc = GalleryService(db)
+    payload = await svc.get_image_with_tags(img.id)
+    assert payload["posted_at"] is not None
+    # Image's own created_at is still surfaced separately.
+    assert payload["created_at"] != payload["posted_at"]

tests/test_importer.py

@@ -134,3 +134,58 @@ def test_root_level_file_has_no_artist(importer, import_layout):
     importer.import_one(src)
     artists = importer.session.execute(select(Artist)).scalars().all()
     assert artists == []
+
+
+def test_pil_load_oserror_in_transparency_check_skips_not_raises(
+    importer, import_layout, monkeypatch,
+):
+    """PIL.verify() only validates header structure — broken pixel data
+    only surfaces when load() actually decodes. The importer must catch
+    the OSError and return a skipped: invalid_image result so the Celery
+    autoretry loop doesn't bounce the same broken file forever.
+    Operator hit this 2026-05-25 with a corrupt JPEG in the IR set."""
+    import_root, _ = import_layout
+    src = import_root / "Bob" / "corrupt.png"
+    # Make a real RGBA PNG so the has_alpha path engages.
+    _make_png_rgba(src, (100, 100), alpha=128)
+
+    importer.settings.skip_transparent = True
+    importer.settings.transparency_threshold = 0.5
+
+    # Force the next _transparency_pct call to raise as if PIL's load()
+    # blew up on truncated pixel data.
+    def _boom(_self, _src):
+        raise OSError("broken data stream when reading image file")
+    monkeypatch.setattr(
+        type(importer), "_transparency_pct", _boom,
+    )
+
+    result = importer.import_one(src)
+    assert result.status == "skipped"
+    assert result.skip_reason == SkipReason.invalid_image
+    assert "transparency check" in (result.error or "")
+
+
+def test_pil_load_oserror_in_phash_compute_skips_not_raises(
+    importer, import_layout, monkeypatch,
+):
+    """Same shape as the transparency-check guard, but for the phash
+    compute block — the OTHER place PIL.load() runs implicitly during
+    the dedup pipeline."""
+    import_root, _ = import_layout
+    src = import_root / "Carol" / "corrupt.jpg"
+    _make_jpeg(src)
+
+    # Disable transparency check so we reach the phash compute block.
+    importer.settings.skip_transparent = False
+
+    from backend.app.services import importer as importer_module
+
+    def _boom(_im):
+        raise OSError("broken data stream when reading image file")
+    monkeypatch.setattr(importer_module, "compute_phash", _boom)
+
+    result = importer.import_one(src)
+    assert result.status == "skipped"
+    assert result.skip_reason == SkipReason.invalid_image
+    assert "phash compute" in (result.error or "")

tests/test_importer_attachments.py

@@ -59,3 +59,25 @@ def test_json_sidecar_is_not_attached(importer, import_layout):
         select(func.count()).select_from(PostAttachment)
     ).scalar_one()
     assert n == 0
+
+
+def test_mangled_filename_extension_is_sanitized(importer, import_layout):
+    """gallery-dl sometimes URL-encodes a query string into the basename
+    (`...https___www.patreon.com_media-u_Z0F...`). Python's Path.suffix
+    returns 50+ chars of base64-ish junk for those, which blows the
+    PostAttachment.ext varchar(32) column. Operator-flagged 2026-05-25.
+    The importer should record an empty ext rather than crash."""
+    import_root, _ = import_layout
+    f = (
+        import_root / "Alice"
+        / "79507046_media_https___www.patreon.com_media-u_Z0FBQUFBQm5q"
+    )
+    f.parent.mkdir(parents=True, exist_ok=True)
+    f.write_bytes(b"binary blob")
+    r = importer.import_one(f)
+    assert r.status == "attached"
+    att = importer.session.execute(select(PostAttachment)).scalar_one()
+    # Junk "extension" -> stored as empty string (not the 50-char garbage).
+    assert att.ext == ""
+    # original_filename is Text-typed so the full name survives intact.
+    assert att.original_filename.endswith("_Z0FBQUFBQm5q")

tests/test_importer_deep.py

@@ -64,9 +64,13 @@ def test_deep_rederives_phash_and_provenance(db_sync, import_layout):
 
     deep = _mk(db_sync, import_layout, deep=True)
     r2 = deep.import_one(src)
-    assert r2.status == "skipped"
-    assert "deep" in (r2.error or "")
+    # Outcome flipped from "skipped+duplicate_hash" to "refreshed" 2026-05-25
+    # so the UI can surface deep scan's actual work instead of showing it as
+    # a no-op. See ImportResult.status comment + _deep_rederive docstring.
+    assert r2.status == "refreshed"
     assert r2.image_id == rec.id
+    assert r2.skip_reason is None
+    assert r2.error is None
 
     db_sync.expire_all()
     rec2 = db_sync.get(ImageRecord, rec.id)

tests/test_maintenance.py

@@ -68,6 +68,102 @@ def test_recover_interrupted_only_old(db_sync, monkeypatch):
     assert dispatched == [stale.id]
 
 
+def test_recover_interrupted_sweeps_pending_orphans_to_failed(db_sync, monkeypatch):
+    """A scan that creates ImportTask rows but crashes before the second
+    pass (transition to 'queued' + .delay()) leaves rows orphaned at
+    status='pending'. The sweep flips them to 'failed' so the operator
+    can drain via /api/import/retry-failed without thundering-herding.
+    Banked 2026-05-25 after operator hit 5490 stuck pending rows.
+    """
+    from backend.app.tasks import import_file
+    monkeypatch.setattr(import_file.import_media_file, "delay", lambda *_: None)
+
+    batch_id = _make_batch(db_sync)
+    now = datetime.now(UTC)
+
+    fresh_pending = ImportTask(
+        batch_id=batch_id, source_path="/import/fresh.jpg", task_type="media",
+        status="pending",
+    )
+    db_sync.add(fresh_pending)
+    db_sync.flush()
+    # created_at defaults to now() server-side; fresh row stays untouched.
+
+    # Two stale rows simulating the orphan pile: one 'pending', one
+    # 'queued' (scanner crashed AFTER transitioning some rows but
+    # before all). Both should sweep.
+    stale_pending = ImportTask(
+        batch_id=batch_id, source_path="/import/stale1.jpg", task_type="media",
+        status="pending",
+    )
+    stale_queued = ImportTask(
+        batch_id=batch_id, source_path="/import/stale2.jpg", task_type="media",
+        status="queued",
+    )
+    db_sync.add_all([stale_pending, stale_queued])
+    db_sync.flush()
+    # Backdate created_at past the orphan cutoff (30 min).
+    from sqlalchemy import update as _upd
+    db_sync.execute(
+        _upd(ImportTask)
+        .where(ImportTask.id.in_([stale_pending.id, stale_queued.id]))
+        .values(created_at=now - timedelta(hours=2))
+    )
+    db_sync.commit()
+
+    from backend.app.tasks.maintenance import recover_interrupted_tasks
+    touched = recover_interrupted_tasks.apply().get()
+    assert touched == 2
+
+    db_sync.refresh(fresh_pending)
+    db_sync.refresh(stale_pending)
+    db_sync.refresh(stale_queued)
+    assert fresh_pending.status == "pending"  # fresh row untouched
+    assert stale_pending.status == "failed"
+    assert stale_queued.status == "failed"
+    assert "orphan" in (stale_pending.error or "")
+
+
+def test_recover_interrupted_handles_both_stuck_and_orphans(db_sync, monkeypatch):
+    """One sweep tick handles both 'processing' crashes AND
+    'pending'/'queued' orphans in a single pass."""
+    from backend.app.tasks import import_file
+    dispatched: list[int] = []
+    monkeypatch.setattr(
+        import_file.import_media_file, "delay", dispatched.append
+    )
+
+    batch_id = _make_batch(db_sync)
+    now = datetime.now(UTC)
+
+    stuck = ImportTask(
+        batch_id=batch_id, source_path="/import/stuck.jpg", task_type="media",
+        status="processing", started_at=now - timedelta(hours=2),
+    )
+    orphan = ImportTask(
+        batch_id=batch_id, source_path="/import/orphan.jpg", task_type="media",
+        status="pending",
+    )
+    db_sync.add_all([stuck, orphan])
+    db_sync.flush()
+    from sqlalchemy import update as _upd
+    db_sync.execute(
+        _upd(ImportTask).where(ImportTask.id == orphan.id)
+        .values(created_at=now - timedelta(hours=2))
+    )
+    db_sync.commit()
+
+    from backend.app.tasks.maintenance import recover_interrupted_tasks
+    touched = recover_interrupted_tasks.apply().get()
+    assert touched == 2
+
+    db_sync.refresh(stuck)
+    db_sync.refresh(orphan)
+    assert stuck.status == "queued"
+    assert orphan.status == "failed"
+    assert dispatched == [stuck.id]  # stuck rows re-enqueue; orphans don't
+
+
 def test_cleanup_old_deletes_finished_old(db_sync):
     batch_id = _make_batch(db_sync)
     now = datetime.now(UTC)

tests/test_ml_tagger.py

@@ -40,5 +40,8 @@ def test_get_tagger_singleton():
 
 def test_load_raises_when_model_missing(tmp_path):
     t = Tagger(model_dir=tmp_path / "nonexistent")
-    with pytest.raises(RuntimeError, match="model.onnx missing"):
+    # Match the trailing "missing at <path>" rather than the specific
+    # filename, so a future model-version bump (camie-tagger-v3.onnx, etc.)
+    # doesn't bounce this test.
+    with pytest.raises(RuntimeError, match=r"\.onnx missing at "):
         t.load()

tests/test_phash_dedup.py

@@ -10,7 +10,15 @@ import pytest
 from PIL import Image
 from sqlalchemy import func, select
 
-from backend.app.models import ImageRecord, ImportSettings, Tag, TagKind
+from backend.app.models import (
+    ImageProvenance,
+    ImageRecord,
+    ImportSettings,
+    Post,
+    Source,
+    Tag,
+    TagKind,
+)
 from backend.app.models.tag import image_tag
 from backend.app.services.importer import Importer, SkipReason
 from backend.app.services.thumbnailer import Thumbnailer
@@ -141,6 +149,67 @@ def test_smaller_existing_is_superseded(importer, import_layout):
     assert Path(row.path).exists()
 
 
+def test_supersede_applies_new_file_sidecar(importer, import_layout):
+    """Operator-flagged 2026-05-25: scanning the GS download dir should
+    supersede smaller IR-migrated images AND wire up the GS sidecar's
+    Post/Source/ImageProvenance. Previously _supersede swapped the file
+    but ignored the sidecar entirely."""
+    import json
+    import_root, _ = import_layout
+
+    # Stage 1: a small, sidecar-less image (the "IR migration" precondition).
+    small = import_root / "ir-migration-folder" / "small.png"
+    _write(small, (60, 130, 200), (200, 200))
+    r1 = importer.import_one(small)
+    assert r1.status == "imported"
+    eid = r1.image_id
+
+    # Stage 2: a larger version of the same image (same phash) WITH a
+    # gallery-dl JSON sidecar adjacent. Live in a separate folder to
+    # simulate the GS download dir.
+    big = import_root / "Maewix" / "patreon" / "01_big.png"
+    _write(big, (60, 130, 200), (900, 900))
+    sidecar_path = big.with_suffix(big.suffix + ".json")
+    sidecar_path.parent.mkdir(parents=True, exist_ok=True)
+    sidecar_path.write_text(json.dumps({
+        "category": "patreon",
+        "id": 555,
+        "url": "https://www.patreon.com/posts/555",
+        "title": "Set 1",
+        "content": "<p>The big version</p>",
+        "page_count": 1,
+        "published_at": "2025-08-01T00:00:00+00:00",
+        "artist": "Maewix",
+    }))
+
+    r2 = importer.import_one(big)
+    assert r2.status == "superseded"
+    assert r2.image_id == eid
+
+    importer.session.expire_all()
+    # Row preserved, file replaced, sidecar metadata wired up.
+    row = importer.session.get(ImageRecord, eid)
+    assert row.width == 900 and row.height == 900
+
+    post = importer.session.execute(
+        select(Post).where(Post.external_post_id == "555")
+    ).scalar_one()
+    assert post.post_title == "Set 1"
+    assert "big version" in (post.description or "")
+
+    source = importer.session.execute(
+        select(Source).where(Source.id == post.source_id)
+    ).scalar_one()
+    assert source.platform == "patreon"
+
+    prov_count = importer.session.execute(
+        select(func.count(ImageProvenance.id))
+        .where(ImageProvenance.image_record_id == eid)
+        .where(ImageProvenance.post_id == post.id)
+    ).scalar_one()
+    assert prov_count == 1
+
+
 def test_threshold_controls_match(importer, import_layout):
     # Structurally distinct images (orthogonal splits) are far apart in
     # phash space, so a tight threshold keeps them independent rather than
@@ -172,3 +241,9 @@ def test_import_task_maps_superseded_to_complete_and_requeues():
     assert _map_result_to_status(
         ImportResult(status="failed", error="boom")
     ) == ("failed", False)
+    # Refreshed (deep scan): complete + no ML/thumb re-derive (pixels
+    # unchanged). Added 2026-05-25 alongside ImportBatch.refreshed
+    # counter so deep scan reports "X refreshed" instead of "no work".
+    assert _map_result_to_status(
+        ImportResult(status="refreshed", image_id=5)
+    ) == ("complete", False)

tests/test_scan_deep_requeue.py

@@ -0,0 +1,109 @@
+"""Deep scan re-queues already-completed ImportTasks.
+
+Operator-flagged 2026-05-25: deep scan used to skip everything that
+already had a non-failed ImportTask row, making a deep re-scan a no-op
+when no new files were added. That defeated the entire point of deep
+scan (re-apply sidecar metadata to existing rows). The skip-set now
+splits by mode — quick keeps the old "any non-failed" semantics; deep
+skips ONLY actively-in-flight statuses.
+"""
+
+import pytest
+from PIL import Image
+from sqlalchemy import func, select
+
+from backend.app.models import ImportBatch, ImportSettings, ImportTask
+from backend.app.tasks.scan import scan_directory
+
+pytestmark = pytest.mark.integration
+
+
+def _img(path):
+    path.parent.mkdir(parents=True, exist_ok=True)
+    Image.new("RGB", (40, 40), (10, 200, 80)).save(path, "JPEG")
+
+
+def test_deep_scan_requeues_completed_task(db_sync, tmp_path, monkeypatch):
+    """Quick scan then deep scan of the same /import: the file completed
+    in the first run should be re-enqueued by the deep run (different
+    ImportTask id, same source_path)."""
+    import_root = tmp_path / "import"
+    settings = db_sync.execute(
+        select(ImportSettings).where(ImportSettings.id == 1)
+    ).scalar_one()
+    settings.import_scan_path = str(import_root)
+    db_sync.commit()
+
+    src = import_root / "Mae" / "p.jpg"
+    _img(src)
+
+    from backend.app import celery_app
+
+    celery_app.celery.conf.task_always_eager = False  # explicit
+    try:
+        first_batch_id = scan_directory.run(triggered_by="manual", mode="quick")
+        first_task = db_sync.execute(
+            select(ImportTask).where(ImportTask.batch_id == first_batch_id)
+        ).scalar_one()
+        # Simulate the worker having finished it.
+        first_task.status = "complete"
+        db_sync.commit()
+
+        # Now deep scan: the SAME file should get a NEW ImportTask row.
+        second_batch_id = scan_directory.run(triggered_by="manual", mode="deep")
+        assert second_batch_id != first_batch_id
+
+        new_tasks_in_second_batch = db_sync.execute(
+            select(func.count())
+            .select_from(ImportTask)
+            .where(ImportTask.batch_id == second_batch_id)
+        ).scalar_one()
+        assert new_tasks_in_second_batch == 1, (
+            "deep scan did not re-queue the completed file"
+        )
+
+        # And the second batch's task should be for the same source_path
+        # as the first (proves it's a re-queue, not a different file).
+        sp = db_sync.execute(
+            select(ImportTask.source_path)
+            .where(ImportTask.batch_id == second_batch_id)
+        ).scalar_one()
+        assert sp == str(src)
+    finally:
+        celery_app.celery.conf.task_always_eager = False
+
+
+def test_quick_scan_does_not_requeue_completed_task(db_sync, tmp_path):
+    """The flip side: quick scan still keeps the old skip semantics —
+    a file with a completed ImportTask row from a prior batch is NOT
+    re-enqueued on a fresh quick scan."""
+    import_root = tmp_path / "import"
+    settings = db_sync.execute(
+        select(ImportSettings).where(ImportSettings.id == 1)
+    ).scalar_one()
+    settings.import_scan_path = str(import_root)
+    db_sync.commit()
+
+    src = import_root / "Mae" / "q.jpg"
+    _img(src)
+
+    first_batch_id = scan_directory.run(triggered_by="manual", mode="quick")
+    first_task = db_sync.execute(
+        select(ImportTask).where(ImportTask.batch_id == first_batch_id)
+    ).scalar_one()
+    first_task.status = "complete"
+    db_sync.commit()
+
+    second_batch_id = scan_directory.run(triggered_by="manual", mode="quick")
+    second_batch_count = db_sync.execute(
+        select(func.count())
+        .select_from(ImportTask)
+        .where(ImportTask.batch_id == second_batch_id)
+    ).scalar_one()
+    assert second_batch_count == 0, (
+        "quick scan re-queued an already-completed task — should have skipped"
+    )
+
+    # And the second batch should self-finalize (files_seen=0).
+    second_batch = db_sync.get(ImportBatch, second_batch_id)
+    assert second_batch.status == "complete"

tests/test_tag_apply.py

@@ -227,6 +227,67 @@ async def test_image_posts_creates_source_post_provenance(db, tmp_path):
     )).scalar_one()
     assert prov_count == 1
 
+    # Phase 4 must also set ImageRecord.primary_post_id so the gallery's
+    # effective_date COALESCE can surface Post.post_date. Operator-flagged
+    # 2026-05-25: without this, IR-migrated images keep sorting by FC's
+    # scan date instead of the original publish date.
+    primary_post_id = (await db.execute(
+        select(ImageRecord.primary_post_id).where(ImageRecord.id == img_id)
+    )).scalar_one()
+    canonical_post_id = (await db.execute(
+        select(Post.id).where(Post.external_post_id == "10001")
+    )).scalar_one()
+    assert primary_post_id == canonical_post_id
+
+
+@pytest.mark.asyncio
+async def test_image_posts_primary_post_id_not_clobbered(db, tmp_path):
+    """If the importer already set primary_post_id (e.g. a downloaded
+    image with a known provenance), phase 4 must NOT overwrite it when
+    re-running tag_apply against the IR migration. The existing
+    download-time linkage is the source of truth."""
+    sha = "9" * 64
+    await _seed_image(db, sha, suffix="9")
+    # Pre-set primary_post_id to a sentinel Post so we can detect a clobber.
+    img_id = (await db.execute(
+        select(ImageRecord.id).where(ImageRecord.sha256 == sha)
+    )).scalar_one()
+    # Build an existing Source + Post for the sentinel.
+    art = Artist(name="Pre-existing", slug="pre-existing")
+    db.add(art)
+    await db.flush()
+    src = Source(
+        artist_id=art.id, platform="patreon",
+        url="https://www.patreon.com/pre-existing",
+    )
+    db.add(src)
+    await db.flush()
+    sentinel_post = Post(
+        source_id=src.id, external_post_id="sentinel-99",
+        post_title="Pre-existing",
+    )
+    db.add(sentinel_post)
+    await db.flush()
+    await db.execute(
+        ImageRecord.__table__.update()
+        .where(ImageRecord.id == img_id)
+        .values(primary_post_id=sentinel_post.id)
+    )
+    await db.commit()
+
+    _write_manifest(tmp_path, image_posts=[
+        {**_POST_ENTRY, "image_sha256s": [sha]},
+    ])
+    await tag_apply.apply_async(db, images_root=tmp_path, dry_run=False)
+
+    # The migration created a NEW Post (external_post_id="10001") and a
+    # new ImageProvenance, but primary_post_id must still point at the
+    # original sentinel.
+    primary_post_id = (await db.execute(
+        select(ImageRecord.primary_post_id).where(ImageRecord.id == img_id)
+    )).scalar_one()
+    assert primary_post_id == sentinel_post.id
+
 
 @pytest.mark.asyncio
 async def test_image_posts_idempotent_on_rerun(db, tmp_path):
@@ -281,6 +342,42 @@ async def test_image_posts_unknown_platform_skipped(db, tmp_path):
     assert src_count == 0
 
 
+@pytest.mark.parametrize("platform,expected_url", [
+    ("deviantart", "https://www.deviantart.com/maewix"),
+    ("pixiv",      "https://www.pixiv.net/users/maewix"),
+])
+@pytest.mark.asyncio
+async def test_image_posts_extended_platforms_create_source(
+    db, tmp_path, platform, expected_url,
+):
+    """Regression for 2026-05-25 operator-reported bug: phase 4's
+    _PLATFORM_PROFILE_URL had only patreon/subscribestar/hentaifoundry,
+    silently dropping deviantart + pixiv PostMetadata from the IR migration."""
+    sha = f"{platform[0]}" * 64
+    await _seed_image(db, sha, suffix=f"_{platform}")
+    await db.commit()
+
+    _write_manifest(tmp_path, image_posts=[
+        {**_POST_ENTRY, "platform": platform, "image_sha256s": [sha]},
+    ])
+
+    result = await tag_apply.apply_async(db, images_root=tmp_path, dry_run=False)
+    assert result["counts"]["rows_inserted"] >= 1
+
+    src_url = (await db.execute(
+        select(Source.url).where(Source.platform == platform)
+    )).scalar_one()
+    assert src_url == expected_url
+
+    # ImageProvenance row was created.
+    prov_count = (await db.execute(
+        select(func.count(ImageProvenance.id))
+        .join(Source, Source.id == ImageProvenance.source_id)
+        .where(Source.platform == platform)
+    )).scalar_one()
+    assert prov_count == 1
+
+
 @pytest.mark.asyncio
 async def test_image_posts_dry_run_makes_no_writes(db, tmp_path):
     sha = "2" * 64