Merge pull request '#69 dev→main: release v26.06.04.0' from dev into main

pg_dump + zstd in runtime image, lift Quart body cap to 1 GiB. See PR #2.

.gitignore

@@ -69,4 +69,3 @@ Thumbs.db
 alembic/versions/__pycache__/
 *.sqlite
 *.sqlite-journal
-.superpowers/

alembic/env.py

@@ -2,21 +2,12 @@
 
 from logging.config import fileConfig
 
-from sqlalchemy import engine_from_config, pool, text
+from sqlalchemy import engine_from_config, pool
 
 from alembic import context
 from backend.app.config import get_config
 from backend.app.models import Base
 
-# Arbitrary fixed 64-bit key for the session/transaction advisory lock that
-# serializes concurrent `alembic upgrade head` runs. Every `web` replica runs
-# migrations in its entrypoint, so under `docker stack deploy` two replicas can
-# boot at once and race the same DDL — duplicate CREATE TABLE, then a crashed
-# replica (operator-flagged 2026-06-07: 0040 raced; one backend died with
-# AdminShutdown). The first replica to reach the lock migrates; the rest block,
-# then find the version table already at head and apply nothing.
-_MIGRATION_LOCK_KEY = 0xFCA1E35C
-
 config = context.config
 
 if config.config_file_name is not None:
@@ -53,18 +44,6 @@ def run_migrations_online() -> None:
             compare_type=True,
         )
         with context.begin_transaction():
-            # Serialize concurrent migrators (see _MIGRATION_LOCK_KEY). A
-            # transaction-scoped advisory lock: the first replica to get here
-            # holds it for the whole upgrade and is auto-released when this
-            # transaction ends. A sibling replica blocks on this line, and only
-            # once the leader commits does it proceed to read the version table
-            # — now at head — so it runs zero migrations instead of re-applying
-            # the same DDL. The lock is acquired BEFORE run_migrations() reads
-            # the current revision, which is what makes the no-op correct.
-            connection.execute(
-                text("SELECT pg_advisory_xact_lock(:k)"),
-                {"k": _MIGRATION_LOCK_KEY},
-            )
             context.run_migrations()
 
 

alembic/versions/0037_patreon_seen_media.py

@@ -1,53 +0,0 @@
-"""patreon_seen_media: per-source ledger of already-ingested Patreon media
-
-Revision ID: 0037
-Revises: 0036
-Create Date: 2026-06-05
-
-Native Patreon ingester (build step 2a). Replaces gallery-dl's
-archive.sqlite3 with our own queryable table. The downloader upserts one
-row per (source, media) so routine walks skip media we've already
-processed; a future "recovery" mode bypasses the ledger to re-walk.
-
-`filehash` is a 32-hex Patreon CDN MD5, OR a video sentinel of the form
-``video:<post_id>:<media_id>`` — hence String(128). The unique
-constraint on (source_id, filehash) is the dedup upsert key.
-"""
-from typing import Sequence, Union
-
-import sqlalchemy as sa
-from alembic import op
-
-revision: str = "0037"
-down_revision: Union[str, None] = "0036"
-branch_labels: Union[str, Sequence[str], None] = None
-depends_on: Union[str, Sequence[str], None] = None
-
-
-def upgrade() -> None:
-    op.create_table(
-        "patreon_seen_media",
-        sa.Column("id", sa.Integer, primary_key=True),
-        sa.Column(
-            "source_id",
-            sa.Integer,
-            sa.ForeignKey("source.id", ondelete="CASCADE"),
-            nullable=False,
-            index=True,
-        ),
-        sa.Column("filehash", sa.String(128), nullable=False),
-        sa.Column("post_id", sa.String(64), nullable=True),
-        sa.Column(
-            "seen_at",
-            sa.DateTime(timezone=True),
-            nullable=False,
-            server_default=sa.text("NOW()"),
-        ),
-        sa.UniqueConstraint(
-            "source_id", "filehash", name="uq_patreon_seen_media_source_id"
-        ),
-    )
-
-
-def downgrade() -> None:
-    op.drop_table("patreon_seen_media")

alembic/versions/0038_patreon_failed_media.py

@@ -1,58 +0,0 @@
-"""patreon_failed_media: per-source dead-letter ledger for failing Patreon media
-
-Revision ID: 0038
-Revises: 0037
-Create Date: 2026-06-06
-
-Plan #705 (#7). Media that keeps failing to download/validate (404'd CDN,
-deleted post, geo-blocked Mux, persistently-corrupt bytes) gets recorded here
-with an attempt counter; once it crosses the dead-letter threshold the ingester
-skips it on routine walks (recovery still re-attempts). A clean download clears
-the row. UNIQUE (source_id, filehash) is the upsert key (same media key the
-seen-ledger uses).
-"""
-from typing import Sequence, Union
-
-import sqlalchemy as sa
-from alembic import op
-
-revision: str = "0038"
-down_revision: Union[str, None] = "0037"
-branch_labels: Union[str, Sequence[str], None] = None
-depends_on: Union[str, Sequence[str], None] = None
-
-
-def upgrade() -> None:
-    op.create_table(
-        "patreon_failed_media",
-        sa.Column("id", sa.Integer, primary_key=True),
-        sa.Column(
-            "source_id",
-            sa.Integer,
-            sa.ForeignKey("source.id", ondelete="CASCADE"),
-            nullable=False,
-            index=True,
-        ),
-        sa.Column("filehash", sa.String(128), nullable=False),
-        sa.Column("attempts", sa.Integer, nullable=False, server_default="1"),
-        sa.Column("last_error", sa.Text, nullable=True),
-        sa.Column(
-            "first_failed_at",
-            sa.DateTime(timezone=True),
-            nullable=False,
-            server_default=sa.text("NOW()"),
-        ),
-        sa.Column(
-            "last_failed_at",
-            sa.DateTime(timezone=True),
-            nullable=False,
-            server_default=sa.text("NOW()"),
-        ),
-        sa.UniqueConstraint(
-            "source_id", "filehash", name="uq_patreon_failed_media_source_id"
-        ),
-    )
-
-
-def downgrade() -> None:
-    op.drop_table("patreon_failed_media")

alembic/versions/0039_library_audit_resume.py

@@ -1,40 +0,0 @@
-"""library_audit_run: resume cursor + progress timestamp for chunked scans
-
-Revision ID: 0039
-Revises: 0038
-Create Date: 2026-06-07
-
-scan_library_for_rule used to run one 2h pass that timed out on large libraries
-and monopolized the concurrency-1 maintenance queue (operator-flagged). It now
-runs short time-boxed chunks that re-enqueue: `resume_after_id` persists the
-keyset cursor so the next chunk continues where it left off, and
-`last_progress_at` lets the recovery sweep tell a progressing multi-chunk audit
-from a genuinely stuck one.
-"""
-from typing import Sequence, Union
-
-import sqlalchemy as sa
-from alembic import op
-
-revision: str = "0039"
-down_revision: Union[str, None] = "0038"
-branch_labels: Union[str, Sequence[str], None] = None
-depends_on: Union[str, Sequence[str], None] = None
-
-
-def upgrade() -> None:
-    op.add_column(
-        "library_audit_run",
-        sa.Column(
-            "resume_after_id", sa.Integer, nullable=False, server_default="0"
-        ),
-    )
-    op.add_column(
-        "library_audit_run",
-        sa.Column("last_progress_at", sa.DateTime(timezone=True), nullable=True),
-    )
-
-
-def downgrade() -> None:
-    op.drop_column("library_audit_run", "last_progress_at")
-    op.drop_column("library_audit_run", "resume_after_id")

alembic/versions/0040_series_chapters.py

@@ -1,108 +0,0 @@
-"""series chapters: chapter layer over series_page (FC-6.1)
-
-Revision ID: 0040
-Revises: 0039
-Create Date: 2026-06-07
-
-A series (Tag kind='series') gains an ordered chapter layer. Reading order
-becomes (series_chapter.chapter_number, series_page.page_number). Every existing
-series is backfilled into a single auto-chapter (chapter_number=1) holding its
-current flat pages, so no data is lost and the old flat ordering is preserved.
-"""
-from typing import Sequence, Union
-
-import sqlalchemy as sa
-from alembic import op
-
-revision: str = "0040"
-down_revision: Union[str, None] = "0039"
-branch_labels: Union[str, Sequence[str], None] = None
-depends_on: Union[str, Sequence[str], None] = None
-
-
-def upgrade() -> None:
-    op.create_table(
-        "series_chapter",
-        sa.Column("id", sa.Integer, primary_key=True),
-        sa.Column(
-            "series_tag_id",
-            sa.Integer,
-            sa.ForeignKey("tag.id", ondelete="CASCADE"),
-            nullable=False,
-        ),
-        sa.Column("chapter_number", sa.Integer, nullable=False),
-        sa.Column("title", sa.Text, nullable=True),
-        sa.Column(
-            "is_placeholder", sa.Boolean, nullable=False, server_default="false"
-        ),
-        sa.Column("stated_page_start", sa.Integer, nullable=True),
-        sa.Column("stated_page_end", sa.Integer, nullable=True),
-        sa.Column(
-            "created_at",
-            sa.DateTime(timezone=True),
-            nullable=False,
-            server_default=sa.text("now()"),
-        ),
-        sa.Column(
-            "updated_at",
-            sa.DateTime(timezone=True),
-            nullable=False,
-            server_default=sa.text("now()"),
-        ),
-    )
-    op.create_index(
-        "ix_series_chapter_series_tag_id", "series_chapter", ["series_tag_id"]
-    )
-
-    # New columns on series_page; chapter_id starts nullable so we can backfill.
-    op.add_column(
-        "series_page", sa.Column("chapter_id", sa.Integer, nullable=True)
-    )
-    op.add_column(
-        "series_page", sa.Column("stated_page", sa.Integer, nullable=True)
-    )
-
-    conn = op.get_bind()
-    # One auto-chapter per existing series (any series_tag_id present in pages).
-    conn.execute(
-        sa.text(
-            "INSERT INTO series_chapter "
-            "(series_tag_id, chapter_number, is_placeholder, created_at, updated_at) "
-            "SELECT DISTINCT series_tag_id, 1, false, now(), now() "
-            "FROM series_page"
-        )
-    )
-    # Point every existing page at its series' auto-chapter.
-    conn.execute(
-        sa.text(
-            "UPDATE series_page sp "
-            "SET chapter_id = sc.id "
-            "FROM series_chapter sc "
-            "WHERE sc.series_tag_id = sp.series_tag_id"
-        )
-    )
-
-    # Now lock chapter_id down: NOT NULL + FK (cascade) + index.
-    op.alter_column("series_page", "chapter_id", nullable=False)
-    op.create_foreign_key(
-        "fk_series_page_chapter_id",
-        "series_page",
-        "series_chapter",
-        ["chapter_id"],
-        ["id"],
-        ondelete="CASCADE",
-    )
-    op.create_index(
-        "ix_series_page_chapter_id", "series_page", ["chapter_id"]
-    )
-
-
-def downgrade() -> None:
-    op.drop_index("ix_series_page_chapter_id", table_name="series_page")
-    op.drop_constraint(
-        "fk_series_page_chapter_id", "series_page", type_="foreignkey"
-    )
-    op.drop_column("series_page", "stated_page")
-    op.drop_column("series_page", "chapter_id")
-    op.drop_index("ix_series_chapter_series_tag_id", table_name="series_chapter")
-    op.drop_table("series_chapter")

alembic/versions/0041_series_suggestions.py

@@ -1,98 +0,0 @@
-"""series suggestions: assisted-continuation matcher (FC-6.3)
-
-Revision ID: 0041
-Revises: 0040
-Create Date: 2026-06-07
-
-A confirm-only queue of "this post may continue this series" hints, plus two
-import_settings knobs (enable + score threshold) for the matcher.
-"""
-from typing import Sequence, Union
-
-import sqlalchemy as sa
-from alembic import op
-
-revision: str = "0041"
-down_revision: Union[str, None] = "0040"
-branch_labels: Union[str, Sequence[str], None] = None
-depends_on: Union[str, Sequence[str], None] = None
-
-
-def upgrade() -> None:
-    op.create_table(
-        "series_suggestion",
-        sa.Column("id", sa.Integer, primary_key=True),
-        sa.Column(
-            "post_id",
-            sa.Integer,
-            sa.ForeignKey("post.id", ondelete="CASCADE"),
-            nullable=False,
-        ),
-        sa.Column(
-            "series_tag_id",
-            sa.Integer,
-            sa.ForeignKey("tag.id", ondelete="CASCADE"),
-            nullable=False,
-        ),
-        sa.Column("score", sa.Float, nullable=False),
-        sa.Column("signals", sa.JSON, nullable=True),
-        sa.Column(
-            "status", sa.String(16), nullable=False, server_default="pending"
-        ),
-        sa.Column(
-            "created_at",
-            sa.DateTime(timezone=True),
-            nullable=False,
-            server_default=sa.text("now()"),
-        ),
-        sa.Column(
-            "updated_at",
-            sa.DateTime(timezone=True),
-            nullable=False,
-            server_default=sa.text("now()"),
-        ),
-        sa.UniqueConstraint(
-            "post_id", "series_tag_id", name="uq_series_suggestion_post_series"
-        ),
-    )
-    op.create_index(
-        "ix_series_suggestion_post_id", "series_suggestion", ["post_id"]
-    )
-    op.create_index(
-        "ix_series_suggestion_series_tag_id",
-        "series_suggestion",
-        ["series_tag_id"],
-    )
-    op.create_index(
-        "ix_series_suggestion_status", "series_suggestion", ["status"]
-    )
-
-    op.add_column(
-        "import_settings",
-        sa.Column(
-            "series_suggest_enabled",
-            sa.Boolean,
-            nullable=False,
-            server_default=sa.true(),
-        ),
-    )
-    op.add_column(
-        "import_settings",
-        sa.Column(
-            "series_suggest_threshold",
-            sa.Float,
-            nullable=False,
-            server_default="0.5",
-        ),
-    )
-
-
-def downgrade() -> None:
-    op.drop_column("import_settings", "series_suggest_threshold")
-    op.drop_column("import_settings", "series_suggest_enabled")
-    op.drop_index("ix_series_suggestion_status", table_name="series_suggestion")
-    op.drop_index(
-        "ix_series_suggestion_series_tag_id", table_name="series_suggestion"
-    )
-    op.drop_index("ix_series_suggestion_post_id", table_name="series_suggestion")
-    op.drop_table("series_suggestion")

alembic/versions/0042_series_chapter_stated_part.py

@@ -1,32 +0,0 @@
-"""series chapter stated_part: operator-facing Part N label (FC-6.4)
-
-Revision ID: 0042
-Revises: 0041
-Create Date: 2026-06-07
-
-A chapter's positional chapter_number is auto-managed (rewritten 1..N on
-reorder/delete), so it can't double as the installment number the operator wants
-to type (e.g. a series authored from a post that is Part 2). Add a nullable
-stated_part alongside it — the same split as series_page.page_number (order) vs
-series_page.stated_page (printed number). Nullable; the UI falls back to
-chapter_number when unset.
-"""
-from typing import Sequence, Union
-
-import sqlalchemy as sa
-from alembic import op
-
-revision: str = "0042"
-down_revision: Union[str, None] = "0041"
-branch_labels: Union[str, Sequence[str], None] = None
-depends_on: Union[str, Sequence[str], None] = None
-
-
-def upgrade() -> None:
-    op.add_column(
-        "series_chapter", sa.Column("stated_part", sa.Integer, nullable=True)
-    )
-
-
-def downgrade() -> None:
-    op.drop_column("series_chapter", "stated_part")

backend/app/api/admin.py

@@ -245,32 +245,6 @@ async def tags_reset_content():
     return jsonify(result)
 
 
-@admin_bp.route("/tags/normalize", methods=["POST"])
-async def tags_normalize():
-    """#714: retro-normalize existing tags to the #701 canonical form (Title
-    Case + collapsed whitespace) and merge case/whitespace-variant duplicates.
-
-    dry_run=true (default) returns a projection inline — group/collision/rename
-    counts + a sample of the changes — so the UI shows exactly what'll happen.
-    dry_run=false dispatches the long-running maintenance task (the merge FK
-    repoints can touch many tags); the UI tails the activity dashboard for the
-    summary. Idempotent; back up first (the merges are irreversible)."""
-    from ..services.tag_service import normalize_existing_tags
-
-    body = await request.get_json(silent=True) or {}
-    dry_run = bool(body.get("dry_run", True))
-
-    if dry_run:
-        async with get_session() as session:
-            result = await normalize_existing_tags(session, dry_run=True)
-        return jsonify(result)
-
-    from ..tasks.admin import normalize_tags_task
-
-    async_result = normalize_tags_task.delay()
-    return jsonify({"task_id": async_result.id, "status": "queued"}), 202
-
-
 @admin_bp.route("/maintenance/db-stats", methods=["GET"])
 async def db_stats():
     """Per-table bloat readout (pg_stat_user_tables) for the high-churn tables
@@ -315,14 +289,3 @@ async def trigger_vacuum():
 
     vacuum_analyze.delay()
     return jsonify({"status": "queued"}), 202
-
-
-@admin_bp.route("/maintenance/reextract-archives", methods=["POST"])
-async def trigger_reextract_archives():
-    """Operator-triggered re-extract (#713): PostAttachments that are actually
-    archives but were filed opaquely (pre magic-byte gate) get extracted and
-    their members linked to the post. Idempotent; runs on the maintenance queue."""
-    from ..tasks.admin import reextract_archive_attachments_task
-
-    async_result = reextract_archive_attachments_task.delay()
-    return jsonify({"task_id": async_result.id, "status": "queued"}), 202

backend/app/api/credentials.py

@@ -121,15 +121,13 @@ async def delete_credential(platform: str):
 
 @credentials_bp.route("/<platform>/verify", methods=["POST"])
 async def verify_credential(platform: str):
-    """Test the stored credential against one of the platform's enabled sources,
-    WITHOUT downloading. Routes through the platform's backend
-    (download_backends.verify_credential) — native ingester for Patreon, an
-    authenticated API page; gallery-dl --simulate for the rest. On success
-    stamps last_verified. Returns {valid: bool|null, reason, last_verified?};
-    valid=null means "couldn't test" (no credential, no enabled source, or an
-    inconclusive network/drift result)."""
+    """Test the stored credential by running gallery-dl --simulate
+    against one of the platform's enabled sources. On success stamps
+    last_verified. Returns {valid: bool|null, reason, last_verified?}.
+    valid=null means "couldn't test" (no credential, or no enabled
+    source to point at)."""
     from ..models import Artist, Source
-    from ..services.download_backends import verify_source_credential
+    from ..services.gallery_dl import GalleryDLService, SourceConfig
 
     async with get_session() as session:
         if not await _ext_key_ok(session):
@@ -156,14 +154,14 @@ async def verify_credential(platform: str):
         cookies_path = await svc.get_cookies_path(platform)
         auth_token = await svc.get_token(platform)
 
-    ok, message = await verify_source_credential(
-        platform=platform,
+    gdl = GalleryDLService(images_root=Path("/images"))
+    ok, message = await gdl.verify(
         url=source.url,
         artist_slug=artist.slug,
-        config_overrides=source.config_overrides or {},
+        platform=platform,
+        source_config=SourceConfig.from_dict(source.config_overrides or {}),
         cookies_path=str(cookies_path) if cookies_path else None,
         auth_token=auth_token,
-        images_root=Path("/images"),
     )
 
     last_verified = None

backend/app/api/downloads.py

@@ -44,9 +44,6 @@ def _list_record(event: DownloadEvent, source: Source | None, artist: Artist | N
         "bytes_downloaded": event.bytes_downloaded,
         "error": event.error,
         "summary": _summary_from_metadata(event.metadata_),
-        # plan #709: mid-walk live counts for a RUNNING native-ingester event
-        # (None otherwise; phase 3 overwrites metadata with run_stats on finish).
-        "live": (event.metadata_ or {}).get("live"),
     }
 
 

backend/app/api/settings.py

@@ -25,8 +25,6 @@ _EDITABLE_FIELDS = (
     "download_schedule_default_seconds",
     "download_event_retention_days",
     "download_failure_warning_threshold",
-    "series_suggest_enabled",
-    "series_suggest_threshold",
 )
 
 
@@ -48,8 +46,6 @@ async def get_import_settings():
         "download_schedule_default_seconds": row.download_schedule_default_seconds,
         "download_event_retention_days": row.download_event_retention_days,
         "download_failure_warning_threshold": row.download_failure_warning_threshold,
-        "series_suggest_enabled": row.series_suggest_enabled,
-        "series_suggest_threshold": row.series_suggest_threshold,
     })
 
 
@@ -100,19 +96,6 @@ async def update_import_settings():
         if not isinstance(v, int) or isinstance(v, bool) or v < 1 or v > 100:
             return _bad_int("download_failure_warning_threshold", 1, 100)
 
-    if "series_suggest_enabled" in body and not isinstance(
-        body["series_suggest_enabled"], bool
-    ):
-        return jsonify(
-            {"error": "series_suggest_enabled must be a boolean"}
-        ), 400
-    if "series_suggest_threshold" in body:
-        v = body["series_suggest_threshold"]
-        if not isinstance(v, (int, float)) or isinstance(v, bool) or v < 0 or v > 1:
-            return jsonify(
-                {"error": "series_suggest_threshold must be a number in [0, 1]"}
-            ), 400
-
     async with get_session() as session:
         row = await ImportSettings.load(session)
         for field in _EDITABLE_FIELDS:

backend/app/api/sources.py

@@ -85,22 +85,6 @@ async def create_source():
             return _bad("empty_url", detail=str(exc))
         except DuplicateSourceError as exc:
             return _bad("duplicate", status=409, existing_id=exc.existing_id)
-
-        # Immediate kickoff: a new enabled source is armed for backfill (#693)
-        # but would otherwise sit idle until the next scheduler tick (~60s).
-        # Enqueue the first walk now, skipping only if the platform is in a
-        # rate-limit cooldown (the scheduler picks it up when that clears).
-        dispatch_id = None
-        if record.enabled:
-            cooldowns = await active_platform_cooldowns(session)
-            if record.platform not in cooldowns:
-                session.add(DownloadEvent(source_id=record.id, status="pending"))
-                await session.commit()
-                dispatch_id = record.id
-
-    if dispatch_id is not None:
-        from ..tasks.download import download_source
-        download_source.delay(dispatch_id)
     return jsonify(record.to_dict()), 201
 
 
@@ -138,119 +122,29 @@ async def delete_source(source_id: int):
 
 @sources_bp.route("/<int:source_id>/backfill", methods=["POST"])
 async def set_backfill(source_id: int):
-    """Plan #693/#697: start/stop a run-until-done backfill, or start a recovery.
-    Body: `{"action": "start" | "stop" | "recover"}` (default "start"). 'start'
-    walks the full post history in time-boxed chunks until it reaches the bottom
-    (then the source shows 'complete'); 'recover' is the same walk but bypasses
-    the Patreon seen-ledger to re-fetch dropped-and-deleted near-dups under the
-    current pHash threshold; 'stop' cancels either back to tick mode. Returns the
-    updated source dict (incl. backfill_state / backfill_chunks /
-    backfill_bypass_seen)."""
-    from pathlib import Path
-
-    from ..services.credential_service import CredentialService
-    from ..services.download_backends import (
-        uses_native_ingester,
-        verify_source_credential,
-    )
-    from .credentials import _get_crypto
-
+    """Plan #544: arm a source for backfill mode for the next N download
+    runs. Body: `{"runs": int}` (1..10, default 3). Returns the updated
+    source dict. While backfill_runs_remaining > 0, downloads use
+    gallery-dl's full-walk config (skip: True + 30-min timeout) instead
+    of the catch-up default (skip: "exit:20" + 14.5-min timeout)."""
     payload = await request.get_json(silent=True) or {}
-    action = payload.get("action", "start")
-    if action not in ("start", "stop", "recover"):
-        return _bad(
-            "invalid_action",
-            detail="action must be 'start', 'stop', or 'recover'",
-        )
-
-    # Pre-flight (plan #703 #2): before arming a deep walk on a native-ingester
-    # platform (where verify is one cheap API page), refuse if the credential is
-    # DEFINITIVELY rejected — don't burn chunks against expired cookies. Proceed
-    # on valid OR inconclusive (a network blip shouldn't block). Gated to native
-    # platforms: gallery-dl verify is a slow --simulate subprocess, too heavy for
-    # an arm action. The credential read happens in a session that's CLOSED
-    # before the verify network call (don't hold a DB conn across the request).
-    if action in ("start", "recover"):
-        async with get_session() as session:
-            rec = await SourceService(session).get(source_id)
-            if rec is None:
-                return _bad("not_found", status=404)
-            native = uses_native_ingester(rec.platform)
-            if native:
-                cred = CredentialService(session, _get_crypto())
-                cookies_path = await cred.get_cookies_path(rec.platform)
-                auth_token = await cred.get_token(rec.platform)
-        if native:
-            ok, message = await verify_source_credential(
-                platform=rec.platform,
-                url=rec.url,
-                artist_slug=rec.artist_slug,
-                config_overrides=rec.config_overrides or {},
-                cookies_path=str(cookies_path) if cookies_path else None,
-                auth_token=auth_token,
-                images_root=Path("/images"),
-            )
-            if ok is False:
-                return _bad("credential_rejected", detail=message, status=409)
-
+    runs = payload.get("runs", 3)
+    try:
+        runs = int(runs)
+    except (TypeError, ValueError):
+        return _bad("invalid_runs", detail="runs must be an integer")
     async with get_session() as session:
         try:
-            svc = SourceService(session)
-            if action == "start":
-                record = await svc.start_backfill(source_id)
-            elif action == "recover":
-                record = await svc.start_recovery(source_id)
-            else:
-                record = await svc.stop_backfill(source_id)
+            record = await SourceService(session).set_backfill_runs(
+                source_id, runs,
+            )
         except LookupError:
             return _bad("not_found", status=404)
+        except ValueError as exc:
+            return _bad("invalid_runs", detail=str(exc))
     return jsonify(record.to_dict())
 
 
-@sources_bp.route("/<int:source_id>/preview", methods=["POST"])
-async def preview_source_endpoint(source_id: int):
-    """Plan #708 B4: dry-run — count what a backfill WOULD download for a native
-    platform (Patreon today), without downloading. Walks the first few feed pages
-    and counts media not already in the seen/dead ledgers. Returns
-    {total_new, posts_scanned, pages_scanned, has_more, sample[]} or 409 + reason
-    (unresolvable campaign id / auth / drift). 400 for gallery-dl platforms (no
-    cheap dry-run — their verify is a slow --simulate)."""
-    from pathlib import Path
-
-    from ..services.credential_service import CredentialService
-    from ..services.download_backends import preview_source, uses_native_ingester
-    from ..tasks._sync_engine import sync_session_factory
-    from .credentials import _get_crypto
-
-    async with get_session() as session:
-        rec = await SourceService(session).get(source_id)
-        if rec is None:
-            return _bad("not_found", status=404)
-        if not uses_native_ingester(rec.platform):
-            return _bad(
-                "unsupported",
-                detail="Preview is only available for native-ingester platforms.",
-                status=400,
-            )
-        cred = CredentialService(session, _get_crypto())
-        cookies_path = await cred.get_cookies_path(rec.platform)
-
-    # The walk + ledger reads are sync (run off the request loop); the process
-    # sync engine is the same one the download task uses.
-    result = await preview_source(
-        platform=rec.platform,
-        url=rec.url,
-        source_id=source_id,
-        config_overrides=rec.config_overrides or {},
-        cookies_path=str(cookies_path) if cookies_path else None,
-        images_root=Path("/images"),
-        sync_session_factory=sync_session_factory(),
-    )
-    if "error" in result:
-        return _bad("preview_failed", detail=result["error"], status=409)
-    return jsonify(result)
-
-
 @sources_bp.route("/<int:source_id>/check", methods=["POST"])
 async def check_source(source_id: int):
     """FC-3c: enqueue a download for this source.

backend/app/api/system_activity.py

@@ -31,7 +31,7 @@ system_activity_bp = Blueprint(
 # absent.
 _QUEUE_NAMES = (
     "default", "import", "thumbnail", "ml",
-    "download", "scan", "maintenance", "maintenance_long",
+    "download", "scan", "maintenance",
 )
 
 # Cache module-level so all requests share the cache between polls.

backend/app/api/tags.py

@@ -8,14 +8,12 @@ from ..extensions import get_session
 from ..models import Tag, TagKind
 from ..models.tag_allowlist import TagAllowlist
 from ..services.bulk_tag_service import BulkTagService
-from ..services.series_match_service import SeriesMatchService
 from ..services.series_service import SeriesError, SeriesService
 from ..services.tag_directory_service import TagDirectoryService
 from ..services.tag_service import (
     TagMergeConflict,
     TagService,
     TagValidationError,
-    normalize_tag_name,
 )
 from ..utils.tag_prefix import parse_kind_prefix
 
@@ -143,11 +141,6 @@ async def create_tag():
 
     fandom_id = body.get("fandom_id")
 
-    # #701: Title-Case operator-entered tags. Only here (the explicit create
-    # endpoint), NOT in the shared find_or_create — the ML tagger uses that path
-    # and must keep the booru vocabulary's casing for allowlist matching.
-    name = normalize_tag_name(name)
-
     async with get_session() as session:
         svc = TagService(session)
         try:
@@ -369,31 +362,6 @@ def _series_err(exc: SeriesError):
     return jsonify({"error": msg}), status
 
 
-def _opt_int(body, key: str):
-    """(value, error) — value is None when absent, error is (json, status)."""
-    if not body or body.get(key) is None:
-        return None, None
-    try:
-        return int(body[key]), None
-    except (TypeError, ValueError):
-        return None, (jsonify({"error": f"{key} must be an integer"}), 400)
-
-
-def _parse_int_list(body, key: str, *, max_ids: int = 500):
-    """(list, error) for a required list of ints under `key`."""
-    if not body or key not in body:
-        return None, (jsonify({"error": f"{key} required"}), 400)
-    raw = body[key]
-    if not isinstance(raw, list) or not raw:
-        return None, (jsonify({"error": f"{key} must be a non-empty list"}), 400)
-    if len(raw) > max_ids:
-        return None, (jsonify({"error": f"too many ids (max {max_ids})"}), 400)
-    try:
-        return [int(x) for x in raw], None
-    except (TypeError, ValueError):
-        return None, (jsonify({"error": f"{key} must be integers"}), 400)
-
-
 @tags_bp.route("/series/<int:tag_id>/pages", methods=["GET"])
 async def series_pages(tag_id: int):
     async with get_session() as session:
@@ -410,14 +378,9 @@ async def series_add(tag_id: int):
     ids, err = _parse_bulk_ids(body, max_ids=500)
     if err:
         return err
-    chapter_id, cerr = _opt_int(body, "chapter_id")
-    if cerr:
-        return cerr
     async with get_session() as session:
         try:
-            n = await SeriesService(session).add_images(
-                tag_id, ids, chapter_id=chapter_id
-            )
+            n = await SeriesService(session).add_images(tag_id, ids)
         except SeriesError as exc:
             return _series_err(exc)
         await session.commit()
@@ -470,232 +433,3 @@ async def series_cover(tag_id: int):
             return _series_err(exc)
         await session.commit()
     return jsonify({"ok": True})
-
-
-# ---- chapters (FC-6.1) ----------------------------------------------------
-
-
-@tags_bp.route("/series/<int:tag_id>/chapters", methods=["POST"])
-async def series_chapter_create(tag_id: int):
-    body = await request.get_json() or {}
-    title = body.get("title")
-    if title is not None and not isinstance(title, str):
-        return jsonify({"error": "title must be a string"}), 400
-    is_placeholder = bool(body.get("is_placeholder", False))
-    start, serr = _opt_int(body, "stated_page_start")
-    if serr:
-        return serr
-    end, eerr = _opt_int(body, "stated_page_end")
-    if eerr:
-        return eerr
-    async with get_session() as session:
-        try:
-            ch = await SeriesService(session).create_chapter(
-                tag_id,
-                title=title,
-                is_placeholder=is_placeholder,
-                stated_page_start=start,
-                stated_page_end=end,
-            )
-        except SeriesError as exc:
-            return _series_err(exc)
-        await session.commit()
-    return jsonify(ch)
-
-
-@tags_bp.route("/series/<int:tag_id>/chapters/reorder", methods=["POST"])
-async def series_chapter_reorder(tag_id: int):
-    body = await request.get_json()
-    ids, err = _parse_int_list(body, "chapter_ids")
-    if err:
-        return err
-    async with get_session() as session:
-        try:
-            await SeriesService(session).reorder_chapters(tag_id, ids)
-        except SeriesError as exc:
-            return _series_err(exc)
-        await session.commit()
-    return jsonify({"ok": True})
-
-
-@tags_bp.route(
-    "/series/<int:tag_id>/chapters/<int:chapter_id>", methods=["PATCH"]
-)
-async def series_chapter_update(tag_id: int, chapter_id: int):
-    body = await request.get_json() or {}
-    kwargs: dict = {}
-    if "title" in body:
-        if body["title"] is not None and not isinstance(body["title"], str):
-            return jsonify({"error": "title must be a string"}), 400
-        kwargs.update(set_title=True, title=body["title"])
-    if "stated_part" in body:
-        part, perr = _opt_int(body, "stated_part")
-        if perr:
-            return perr
-        kwargs.update(set_part=True, stated_part=part)
-    if "stated_page_start" in body:
-        start, serr = _opt_int(body, "stated_page_start")
-        if serr:
-            return serr
-        kwargs.update(set_start=True, stated_page_start=start)
-    if "stated_page_end" in body:
-        end, eerr = _opt_int(body, "stated_page_end")
-        if eerr:
-            return eerr
-        kwargs.update(set_end=True, stated_page_end=end)
-    async with get_session() as session:
-        try:
-            await SeriesService(session).update_chapter(
-                tag_id, chapter_id, **kwargs
-            )
-        except SeriesError as exc:
-            return _series_err(exc)
-        await session.commit()
-    return jsonify({"ok": True})
-
-
-@tags_bp.route(
-    "/series/<int:tag_id>/chapters/<int:chapter_id>", methods=["DELETE"]
-)
-async def series_chapter_delete(tag_id: int, chapter_id: int):
-    async with get_session() as session:
-        try:
-            await SeriesService(session).delete_chapter(tag_id, chapter_id)
-        except SeriesError as exc:
-            return _series_err(exc)
-        await session.commit()
-    return jsonify({"ok": True})
-
-
-@tags_bp.route(
-    "/series/<int:tag_id>/chapters/<int:chapter_id>/merge", methods=["POST"]
-)
-async def series_chapter_merge(tag_id: int, chapter_id: int):
-    body = await request.get_json()
-    target, terr = _opt_int(body, "target_chapter_id")
-    if terr:
-        return terr
-    if target is None:
-        return jsonify({"error": "target_chapter_id required"}), 400
-    async with get_session() as session:
-        try:
-            moved = await SeriesService(session).merge_chapter(
-                tag_id, chapter_id, target
-            )
-        except SeriesError as exc:
-            return _series_err(exc)
-        await session.commit()
-    return jsonify({"moved_count": moved})
-
-
-@tags_bp.route(
-    "/series/<int:tag_id>/chapters/<int:chapter_id>/reorder", methods=["POST"]
-)
-async def series_chapter_reorder_pages(tag_id: int, chapter_id: int):
-    body = await request.get_json()
-    ids, err = _parse_bulk_ids(body, max_ids=500)
-    if err:
-        return err
-    async with get_session() as session:
-        try:
-            await SeriesService(session).reorder_pages(tag_id, chapter_id, ids)
-        except SeriesError as exc:
-            return _series_err(exc)
-        await session.commit()
-    return jsonify({"ok": True})
-
-
-# ---- browse list + post→series flows (FC-6.2) -----------------------------
-
-
-@tags_bp.route("/series", methods=["GET"])
-async def series_list():
-    args = request.args
-    sort = args.get("sort", "recent")
-    if sort not in ("recent", "name", "size"):
-        return jsonify({"error": "sort must be recent|name|size"}), 400
-    artist_id = None
-    if args.get("artist_id") is not None:
-        try:
-            artist_id = int(args["artist_id"])
-        except ValueError:
-            return jsonify({"error": "artist_id must be an integer"}), 400
-    async with get_session() as session:
-        rows = await SeriesService(session).list_series(
-            sort=sort, artist_id=artist_id
-        )
-    return jsonify({"series": rows})
-
-
-@tags_bp.route("/series/from-post", methods=["POST"])
-async def series_from_post():
-    body = await request.get_json()
-    post_id, err = _opt_int(body, "post_id")
-    if err:
-        return err
-    if post_id is None:
-        return jsonify({"error": "post_id required"}), 400
-    async with get_session() as session:
-        try:
-            out = await SeriesService(session).promote_post_to_series(post_id)
-        except SeriesError as exc:
-            return _series_err(exc)
-        await session.commit()
-    return jsonify(out)
-
-
-@tags_bp.route("/series/<int:tag_id>/add-post", methods=["POST"])
-async def series_add_post(tag_id: int):
-    body = await request.get_json()
-    post_id, err = _opt_int(body, "post_id")
-    if err:
-        return err
-    if post_id is None:
-        return jsonify({"error": "post_id required"}), 400
-    async with get_session() as session:
-        try:
-            out = await SeriesService(session).add_post_as_chapter(tag_id, post_id)
-        except SeriesError as exc:
-            return _series_err(exc)
-        await session.commit()
-    return jsonify(out)
-
-
-# ---- suggestion queue (FC-6.3) --------------------------------------------
-
-
-@tags_bp.route("/series/suggestions", methods=["GET"])
-async def series_suggestions_list():
-    async with get_session() as session:
-        rows = await SeriesMatchService(session).list_pending()
-    return jsonify({"suggestions": rows})
-
-
-@tags_bp.route("/series/suggestions/<int:sid>/accept", methods=["POST"])
-async def series_suggestion_accept(sid: int):
-    async with get_session() as session:
-        try:
-            out = await SeriesMatchService(session).accept(sid)
-        except SeriesError as exc:
-            return _series_err(exc)
-        await session.commit()
-    return jsonify(out)
-
-
-@tags_bp.route("/series/suggestions/<int:sid>/dismiss", methods=["POST"])
-async def series_suggestion_dismiss(sid: int):
-    async with get_session() as session:
-        try:
-            await SeriesMatchService(session).dismiss(sid)
-        except SeriesError as exc:
-            return _series_err(exc)
-        await session.commit()
-    return jsonify({"ok": True})
-
-
-@tags_bp.route("/series/suggestions/rescan", methods=["POST"])
-async def series_suggestions_rescan():
-    from ..tasks.admin import rescan_series_suggestions_task
-
-    res = rescan_series_suggestions_task.delay()
-    return jsonify({"task_id": res.id})

backend/app/celery_app.py

@@ -43,16 +43,10 @@ def make_celery() -> Celery:
             "backend.app.tasks.thumbnail.*": {"queue": "thumbnail"},
             "backend.app.tasks.download.*": {"queue": "download"},
             "backend.app.tasks.scan.*": {"queue": "scan"},
-            # `maintenance` is the QUICK lane — recovery sweeps, vacuum, cleanup
-            # (concurrency-1 on the scheduler). The long one-shots (DB backups,
-            # library audits, admin maintenance: normalize/re-extract/cascade-
-            # delete) run on a SEPARATE `maintenance_long` lane + worker so they
-            # can never starve the quick self-healing sweeps (operator-flagged
-            # 2026-06-07: a 2h audit blocked vacuum/backup/normalize for hours).
             "backend.app.tasks.maintenance.*": {"queue": "maintenance"},
-            "backend.app.tasks.backup.*": {"queue": "maintenance_long"},
-            "backend.app.tasks.admin.*": {"queue": "maintenance_long"},
-            "backend.app.tasks.library_audit.*": {"queue": "maintenance_long"},
+            "backend.app.tasks.backup.*": {"queue": "maintenance"},
+            "backend.app.tasks.admin.*": {"queue": "maintenance"},
+            "backend.app.tasks.library_audit.*": {"queue": "maintenance"},
         },
         # Heavy ML tasks need fair dispatch — see ImageRepo's precedent.
         task_acks_late=True,

backend/app/models/__init__.py

@@ -14,13 +14,9 @@ from .import_settings import ImportSettings
 from .import_task import ImportTask
 from .library_audit_run import LibraryAuditRun
 from .ml_settings import MLSettings
-from .patreon_failed_media import PatreonFailedMedia
-from .patreon_seen_media import PatreonSeenMedia
 from .post import Post
 from .post_attachment import PostAttachment
-from .series_chapter import SeriesChapter
 from .series_page import SeriesPage
-from .series_suggestion import SeriesSuggestion
 from .source import Source
 from .tag import Tag, TagKind, image_tag
 from .tag_alias import TagAlias
@@ -37,13 +33,9 @@ __all__ = [
     "BackupRun",
     "Source",
     "Credential",
-    "PatreonFailedMedia",
-    "PatreonSeenMedia",
     "Post",
     "PostAttachment",
-    "SeriesChapter",
     "SeriesPage",
-    "SeriesSuggestion",
     "ImageRecord",
     "ImageProvenance",
     "Tag",

backend/app/models/import_settings.py

@@ -64,15 +64,6 @@ class ImportSettings(Base):
         Integer, nullable=False, default=3,
     )
 
-    # FC-6.3 series continuation matcher. enabled gates the rescan; threshold is
-    # the weighted-score cut-off (0..1) above which a pending suggestion is made.
-    series_suggest_enabled: Mapped[bool] = mapped_column(
-        Boolean, nullable=False, default=True,
-    )
-    series_suggest_threshold: Mapped[float] = mapped_column(
-        Float, nullable=False, default=0.5,
-    )
-
     @classmethod
     async def load(cls, session) -> ImportSettings:
         """The singleton settings row (id=1), via an async session."""

backend/app/models/library_audit_run.py

@@ -35,10 +35,3 @@ class LibraryAuditRun(Base):
     matched_count: Mapped[int] = mapped_column(Integer, nullable=False, default=0)
     matched_ids: Mapped[list[int]] = mapped_column(JSONB, nullable=False, default=list)
     error: Mapped[str | None] = mapped_column(Text, nullable=True)
-    # Chunked-scan state (alembic 0039): keyset cursor the next chunk resumes
-    # from, and the last time a chunk made progress (so the recovery sweep can
-    # tell a progressing multi-chunk audit from a stuck one).
-    resume_after_id: Mapped[int] = mapped_column(Integer, nullable=False, default=0)
-    last_progress_at: Mapped[datetime | None] = mapped_column(
-        DateTime(timezone=True), nullable=True,
-    )

backend/app/models/patreon_failed_media.py

@@ -1,45 +0,0 @@
-"""PatreonFailedMedia — per-source dead-letter ledger of Patreon media that
-keeps failing to download/validate.
-
-Plan #705 (#7). A media that fails every walk (404'd CDN URL, deleted post,
-geo-blocked Mux stream, persistently-corrupt bytes) would otherwise re-error
-forever and re-burn backfill chunks. After ``attempts`` reaches the dead-letter
-threshold the ingester skips it on routine tick/backfill walks (recovery still
-re-attempts it — the operator's "try everything again"). A later clean download
-clears the row (the media recovered).
-
-`filehash` is the same per-media key the seen-ledger uses (32-hex CDN MD5, or a
-``video:`` / ``post:filename`` synthesized key) — hence String(128). UNIQUE
-(source_id, filehash) is the upsert key.
-"""
-
-from datetime import datetime
-
-from sqlalchemy import ForeignKey, Integer, String, Text, UniqueConstraint, func
-from sqlalchemy.orm import Mapped, mapped_column
-from sqlalchemy.types import DateTime
-
-from .base import Base
-
-
-class PatreonFailedMedia(Base):
-    __tablename__ = "patreon_failed_media"
-    __table_args__ = (
-        UniqueConstraint(
-            "source_id", "filehash", name="uq_patreon_failed_media_source_id"
-        ),
-    )
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True)
-    source_id: Mapped[int] = mapped_column(
-        ForeignKey("source.id", ondelete="CASCADE"), nullable=False, index=True
-    )
-    filehash: Mapped[str] = mapped_column(String(128), nullable=False)
-    attempts: Mapped[int] = mapped_column(Integer, nullable=False, default=1)
-    last_error: Mapped[str | None] = mapped_column(Text, nullable=True)
-    first_failed_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), nullable=False, server_default=func.now()
-    )
-    last_failed_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), nullable=False, server_default=func.now()
-    )

backend/app/models/patreon_seen_media.py

@@ -1,38 +0,0 @@
-"""PatreonSeenMedia — per-source ledger of Patreon media already
-downloaded+processed.
-
-Replaces gallery-dl's archive.sqlite3 with our own queryable table so
-routine walks can skip media we've already ingested (and a future
-"recovery" mode can deliberately bypass the ledger to re-walk).
-
-`filehash` is normally a Patreon CDN MD5 (32 hex chars), but videos —
-which have no stable content hash at discovery time — use a sentinel of
-the form ``video:<post_id>:<media_id>``, hence String(128) rather than 32.
-"""
-
-from datetime import datetime
-
-from sqlalchemy import ForeignKey, Integer, String, UniqueConstraint, func
-from sqlalchemy.orm import Mapped, mapped_column
-from sqlalchemy.types import DateTime
-
-from .base import Base
-
-
-class PatreonSeenMedia(Base):
-    __tablename__ = "patreon_seen_media"
-    __table_args__ = (
-        # Dedup key the downloader upserts against: one ledger row per
-        # (source, media). A second sighting of the same media is a no-op.
-        UniqueConstraint("source_id", "filehash", name="uq_patreon_seen_media_source_id"),
-    )
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True)
-    source_id: Mapped[int] = mapped_column(
-        ForeignKey("source.id", ondelete="CASCADE"), nullable=False, index=True
-    )
-    filehash: Mapped[str] = mapped_column(String(128), nullable=False)
-    post_id: Mapped[str | None] = mapped_column(String(64), nullable=True)
-    seen_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), nullable=False, server_default=func.now()
-    )

backend/app/models/series_chapter.py

@@ -1,54 +0,0 @@
-"""SeriesChapter — an ordered chapter/part within a series.
-
-A series IS a Tag(kind='series'); a chapter groups ordered SeriesPages under it.
-Reading order is (chapter.chapter_number, series_page.page_number): chapter_number
-sets the order of chapters, page_number orders pages within a chapter.
-
-chapter_number is an ordering key only (not unique) — reorder rewrites 1..N
-wholesale, mirroring series_page.page_number, so a reorder can't transiently
-collide on a unique index.
-
-A chapter may be a placeholder (is_placeholder=True) — a reserved empty slot for
-a section the operator doesn't have yet; it holds no pages and shows as a gap in
-the reader. stated_page_start/end carry the page range parsed from the source
-post (FC-6.2), used to flag missing-page gaps; both are nullable when unknown.
-
-stated_part is the operator-facing "Part N" label (FC-6.4), separate from the
-positional chapter_number: chapter_number is auto-managed ordering (rewritten
-1..N on reorder/delete), while stated_part is the real installment number the
-operator types — e.g. a series authored from a post that is Part 2 of a story.
-Nullable when unset (the UI then falls back to showing chapter_number).
-"""
-
-from datetime import datetime
-
-from sqlalchemy import Boolean, DateTime, ForeignKey, Integer, Text, func
-from sqlalchemy.orm import Mapped, mapped_column
-
-from .base import Base
-
-
-class SeriesChapter(Base):
-    __tablename__ = "series_chapter"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True)
-    series_tag_id: Mapped[int] = mapped_column(
-        ForeignKey("tag.id", ondelete="CASCADE"), nullable=False, index=True
-    )
-    chapter_number: Mapped[int] = mapped_column(Integer, nullable=False)
-    stated_part: Mapped[int | None] = mapped_column(Integer, nullable=True)
-    title: Mapped[str | None] = mapped_column(Text, nullable=True)
-    is_placeholder: Mapped[bool] = mapped_column(
-        Boolean, nullable=False, server_default="false"
-    )
-    stated_page_start: Mapped[int | None] = mapped_column(Integer, nullable=True)
-    stated_page_end: Mapped[int | None] = mapped_column(Integer, nullable=True)
-    created_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), nullable=False, server_default=func.now()
-    )
-    updated_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True),
-        nullable=False,
-        server_default=func.now(),
-        onupdate=func.now(),
-    )

backend/app/models/series_page.py

@@ -1,12 +1,9 @@
 """SeriesPage — ordered image membership for a series-kind Tag.
 
-A series IS a Tag with kind='series'; series_page gives it ordered pages,
-grouped into chapters (FC-6). An image belongs to at most one series
-(UNIQUE image_id) ⇒ at most one chapter. Reading order is
-(chapter.chapter_number, series_page.page_number): page_number orders pages
-WITHIN a chapter and is an ordering key only (not unique) — reorder rewrites
-1..N wholesale. stated_page carries the page number parsed from the source
-post (FC-6.2), nullable when unknown.
+A series IS a Tag with kind='series'; series_page gives it ordered pages.
+An image belongs to at most one series (UNIQUE image_id). Cover = the
+lowest page_number. page_number is an ordering key only (not unique) —
+reorder rewrites 1..N wholesale.
 """
 
 from datetime import datetime
@@ -24,18 +21,12 @@ class SeriesPage(Base):
     series_tag_id: Mapped[int] = mapped_column(
         ForeignKey("tag.id", ondelete="CASCADE"), nullable=False, index=True
     )
-    chapter_id: Mapped[int] = mapped_column(
-        ForeignKey("series_chapter.id", ondelete="CASCADE"),
-        nullable=False,
-        index=True,
-    )
     image_id: Mapped[int] = mapped_column(
         ForeignKey("image_record.id", ondelete="CASCADE"),
         nullable=False,
         unique=True,
     )
     page_number: Mapped[int] = mapped_column(Integer, nullable=False)
-    stated_page: Mapped[int | None] = mapped_column(Integer, nullable=True)
     created_at: Mapped[datetime] = mapped_column(
         DateTime(timezone=True), nullable=False, server_default=func.now()
     )

backend/app/models/series_suggestion.py

@@ -1,55 +0,0 @@
-"""SeriesSuggestion — a confirm-only "this post may continue this series" hint.
-
-The matcher (FC-6.3) scores a (post, candidate series) pair from several weighted
-signals and, above the configured threshold, records a pending suggestion. The
-operator confirms (→ the post is added as a chapter) or dismisses it; FC never
-files a post into a series on its own. status is a plain string (no Postgres
-ENUM — see the check-existing-enums lesson): pending | added | dismissed.
-"""
-
-from datetime import datetime
-
-from sqlalchemy import (
-    JSON,
-    DateTime,
-    Float,
-    ForeignKey,
-    Integer,
-    String,
-    UniqueConstraint,
-    func,
-)
-from sqlalchemy.orm import Mapped, mapped_column
-
-from .base import Base
-
-
-class SeriesSuggestion(Base):
-    __tablename__ = "series_suggestion"
-    __table_args__ = (
-        UniqueConstraint(
-            "post_id", "series_tag_id", name="uq_series_suggestion_post_series"
-        ),
-    )
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True)
-    post_id: Mapped[int] = mapped_column(
-        ForeignKey("post.id", ondelete="CASCADE"), nullable=False, index=True
-    )
-    series_tag_id: Mapped[int] = mapped_column(
-        ForeignKey("tag.id", ondelete="CASCADE"), nullable=False, index=True
-    )
-    score: Mapped[float] = mapped_column(Float, nullable=False)
-    signals: Mapped[dict | None] = mapped_column(JSON, nullable=True)
-    status: Mapped[str] = mapped_column(
-        String(16), nullable=False, server_default="pending", index=True
-    )
-    created_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), nullable=False, server_default=func.now()
-    )
-    updated_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True),
-        nullable=False,
-        server_default=func.now(),
-        onupdate=func.now(),
-    )

backend/app/services/archive_extractor.py

@@ -16,45 +16,9 @@ log = logging.getLogger(__name__)
 
 ARCHIVE_EXTS = {".zip", ".cbz", ".rar", ".7z"}
 
-# Magic-byte signatures, so an archive with a mangled / extension-less filename
-# is still recognised. Patreon attachment download URLs sanitize to names like
-# `01_https___www.patreon.com_media-u_v3_131083093`, whose `Path.suffix` is junk
-# (`.com_media-u_v3_131083093`), never `.zip` — an extension-only gate filed
-# those as opaque PostAttachments and NEVER extracted them (operator-flagged
-# 2026-06-06). Detection is by extension first (cheap), then header sniff.
-_RAR_MAGIC = b"Rar!\x1a\x07"
-_7Z_MAGIC = b"7z\xbc\xaf\x27\x1c"
-
-
-def detect_archive_format(path: Path) -> str | None:
-    """Return "zip" | "rar" | "7z" for an archive, else None.
-
-    Trusts a known extension first, then falls back to magic-byte sniffing so a
-    mis-named or extension-less archive is still handled. (zip covers .cbz too.)
-    """
-    ext = Path(path).suffix.lower()
-    if ext in (".zip", ".cbz"):
-        return "zip"
-    if ext == ".rar":
-        return "rar"
-    if ext == ".7z":
-        return "7z"
-    try:
-        if zipfile.is_zipfile(path):
-            return "zip"
-        with open(path, "rb") as fh:
-            head = fh.read(8)
-    except OSError:
-        return None
-    if head.startswith(_RAR_MAGIC):
-        return "rar"
-    if head.startswith(_7Z_MAGIC):
-        return "7z"
-    return None
-
 
 def is_archive(path: Path) -> bool:
-    return detect_archive_format(path) is not None
+    return Path(path).suffix.lower() in ARCHIVE_EXTS
 
 
 @contextmanager
@@ -68,16 +32,16 @@ def extract_archive(path: Path):
     members: list[tuple[str, Path]] = []
     try:
         try:
-            fmt = detect_archive_format(path)
-            if fmt == "zip":
+            ext = Path(path).suffix.lower()
+            if ext in (".zip", ".cbz"):
                 with zipfile.ZipFile(path) as zf:
                     zf.extractall(base)
-            elif fmt == "rar":
+            elif ext == ".rar":
                 import rarfile
 
                 with rarfile.RarFile(path) as rf:
                     rf.extractall(base)
-            elif fmt == "7z":
+            elif ext == ".7z":
                 import py7zr
 
                 with py7zr.SevenZipFile(path, "r") as zf:

backend/app/services/backup_service.py

@@ -15,10 +15,7 @@ lifecycle + soft/hard time limits + retention bookkeeping.
 from __future__ import annotations
 
 import json
-import os
-import shutil
 import subprocess
-import tempfile
 from datetime import UTC, datetime
 from pathlib import Path
 
@@ -29,35 +26,6 @@ _BACKUPS_DIRNAME = "_backups"
 # blocking syscall ignores that signal. These bound the worst case.
 _DB_SUBPROCESS_TIMEOUT_S = 12 * 60        # 12 min (Celery soft is 10 min)
 _IMAGES_SUBPROCESS_TIMEOUT_S = 7 * 60 * 60  # 7 hr (Celery soft is 6 hr)
-# Grace after SIGKILL to reap the child. If it can't be reaped in this window
-# (an uninterruptible NFS D-state — the failure mode that wedged the
-# concurrency-1 maintenance lane for hours, operator-flagged 2026-06-07), we
-# STOP waiting and fail fast, freeing the worker slot. The orphan is reaped by
-# the OS once its blocking syscall clears.
-_KILL_REAP_GRACE_S = 10
-
-
-def _run_bounded(cmd: list[str], timeout: int) -> None:
-    """subprocess.run(check=True, timeout) whose reaper can't itself hang.
-
-    subprocess.run's timeout path SIGKILLs the child then blocks in wait() to
-    reap it — but a process stuck in uninterruptible I/O (NFS) can't be reaped,
-    so wait() blocks for hours. Here we bound the post-kill reap and re-raise
-    TimeoutExpired regardless, so the caller fails fast instead of wedging."""
-    proc = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
-    try:
-        out, err = proc.communicate(timeout=timeout)
-    except subprocess.TimeoutExpired:
-        proc.kill()
-        try:
-            proc.communicate(timeout=_KILL_REAP_GRACE_S)
-        except subprocess.TimeoutExpired:
-            pass  # unkillable (D-state) — abandon the reap, fail fast
-        raise
-    if proc.returncode != 0:
-        raise subprocess.CalledProcessError(
-            proc.returncode, cmd, output=out, stderr=err
-        )
 
 
 def _libpq_url(sa_url: str) -> str:
@@ -116,25 +84,14 @@ def backup_db(
     ts = _now_ts()
     out_dir = _backups_dir(images_root)
     sql_path = out_dir / f"fc_db_{ts}.sql"
-    # Dump to LOCAL disk first, then move the finished file to the (NFS) backups
-    # dir. pg_dump's long phase is then a DB-socket wait + local writes — both
-    # killable — instead of an NFS write that can hang uninterruptibly. Only the
-    # final move touches NFS, and it's a bounded single-file step.
-    fd, tmp_name = tempfile.mkstemp(prefix="fc_db_", suffix=".sql")
-    os.close(fd)
-    tmp_path = Path(tmp_name)
-    try:
-        _run_bounded(
-            [
-                "pg_dump", "--no-owner", "--no-acl",
-                "-f", str(tmp_path), _libpq_url(db_url),
-            ],
-            _DB_SUBPROCESS_TIMEOUT_S,
-        )
-        shutil.move(str(tmp_path), str(sql_path))
-    finally:
-        if tmp_path.exists():
-            tmp_path.unlink(missing_ok=True)
+    subprocess.run(
+        [
+            "pg_dump", "--no-owner", "--no-acl",
+            "-f", str(sql_path), _libpq_url(db_url),
+        ],
+        capture_output=True, check=True,
+        timeout=_DB_SUBPROCESS_TIMEOUT_S,
+    )
     manifest_path = _write_manifest(
         out_dir, kind="db", ts=ts, tag=tag, triggered_by=triggered_by,
         artifact_path=sql_path,
@@ -157,17 +114,15 @@ def backup_images(
     ts = _now_ts()
     out_dir = _backups_dir(images_root)
     tar_path = out_dir / f"fc_images_{ts}.tar.zst"
-    # No local-temp here (the archive is hundreds of GB — it can't stage in
-    # /tmp), but bounded-kill still applies so a tar wedged on NFS fails fast
-    # rather than holding the lane for hours.
-    _run_bounded(
+    subprocess.run(
         [
             "tar", "--zstd", "-cf", str(tar_path),
             "-C", str(images_root.parent), images_root.name,
             f"--exclude={images_root.name}/_backups",
             f"--exclude={images_root.name}/_quarantine",
         ],
-        _IMAGES_SUBPROCESS_TIMEOUT_S,
+        capture_output=True, check=True,
+        timeout=_IMAGES_SUBPROCESS_TIMEOUT_S,
     )
     manifest_path = _write_manifest(
         out_dir, kind="images", ts=ts, tag=tag, triggered_by=triggered_by,

backend/app/services/cleanup_service.py

@@ -11,8 +11,6 @@ the one-and-done GS/IR migration tooling.)
 """
 from __future__ import annotations
 
-import logging
-import time
 from datetime import UTC, datetime, timedelta
 from pathlib import Path
 from typing import Any
@@ -24,8 +22,6 @@ from ..models import Artist, ImageRecord, LibraryAuditRun, Tag
 from ..models.series_page import SeriesPage
 from ..models.tag import image_tag
 
-log = logging.getLogger(__name__)
-
 
 def project_artist_cascade(session: Session, *, slug: str) -> dict:
     """Read-only projection of what delete_artist_cascade would touch.
@@ -669,170 +665,3 @@ def cancel_audit_run(session: Session, *, audit_id: int) -> None:
         .where(LibraryAuditRun.status == "running")
         .values(status="cancelled", finished_at=datetime.now(UTC))
     )
-
-
-# -- archive-attachment re-extraction (#713 part 2) ------------------------
-
-_ARCHIVE_EXT_FOR_FORMAT = {"zip": ".zip", "rar": ".rar", "7z": ".7z"}
-
-
-def _reextract_archive_to_post(
-    importer, archive_path: Path, post, source_row, artist, images_root: Path,
-) -> list[int]:
-    """Extract one stored archive and link its members to `post`.
-
-    The stored attachment has no adjacent sidecar (it lives in the sha-addressed
-    attachment store). Stage a copy + a reconstructed sidecar UNDER the artist's
-    library dir (`images_root/<slug>/<platform>/<post>/`) — the importer
-    re-derives the artist from the path AND copies members relative to it, so the
-    members land in the real library and resolve to the right artist — then re-run
-    `attach_in_place`: the archive extracts and `find_or_create_post` re-attaches
-    the members to the SAME Post (source_id + external_post_id). Removes only the
-    staged archive + sidecar afterward; the imported member files stay. Returns
-    the new member image ids.
-    """
-    import json
-    import shutil
-
-    from .archive_extractor import detect_archive_format
-
-    fmt = detect_archive_format(archive_path)
-    ext = _ARCHIVE_EXT_FOR_FORMAT.get(fmt or "", ".zip")
-    platform = source_row.platform if source_row is not None else "imported"
-    sidecar = {
-        "category": source_row.platform if source_row is not None
-        else (post.raw_metadata or {}).get("category"),
-        "id": post.external_post_id,
-        "title": post.post_title or "",
-        "content": post.description or "",
-        "published_at": post.post_date.isoformat() if post.post_date else None,
-        "url": post.post_url,
-    }
-    work = images_root / artist.slug / platform / str(post.external_post_id)
-    work.mkdir(parents=True, exist_ok=True)
-    staged = work / f"archive{ext}"  # clean ext → is_archive + find_sidecar
-    sidecar_path = staged.with_suffix(".json")
-    try:
-        shutil.copy2(archive_path, staged)
-        sidecar_path.write_text(json.dumps(sidecar))
-        res = importer.attach_in_place(staged, artist=artist, source=source_row)
-        return list(res.member_image_ids or [])
-    finally:
-        # Drop only the staged archive + sidecar; the extracted member files
-        # were copied into the library alongside them and must stay.
-        staged.unlink(missing_ok=True)
-        sidecar_path.unlink(missing_ok=True)
-
-
-def reextract_archive_attachments(
-    session: Session,
-    *,
-    images_root: Path,
-    time_budget_seconds: float | None = None,
-    after_id: int = 0,
-) -> dict:
-    """Re-process existing PostAttachments that are ACTUALLY archives but were
-    filed opaquely before #713 part 1 (extension-only is_archive missed mangled /
-    extension-less Patreon attachment names). For each: extract the members,
-    import them, and link them to the attachment's post.
-
-    Idempotent — members dedupe by sha256, the archive dedupes by sha — so it's
-    safe to run repeatedly. Returns a summary dict for task_run.metadata.
-
-    Time-boxed + resumable: scans PostAttachments in ascending id order starting
-    after ``after_id``. When ``time_budget_seconds`` elapses, stops and reports
-    ``partial=True`` + ``resume_after_id`` (the last scanned id) so the task can
-    re-enqueue itself and continue — a large archive back-catalog can't run the
-    task into the Celery time limit or hog the maintenance lane. A bare re-run
-    (after_id=0) would never advance because an already-extracted archive is
-    still an archive on disk, so the cursor is what guarantees forward progress.
-    """
-    from ..models import ImportSettings, Post, PostAttachment, Source
-    from ..tasks.ml import tag_and_embed
-    from ..tasks.thumbnail import generate_thumbnail
-    from .archive_extractor import is_archive
-    from .importer import Importer
-    from .thumbnailer import Thumbnailer
-
-    summary = {
-        "scanned": 0, "archives": 0, "members_imported": 0,
-        "posts_touched": 0, "skipped_no_post": 0, "skipped_no_artist": 0,
-        "errors": 0, "partial": False, "resume_after_id": after_id,
-    }
-    settings = ImportSettings.load_sync(session)
-    importer = Importer(
-        session=session, images_root=images_root, import_root=images_root,
-        thumbnailer=Thumbnailer(images_root=images_root), settings=settings,
-    )
-
-    attachments = session.execute(
-        select(PostAttachment)
-        .where(PostAttachment.id > after_id)
-        .order_by(PostAttachment.id)
-    ).scalars().all()
-    enqueue_ids: list[int] = []
-    start = time.monotonic()
-    for att in attachments:
-        summary["scanned"] += 1
-        summary["resume_after_id"] = att.id
-        stored = Path(att.path)
-        try:
-            if not stored.is_file() or not is_archive(stored):
-                continue
-        except OSError:
-            continue
-        summary["archives"] += 1
-        if att.post_id is None:
-            summary["skipped_no_post"] += 1
-            continue
-        post = session.get(Post, att.post_id)
-        if post is None:
-            summary["skipped_no_post"] += 1
-            continue
-        artist = session.get(Artist, att.artist_id) if att.artist_id else None
-        if artist is None and post.artist_id:
-            artist = session.get(Artist, post.artist_id)
-        if artist is None or not artist.slug:
-            # The importer re-derives the artist from the staged path, so we need
-            # a real artist+slug to anchor under. (Shouldn't happen for
-            # subscription posts; skip rather than orphan the members.)
-            summary["skipped_no_artist"] += 1
-            continue
-        source_row = session.get(Source, post.source_id) if post.source_id else None
-        try:
-            ids = _reextract_archive_to_post(
-                importer, stored, post, source_row, artist, images_root,
-            )
-            session.commit()
-        except Exception as exc:  # one bad archive must not strand the rest
-            session.rollback()
-            summary["errors"] += 1
-            log.warning("re-extract failed for attachment %s: %s", att.id, exc)
-            continue
-        if ids:
-            summary["members_imported"] += len(ids)
-            summary["posts_touched"] += 1
-            enqueue_ids.extend(ids)
-
-        # Time-box the chunk. resume_after_id already points at this attachment,
-        # so the next run starts strictly after it. Checked after the commit so a
-        # half-extracted archive never straddles the boundary.
-        if (
-            time_budget_seconds is not None
-            and time.monotonic() - start >= time_budget_seconds
-        ):
-            summary["partial"] = True
-            break
-    else:
-        # Loop ran to exhaustion — nothing left to resume.
-        summary["partial"] = False
-
-    # Thumbnails + ML for the newly-imported members (best-effort; off the
-    # critical path — a Redis hiccup must not fail the whole re-extract).
-    for img_id in enqueue_ids:
-        try:
-            generate_thumbnail.delay(img_id)
-            tag_and_embed.delay(img_id)
-        except Exception as exc:
-            log.warning("re-extract enqueue failed for image %s: %s", img_id, exc)
-    return summary

backend/app/services/download_backends.py

@@ -1,241 +0,0 @@
-"""Platform → download-backend dispatch (one place that knows which platforms
-are served by the native FC ingester vs. the gallery-dl subprocess).
-
-gallery-dl wasn't built to be driven by an automated scheduler — no native
-checkpoint/resume, no structured logs, per-file HEADs that dominate wall-clock.
-The native ingester (services/patreon_ingester.py, plan #697) replaces it for
-Patreon and is the path we grow as more platforms migrate. To keep that
-migration DRY, every caller that has to behave differently per backend —
-download routing, the credential-verify probe, cursor handling — asks THIS
-module instead of testing ``platform == "patreon"`` inline. When a platform gets
-a native ingester, it moves into ``NATIVE_INGESTER_PLATFORMS`` here and both the
-download path and verify switch over together.
-
-The backend surfaces share a UNIFORM signature so a caller invokes the same
-function regardless of platform:
-  - verify_credential(...)  → (ok: bool|None, message: str)
-  - (download stays in download_service for now; uses_native_ingester() is the
-    shared predicate it routes on, so the decision lives here too.)
-"""
-
-from __future__ import annotations
-
-import asyncio
-from pathlib import Path
-
-from .gallery_dl import DownloadResult, ErrorType
-from .patreon_ingester import PatreonIngester
-from .patreon_resolver import extract_vanity, resolve_campaign_id_for_source
-
-# Platforms whose download + verify go through the native ingester rather than
-# gallery-dl. gallery-dl still serves every other platform (subscribestar,
-# hentaifoundry, discord, pixiv, deviantart) unchanged.
-NATIVE_INGESTER_PLATFORMS = frozenset({"patreon"})
-
-# Mirrors patreon_resolver._CAMPAIGNS_URL — surfaced in resolution-failure
-# messages so the operator sees the exact lookup endpoint that was hit.
-_CAMPAIGNS_API = "https://www.patreon.com/api/campaigns"
-
-
-def uses_native_ingester(platform: str) -> bool:
-    """True when `platform` is served by the native ingester (not gallery-dl).
-    The single predicate the download path and verify both route on."""
-    return platform in NATIVE_INGESTER_PLATFORMS
-
-
-async def run_download(
-    *,
-    ctx: dict,
-    source_config,
-    skip_value: bool | str,
-    mode: str | None,
-    gdl,
-    sync_session_factory,
-) -> tuple[DownloadResult, str | None]:
-    """Uniform download across backends — the download counterpart to
-    `verify_source_credential`, so this module is the ONE place that knows how
-    each platform both downloads AND verifies (the seam that makes adding a
-    platform a bounded job).
-
-    Returns `(DownloadResult, resolved_campaign_id)`; `resolved_campaign_id` is
-    non-None only when a native vanity lookup ran this call (so phase 3 caches
-    it). Native platforms route through their ingester in `mode`
-    (tick/backfill/recovery); gallery-dl platforms run the subprocess. The caller
-    (download_service) prepares `source_config`/`skip_value`/`mode` from the
-    backfill state machine and owns phase 3.
-    """
-    platform = ctx["platform"]
-    if uses_native_ingester(platform):
-        return await _run_native_ingester(
-            ctx, source_config, mode, gdl, sync_session_factory
-        )
-    result = await gdl.download(
-        url=ctx["url"],
-        artist_slug=ctx["artist_slug"],
-        platform=platform,
-        source_config=source_config,
-        cookies_path=ctx["cookies_path"],
-        auth_token=ctx["auth_token"],
-        skip_value=skip_value,
-    )
-    return result, None
-
-
-async def _run_native_ingester(
-    ctx: dict, source_config, mode: str | None, gdl, sync_session_factory,
-) -> tuple[DownloadResult, str | None]:
-    """Patreon (today the only native platform): resolve the campaign id, then run
-    the native ingester in a worker thread (it is sync requests/subprocess).
-
-    `resolved_campaign_id` is non-None only when we had to look it up from the
-    vanity URL this run, so phase 3 caches it the way the old gallery-dl retry
-    did. A campaign id we cannot resolve is a loud NOT_FOUND — never a silent
-    empty success.
-    """
-    overrides = ctx["config_overrides"] or {}
-    campaign_id, resolved_campaign_id = await resolve_campaign_id_for_source(
-        ctx["url"], ctx["cookies_path"], overrides
-    )
-    if not campaign_id:
-        url = ctx["url"]
-        vanity = extract_vanity(url)
-        return (
-            DownloadResult(
-                success=False,
-                url=url,
-                artist_slug=ctx["artist_slug"],
-                platform="patreon",
-                error_type=ErrorType.NOT_FOUND,
-                error_message=(
-                    f"Could not resolve Patreon campaign id. source_url={url!r}; "
-                    f"vanity={vanity!r}; "
-                    f"lookup=GET {_CAMPAIGNS_API}?filter[vanity]={vanity or ''} "
-                    "(vanity lookup failed — cookies expired or creator moved?)"
-                ),
-            ),
-            None,
-        )
-
-    # Honor the operator's existing rate-limit knobs on the native path (plan
-    # #703): the global download_rate_limit_seconds (gallery-dl's `rate_limit`,
-    # here on the gdl service) paces media downloads; page fetches use the
-    # per-source sleep_request override, else `max(0.5, rate_limit/4)` — the same
-    # API-pacing default gallery-dl applied as its `sleep-request`.
-    rate_limit = gdl._rate_limit
-    request_sleep = (
-        source_config.sleep_request
-        if source_config.sleep_request is not None
-        else max(0.5, rate_limit / 4)
-    )
-    ingester = PatreonIngester(
-        images_root=gdl.images_root,
-        cookies_path=ctx["cookies_path"],
-        session_factory=sync_session_factory,
-        validate=gdl._validate_files,
-        rate_limit=rate_limit,
-        request_sleep=request_sleep,
-    )
-    loop = asyncio.get_running_loop()
-    dl_result = await loop.run_in_executor(
-        None,
-        lambda: ingester.run(
-            source_id=ctx["source_id"],
-            campaign_id=campaign_id,
-            artist_slug=ctx["artist_slug"],
-            url=ctx["url"],
-            mode=mode,
-            resume_cursor=source_config.resume_cursor,
-            time_budget_seconds=source_config.timeout,
-            posts_base=int(overrides.get("_backfill_posts", 0)),
-            # plan #709: live progress writes to this running event mid-walk.
-            event_id=ctx.get("event_id"),
-        ),
-    )
-    return dl_result, resolved_campaign_id
-
-
-async def preview_source(
-    *,
-    platform: str,
-    url: str,
-    source_id: int,
-    config_overrides: dict | None,
-    cookies_path: str | None,
-    images_root: Path,
-    sync_session_factory,
-    page_limit: int = 3,
-) -> dict:
-    """Dry-run preview for a native platform (plan #708 B4): resolve the campaign
-    id, then walk a few pages counting media not already seen/dead — no download.
-
-    Returns the preview dict (total_new / posts_scanned / pages_scanned /
-    has_more / sample), or `{"error": msg}` on a resolve / auth / drift failure.
-    Native-only — the caller gates on `uses_native_ingester`.
-    """
-    import asyncio
-
-    from .patreon_client import PatreonAPIError
-
-    campaign_id, _ = await resolve_campaign_id_for_source(
-        url, cookies_path, config_overrides or {}
-    )
-    if not campaign_id:
-        vanity = extract_vanity(url)
-        return {
-            "error": (
-                f"Couldn't resolve the campaign id. source_url={url!r}; "
-                f"vanity={vanity!r}; lookup=GET {_CAMPAIGNS_API}?filter[vanity]={vanity or ''} "
-                "(cookies expired, or the creator moved/renamed?)."
-            )
-        }
-    ingester = PatreonIngester(
-        images_root=images_root,
-        cookies_path=cookies_path,
-        session_factory=sync_session_factory,
-    )
-    loop = asyncio.get_running_loop()
-    try:
-        result = await loop.run_in_executor(
-            None,
-            lambda: ingester.preview(source_id, campaign_id, page_limit=page_limit),
-        )
-    except PatreonAPIError as exc:
-        return {"error": f"Couldn't preview: {exc}"}
-    return result
-
-
-async def verify_source_credential(
-    *,
-    platform: str,
-    url: str,
-    artist_slug: str,
-    config_overrides: dict | None,
-    cookies_path: str | None,
-    auth_token: str | None,
-    images_root: Path,
-) -> tuple[bool | None, str]:
-    """Uniform credential probe across backends. Returns `(ok, message)`:
-    True = authenticated, False = rejected, None = inconclusive (drift /
-    network / nothing to test). Callers don't branch on platform — they call
-    this and render the result.
-    """
-    if uses_native_ingester(platform):
-        # Native ingester platforms verify via their own lightweight auth probe
-        # (resolve campaign id + one authenticated API page). Patreon today.
-        from .patreon_ingester import verify_patreon_credential
-
-        return await verify_patreon_credential(url, cookies_path, config_overrides)
-
-    # gallery-dl platforms: --simulate one item; the extractor errors before it
-    # can list if auth is bad.
-    from .gallery_dl import GalleryDLService, SourceConfig
-
-    gdl = GalleryDLService(images_root=images_root)
-    return await gdl.verify(
-        url=url,
-        artist_slug=artist_slug,
-        platform=platform,
-        source_config=SourceConfig.from_dict(config_overrides or {}),
-        cookies_path=cookies_path,
-        auth_token=auth_token,
-    )

backend/app/services/download_service.py

@@ -13,6 +13,7 @@ from __future__ import annotations
 
 import asyncio
 import logging
+import re
 from datetime import UTC, datetime
 from pathlib import Path
 from typing import Any
@@ -24,25 +25,44 @@ from sqlalchemy.orm import joinedload
 
 from ..models import Artist, DownloadEvent, Source
 from .credential_service import CredentialService
-from .download_backends import run_download, uses_native_ingester
 from .gallery_dl import (
-    BACKFILL_CHUNK_SECONDS,
     BACKFILL_SKIP_VALUE,
+    BACKFILL_TIMEOUT_SECONDS,
     TICK_SKIP_VALUE,
-    DownloadResult,
     ErrorType,
     GalleryDLService,
     SourceConfig,
-    extract_errors_warnings,
-    truncate_log,
 )
 from .importer import Importer
+from .patreon_resolver import resolve_campaign_id
 from .platforms import auth_type_for
 from .scheduler_service import set_platform_cooldown
 
 log = logging.getLogger(__name__)
 
 
+_PATREON_VANITY_RE = re.compile(
+    r"^https?://(?:www\.)?patreon\.com/(?:c/)?(?!id:)([^/?#]+)",
+    re.IGNORECASE,
+)
+_CAMPAIGN_ID_FAILURE_PATTERN = "failed to extract campaign id"
+
+
+def _extract_patreon_vanity(url: str) -> str | None:
+    m = _PATREON_VANITY_RE.match(url)
+    return m.group(1) if m else None
+
+
+def _looks_like_campaign_id_failure(stdout: str, stderr: str) -> bool:
+    return _CAMPAIGN_ID_FAILURE_PATTERN in f"{stdout}\n{stderr}".lower()
+
+
+def _effective_url(platform: str, source_url: str, overrides: dict) -> str:
+    if platform == "patreon" and overrides.get("patreon_campaign_id"):
+        return f"https://www.patreon.com/id:{overrides['patreon_campaign_id']}"
+    return source_url
+
+
 class DownloadService:
     """Async orchestrator. The Celery task runs `asyncio.run(svc.download_source(N))`.
 
@@ -57,18 +77,12 @@ class DownloadService:
         gdl: GalleryDLService,
         importer: Importer,
         cred_service: CredentialService,
-        sync_session_factory=None,
     ):
         self.async_session = async_session
         self.sync_session = sync_session
         self.gdl = gdl
         self.importer = importer
         self.cred_service = cred_service
-        # Sync sessionmaker the native Patreon ingester opens SHORT-LIVED
-        # sessions from for its seen-ledger reads/writes (never one held across
-        # the multi-minute walk — see PatreonIngester). Only the patreon branch
-        # of phase 2 uses it; gallery-dl sources leave it None.
-        self.sync_session_factory = sync_session_factory
 
     async def download_source(self, source_id: int) -> int:
         """Returns DownloadEvent.id. Idempotent: in-flight events are returned as-is."""
@@ -93,85 +107,63 @@ class DownloadService:
         self.sync_session.close()
 
         source_config = SourceConfig.from_dict(ctx["config_overrides"] or {})
-        # Backfill mode (plan #693): the source's `_backfill_state == "running"`
-        # selects a time-boxed deep-walk chunk — skip: True (walk full history)
-        # + the BACKFILL_CHUNK_SECONDS budget, resuming from the cursor
-        # checkpoint (plan #689). State is the single source of truth; it stays
-        # "running" across ticks until the walk reaches the bottom (phase 3
-        # flips it to "complete"). Otherwise tick mode: exit gallery-dl after
-        # 20 contiguous archived items (skip: "exit:20" + the default 870s).
-        # Operator drives this via POST /api/sources/{id}/backfill {action}.
-        overrides = ctx["config_overrides"] or {}
-        in_backfill = overrides.get("_backfill_state") == "running"
-        # Recovery (plan #697) reuses the entire #693 backfill state machine —
-        # cursor checkpoint, time-boxed chunks, complete/stall lifecycle — and
-        # differs only in bypassing the tier-1 seen-ledger so dropped-and-deleted
-        # near-dups get re-fetched and re-evaluated under the CURRENT pHash
-        # threshold (tier-2 disk still spares files we kept). The
-        # `_backfill_bypass_seen` flag rides alongside the running backfill state;
-        # download mode is "recovery" when both are set.
-        bypass_seen = bool(overrides.get("_backfill_bypass_seen"))
-        if in_backfill:
+        # alembic 0031 / plan #544: derive skip_value + timeout from the
+        # source's backfill_runs_remaining counter. When > 0, walk the full
+        # post history (skip: True + 1170s); when 0, exit gallery-dl after
+        # 20 contiguous archived items (skip: "exit:20" + the default
+        # 870s). Operator sets backfill via POST /api/sources/{id}/backfill.
+        backfill_remaining = ctx.get("backfill_runs_remaining", 0) or 0
+        if backfill_remaining > 0:
             skip_value: bool | str = BACKFILL_SKIP_VALUE
-            source_config.timeout = BACKFILL_CHUNK_SECONDS
-            pending_cursor = overrides.get("_backfill_cursor")
-            if uses_native_ingester(ctx["platform"]) and pending_cursor:
-                source_config.resume_cursor = pending_cursor
+            source_config.timeout = BACKFILL_TIMEOUT_SECONDS
         else:
             skip_value = TICK_SKIP_VALUE
 
-        # Phase 2 dispatch is uniform across backends (download_backends.
-        # run_download — the download counterpart to verify_source_credential):
-        # native platforms run their ingester in `mode` (zero per-file HEADs,
-        # native cursor/resume, loud drift detection); gallery-dl platforms run
-        # the subprocess. Either returns a DownloadResult-shaped object so phase 3
-        # is untouched. `mode` is None for gallery-dl (run_download ignores it).
-        mode: str | None = None
-        if uses_native_ingester(ctx["platform"]):
-            if in_backfill and bypass_seen:
-                mode = "recovery"
-            elif in_backfill:
-                mode = "backfill"
-            else:
-                mode = "tick"
-        dl_result, resolved_campaign_id = await self._run_download(
-            ctx=ctx, source_config=source_config, skip_value=skip_value, mode=mode,
+        effective_url = _effective_url(
+            ctx["platform"], ctx["url"], ctx["config_overrides"] or {}
         )
 
-        # A backfill chunk that hit its time-box but made forward progress is
-        # NORMAL, not a failure — reclassify TIMEOUT → PARTIAL so it reads as
-        # "ok/progress" (PARTIAL maps to status "ok"), not a red error, and the
-        # next chunk just resumes from the new cursor (plan #693). A chunk that
-        # timed out with NO progress stays TIMEOUT and feeds phase 3's
-        # stall-guard. Leave RATE_LIMITED alone so the platform-cooldown fires.
-        if in_backfill and dl_result.error_type == ErrorType.TIMEOUT:
-            new_cursor = dl_result.cursor  # plan #704: structured, not scraped
-            advanced = bool(
-                (new_cursor and new_cursor != overrides.get("_backfill_cursor"))
-                or dl_result.files_downloaded > 0
-            )
-            if advanced:
-                dl_result.error_type = ErrorType.PARTIAL
-                dl_result.error_message = (
-                    f"Backfill chunk: {dl_result.files_downloaded} file(s) — continuing"
+        dl_result = await self.gdl.download(
+            url=effective_url,
+            artist_slug=ctx["artist_slug"],
+            platform=ctx["platform"],
+            source_config=source_config,
+            cookies_path=ctx["cookies_path"],
+            auth_token=ctx["auth_token"],
+            skip_value=skip_value,
+        )
+
+        resolved_campaign_id: str | None = None
+        if (
+            ctx["platform"] == "patreon"
+            and not dl_result.success
+            and "patreon_campaign_id" not in (ctx["config_overrides"] or {})
+            and _looks_like_campaign_id_failure(dl_result.stdout, dl_result.stderr)
+        ):
+            vanity = _extract_patreon_vanity(ctx["url"])
+            if vanity:
+                log.info(
+                    "Attempting campaign-ID resolution for %s (%s)",
+                    ctx["artist_slug"], vanity,
                 )
+                resolved_campaign_id = await resolve_campaign_id(
+                    vanity, ctx["cookies_path"]
+                )
+                if resolved_campaign_id:
+                    dl_result = await self.gdl.download(
+                        url=f"https://www.patreon.com/id:{resolved_campaign_id}",
+                        artist_slug=ctx["artist_slug"],
+                        platform=ctx["platform"],
+                        source_config=source_config,
+                        cookies_path=ctx["cookies_path"],
+                        auth_token=ctx["auth_token"],
+                        skip_value=skip_value,
+                    )
 
         return await self._phase3_persist(
             ctx["event_id"], ctx, dl_result, resolved_campaign_id,
         )
 
-    async def _run_download(
-        self, *, ctx: dict, source_config, skip_value, mode: str | None,
-    ) -> tuple[DownloadResult, str | None]:
-        """Phase-2 dispatch → `download_backends.run_download`, passing this
-        service's gdl + sync sessionmaker. Kept as a thin instance method so tests
-        can stub the whole phase-2 dispatch on the service, and so the per-backend
-        construction lives in download_backends (the backend registry)."""
-        return await run_download(
-            ctx=ctx, source_config=source_config, skip_value=skip_value, mode=mode,
-            gdl=self.gdl, sync_session_factory=self.sync_session_factory,
-        )
-
     async def _phase1_setup(self, source_id: int) -> dict[str, Any]:
         source = (await self.async_session.execute(
             select(Source).options(joinedload(Source.artist)).where(Source.id == source_id)
@@ -269,11 +261,6 @@ class DownloadService:
         source_row = self.sync_session.get(Source, ctx["source_id"])
 
         import_summary = {"attached": 0, "skipped": 0, "errors": 0}
-        # Archives detected but captured WITHOUT extracting any image (probe
-        # rejected / corrupt / missing extractor backend). Surfaced on the event
-        # so a post showing "no images" beside a zip is diagnosable (plan
-        # follow-up 2026-06-06 — the recurring archive-association report).
-        unextracted_archives: list[dict] = []
         bytes_downloaded = 0
 
         loop = asyncio.get_running_loop()
@@ -327,17 +314,6 @@ class DownloadService:
                 # mirror duplicate_hash cleanup so we don't keep two copies.
                 # Operator-flagged 2026-06-02 (Lustria OST zip).
                 import_summary["attached"] += 1
-                # An archive captured WITHOUT extracting any image carries a
-                # reason on result.error — record it so the event explains the
-                # "no images beside a zip" symptom instead of staying silent.
-                if result.error:
-                    unextracted_archives.append(
-                        {"file": path.name, "reason": result.error}
-                    )
-                    log.warning(
-                        "archive captured unextracted (%s): %s",
-                        path.name, result.error,
-                    )
                 try:
                     bytes_downloaded += path.stat().st_size  # noqa: ASYNC240
                 except OSError:
@@ -384,16 +360,11 @@ class DownloadService:
             select(DownloadEvent).where(DownloadEvent.id == event_id)
         )).scalar_one()
 
-        # plan #704: the native ingester returns structured run_stats; only the
-        # gallery-dl path needs the regex-over-stdout reconstruction.
-        if dl_result.run_stats is not None:
-            run_stats = dict(dl_result.run_stats)
-        else:
-            run_stats = self.gdl._compute_run_stats(
-                dl_result.return_code, dl_result.stdout, dl_result.stderr
-            )
+        run_stats = self.gdl._compute_run_stats(
+            dl_result.return_code, dl_result.stdout, dl_result.stderr
+        )
         run_stats["quarantined_count"] = dl_result.files_quarantined
-        stderr_summary = extract_errors_warnings(dl_result.stderr)
+        stderr_summary = self.gdl._extract_errors_warnings(dl_result.stderr)
 
         # Plan #544: PARTIAL means the run downloaded ≥1 file but the
         # subprocess didn't finish in budget (typically wall-clock timeout
@@ -413,116 +384,50 @@ class DownloadService:
         ev.metadata_ = {
             "run_stats": run_stats,
             "error_type": dl_result.error_type.value if dl_result.error_type else None,
-            "stdout": truncate_log(dl_result.stdout) or None,
-            "stderr": truncate_log(dl_result.stderr) or None,
+            "stdout": self.gdl._truncate_log(dl_result.stdout) or None,
+            "stderr": self.gdl._truncate_log(dl_result.stderr) or None,
             "stderr_errors_warnings": stderr_summary or None,
             "duration_seconds": dl_result.duration_seconds,
             "quarantined_paths": dl_result.quarantined_paths or None,
             "import_summary": import_summary,
-            # Archives detected but captured without extracting an image — the
-            # recurring "post shows a zip but no images" report. Each entry is
-            # {file, reason}; None when every archive extracted cleanly.
-            "unextracted_archives": unextracted_archives or None,
         }
         await self._update_source_health(
             source_id=ctx["source_id"], status=status, error_message=ev.error,
             error_type=dl_result.error_type.value if dl_result.error_type else None,
-            retry_after_seconds=getattr(dl_result, "retry_after_seconds", None),
         )
-        await self._apply_backfill_lifecycle(ctx, dl_result)
+        # Plan #544: backfill lifecycle — auto-complete when a clean
+        # backfill run drained the queue (gallery-dl exited 0 + zero files
+        # downloaded means there was nothing to fetch); otherwise decrement
+        # the counter. Next tick falls back to tick mode once it hits 0.
+        #
+        # Audit 2026-06-02 gating: VALIDATION_FAILED also exits the
+        # subprocess with return_code=0 and files_downloaded=0 (every
+        # file was quarantined), which used to match the auto-complete
+        # predicate exactly — zeroing the operator's armed budget on
+        # the FIRST quarantine run instead of decrementing. Require
+        # dl_result.success + no error_type so only genuinely-empty
+        # successful runs drain the counter.
+        backfill_remaining = ctx.get("backfill_runs_remaining", 0) or 0
+        if backfill_remaining > 0:
+            src = (await self.async_session.execute(
+                select(Source).where(Source.id == ctx["source_id"])
+            )).scalar_one()
+            queue_drained = (
+                dl_result.success
+                and dl_result.error_type is None
+                and dl_result.return_code == 0
+                and dl_result.files_downloaded == 0
+            )
+            if queue_drained:
+                src.backfill_runs_remaining = 0
+            else:
+                src.backfill_runs_remaining = max(0, backfill_remaining - 1)
         await self.async_session.commit()
         return event_id
 
-    async def _apply_backfill_lifecycle(self, ctx: dict, dl_result) -> None:
-        """Backfill state machine (plan #693, building on the cursor of #689).
-
-        A backfill runs in time-boxed chunks while
-        `config_overrides["_backfill_state"] == "running"`. Each chunk:
-          - COMPLETES the walk → clean rc=0 (gallery-dl with skip:True exits 0
-            only after exhausting the newest→oldest walk; a chunk cut short by
-            its time-box returns success=False / rc<0 via TimeoutExpired). On
-            completion: state="complete", clear the cursor, return to tick mode.
-          - made PROGRESS (cursor advanced and/or files written) → stay
-            "running", checkpoint the new cursor, bump the chunk counter, and
-            spend one of the safety-cap chunks (backfill_runs_remaining). If the
-            cap is exhausted without finishing → state="stalled".
-          - made NO progress → increment the stall counter; two strikes →
-            state="stalled", clear the cursor (a wedged walk can't loop).
-
-        State is the single source of truth; the cursor lives only inside a
-        running backfill. config_overrides is reassigned (not mutated in place)
-        for SQLAlchemy JSON change detection.
-        """
-        overrides = ctx["config_overrides"] or {}
-        if overrides.get("_backfill_state") != "running":
-            return  # not backfilling — tick mode, nothing to do
-
-        old_cursor = overrides.get("_backfill_cursor")
-        cap_remaining = ctx.get("backfill_runs_remaining", 0) or 0
-
-        src = (await self.async_session.execute(
-            select(Source).where(Source.id == ctx["source_id"])
-        )).scalar_one()
-        new_overrides = dict(src.config_overrides or {})
-        chunks = int(new_overrides.get("_backfill_chunks", 0)) + 1
-        new_overrides["_backfill_chunks"] = chunks
-        # plan #704 (#5): _backfill_posts (the live progress badge) is OWNED by the
-        # ingester now — it writes a monotonic absolute mid-walk at each page
-        # boundary (ingest_core._checkpoint_posts), so the badge climbs DURING a
-        # chunk instead of jumping once per chunk here, and the re-walked resume
-        # page is no longer double-counted. new_overrides (read fresh above)
-        # carries the ingester's committed value forward untouched.
-
-        completed = (
-            dl_result.success
-            and dl_result.error_type is None
-            and dl_result.return_code == 0
-        )
-        if completed:
-            new_overrides["_backfill_state"] = "complete"
-            new_overrides.pop("_backfill_cursor", None)
-            new_overrides.pop("_backfill_cursor_stalls", None)
-            # plan #697: a recovery walk shares this lifecycle; clear its bypass
-            # flag on completion so the next routine tick honors the seen-ledger.
-            new_overrides.pop("_backfill_bypass_seen", None)
-            src.config_overrides = new_overrides
-            src.backfill_runs_remaining = 0
-            return
-
-        # Did not finish. The native ingester checkpoints + resumes via cursor
-        # (carried structurally on the result, plan #704); gallery-dl platforms
-        # have no resumable cursor (every chunk re-walks from the top), so they
-        # advance only by the download archive growing.
-        new_cursor = (
-            dl_result.cursor if uses_native_ingester(ctx["platform"]) else None
-        )
-        advanced = bool(
-            (new_cursor and new_cursor != old_cursor)
-            or dl_result.files_downloaded > 0
-        )
-        if advanced:
-            if new_cursor:
-                new_overrides["_backfill_cursor"] = new_cursor
-            new_overrides.pop("_backfill_cursor_stalls", None)
-            cap_remaining = max(0, cap_remaining - 1)
-            src.backfill_runs_remaining = cap_remaining
-            if cap_remaining == 0:
-                # Safety cap hit before reaching the bottom — pause, don't loop.
-                new_overrides["_backfill_state"] = "stalled"
-        else:
-            stalls = int(new_overrides.get("_backfill_cursor_stalls", 0)) + 1
-            if stalls >= 2:
-                new_overrides["_backfill_state"] = "stalled"
-                new_overrides.pop("_backfill_cursor", None)
-                new_overrides.pop("_backfill_cursor_stalls", None)
-            else:
-                new_overrides["_backfill_cursor_stalls"] = stalls
-
-        src.config_overrides = new_overrides
-
     async def _update_source_health(
         self, *, source_id: int, status: str, error_message: str | None,
-        error_type: str | None = None, retry_after_seconds: float | None = None,
+        error_type: str | None = None,
     ) -> None:
         """FC-3d: update Source.{consecutive_failures, last_error, last_checked_at}.
 
@@ -554,17 +459,7 @@ class DownloadService:
             # by error class without opening Logs per row.
             source.error_type = error_type
             if error_type == "rate_limited":
-                # plan #708 B1: honor the server's Retry-After when the native
-                # client surfaced one (clamped to a sane [60, 3600] window so a
-                # tiny hint can't leave the platform effectively un-cooled and a
-                # huge one can't strand it for hours); else the flat default.
-                if retry_after_seconds is not None:
-                    seconds = int(min(max(retry_after_seconds, 60), 3600))
-                    await set_platform_cooldown(
-                        self.async_session, source.platform, seconds=seconds,
-                    )
-                else:
-                    await set_platform_cooldown(self.async_session, source.platform)
+                await set_platform_cooldown(self.async_session, source.platform)
         elif status == "skipped":
             source.last_error = None
         source.last_checked_at = now

backend/app/services/extension_service.py

@@ -16,7 +16,7 @@ from sqlalchemy.ext.asyncio import AsyncSession
 
 from ..models import Artist, Source
 from ..utils.slug import slugify
-from .source_service import BACKFILL_MAX_CHUNKS
+from .source_service import NEW_SOURCE_BACKFILL_RUNS
 
 
 class UnknownPlatformError(Exception):
@@ -205,17 +205,16 @@ class ExtensionService:
             return existing, False
         sp = await self.session.begin_nested()
         try:
-            # New subscription sources arm run-until-done backfill (plan #693)
-            # so the first ticks walk the full history (otherwise gallery-dl's
-            # exit:20 short-circuits before the archive is built). Mirrors
-            # SourceService.create — without it, Firefox quick-add on a creator
-            # with >20 unsynced posts would surface as "check failed" with no
-            # diagnosis. Audit 2026-06-02.
+            # New subscription sources arm a few backfill runs so the
+            # first ticks walk the full history (otherwise gallery-dl's
+            # exit:20 short-circuits before the archive is built).
+            # Mirrors SourceService.create — without it, Firefox quick-
+            # add on a creator with >20 unsynced posts would surface
+            # as "check failed" with no diagnosis. Audit 2026-06-02.
             src = Source(
                 artist_id=artist_id, platform=platform,
                 url=url, enabled=True,
-                config_overrides={"_backfill_state": "running"},
-                backfill_runs_remaining=BACKFILL_MAX_CHUNKS,
+                backfill_runs_remaining=NEW_SOURCE_BACKFILL_RUNS,
             )
             self.session.add(src)
             await self.session.flush()

backend/app/services/file_validator.py

@@ -11,15 +11,9 @@ expected to write.
 
 from __future__ import annotations
 
-import json
-import logging
-import shutil
 from dataclasses import dataclass
-from datetime import UTC, datetime
 from pathlib import Path
 
-log = logging.getLogger(__name__)
-
 JPEG_HEAD = b"\xff\xd8\xff"
 JPEG_TAIL = b"\xff\xd9"
 
@@ -114,52 +108,3 @@ def validate_file(path: Path) -> ValidationResult:
         return ValidationResult(ok=True, format="webp", size=size)
 
     return ValidationResult(ok=True, format=None, size=size)
-
-
-def quarantine_file(
-    images_root: Path,
-    path: Path,
-    artist_slug: str,
-    platform: str,
-    *,
-    url: str | None,
-    result: ValidationResult,
-) -> Path | None:
-    """Move a validation-failed file to `_quarantine/<slug>/<platform>` and write
-    a `.quarantine.json` provenance sidecar next to it.
-
-    Returns the destination path, or None if the move itself failed (file left
-    in place — the caller decides what to report). The caller has already run
-    `validate_file` and seen `result.ok is False`. ONE implementation for both
-    download backends — gallery-dl's batch post-process and the native ingester's
-    per-media path — so the quarantine layout + provenance sidecar can't drift.
-    """
-    quarantine_root = images_root / "_quarantine" / artist_slug / platform
-    try:
-        quarantine_root.mkdir(parents=True, exist_ok=True)
-        dest = quarantine_root / path.name
-        counter = 1
-        while dest.exists():
-            dest = quarantine_root / f"{path.stem}.{counter}{path.suffix}"
-            counter += 1
-        shutil.move(str(path), str(dest))
-        sidecar = dest.with_suffix(dest.suffix + ".quarantine.json")
-        sidecar.write_text(
-            json.dumps(
-                {
-                    "original_path": str(path),
-                    "source_url": url,
-                    "artist_slug": artist_slug,
-                    "platform": platform,
-                    "format": result.format,
-                    "reason": result.reason,
-                    "size": result.size,
-                    "quarantined_at": datetime.now(UTC).isoformat(),
-                },
-                indent=2,
-            )
-        )
-    except OSError as exc:
-        log.error("Failed to quarantine %s: %s. File left in place.", path, exc)
-        return None
-    return dest

backend/app/services/gallery_dl.py

@@ -22,7 +22,7 @@ from datetime import UTC, datetime
 from enum import StrEnum
 from pathlib import Path
 
-from .file_validator import is_validatable, quarantine_file, validate_file
+from .file_validator import is_validatable, validate_file
 
 log = logging.getLogger(__name__)
 
@@ -39,12 +39,6 @@ class ErrorType(StrEnum):
     HTTP_ERROR = "http_error"
     UNSUPPORTED_URL = "unsupported_url"
     VALIDATION_FAILED = "validation_failed"
-    # Native Patreon ingester only (plan #697): a response parsed as JSON but
-    # didn't match the JSON:API shape the ingester depends on (missing `data`,
-    # media lacking `file_name`/`url`, etc.). Distinct from AUTH_ERROR — the fix
-    # is updating the ingester's field-set/parser, not rotating credentials. The
-    # contract test guards against silently shipping this.
-    API_DRIFT = "api_drift"
     # Run made real progress (downloaded ≥1 file) but did not finish in the
     # subprocess budget. Distinct from UNKNOWN_ERROR — the downstream status
     # mapping classifies this as "ok" because the next tick continues.
@@ -61,24 +55,25 @@ class ErrorType(StrEnum):
 # 20 contiguous HEADs is still negligible.
 TICK_SKIP_VALUE = "exit:20"
 
-# Backfill mode (operator-triggered deep scan): walk the full history,
-# one TIME-BOXED CHUNK per run (plan #693). config_overrides["_backfill_state"]
-# == "running" selects this mode; the cursor checkpoint (plan #689) lets each
-# chunk resume where the last stopped, so the walk advances across chunks
-# until gallery-dl exits cleanly (= reached the bottom).
+# Backfill mode (operator-triggered deep scan): walk the full history.
+# Source.backfill_runs_remaining > 0 selects this mode; the longer
+# timeout below absorbs creators with thousands of posts.
 #
-# The chunk budget is deliberately FAR below download_source's Celery
-# soft_time_limit (DOWNLOAD_SOFT_TIME_LIMIT=1350, tasks/download.py), not
-# "just under" it. Hitting this budget is the NORMAL chunk boundary, not a
-# failure: subprocess.run raises TimeoutExpired (which captures partial
-# stdout/stderr + the last emitted cursor), and download_service reclassifies
-# a chunk that made progress as PARTIAL (status "ok"), not an error. The huge
-# headroom means a stuck file or a slow chunk can never let Celery's
-# SoftTimeLimitExceeded preempt TimeoutExpired (the failure mode behind Knuxy
-# #38275 / Anduo #39912/#40411). Earlier we ran one ~1170s run-to-the-wall
-# per arming; that died as a timeout error every time on large catalogs.
+# Sits below download_source's Celery soft_time_limit
+# (DOWNLOAD_SOFT_TIME_LIMIT=1350, tasks/download.py) with ~180s of
+# headroom for phase-3 persist. subprocess.run MUST raise TimeoutExpired
+# before Celery raises SoftTimeLimitExceeded — that exception path
+# captures partial stdout/stderr and finalizes the event; the soft-limit
+# path (until the 2026-06-03 fix) did not. Audit history: 1800 guaranteed
+# SIGKILL against the old hard limit (Knuxy #38275); 1170 was then sized
+# "30s shy of the hard limit (1200)" but still EXCEEDED the soft limit
+# (900), so SoftTimeLimitExceeded preempted TimeoutExpired and every
+# backfill stranded empty (Anduo #39912). Raising the Celery soft/hard
+# limits to 1350/1500 (tasks/download.py) is what made 1170 safe.
+# backfill_runs_remaining=3 still gives ~58 minutes of cumulative walk
+# across three runs for prolific creators.
 BACKFILL_SKIP_VALUE = True
-BACKFILL_CHUNK_SECONDS = 600
+BACKFILL_TIMEOUT_SECONDS = 1170
 
 
 # Sits well below download_source's Celery soft_time_limit
@@ -103,17 +98,6 @@ class SourceConfig:
     Source.backfill_runs_remaining column at the download_service layer
     and is passed to _build_config_for_source as `skip_value`. Same for
     the per-run subprocess timeout.
-
-    `resume_cursor` is RUNTIME state, not persisted operator config: the
-    cursor-paged backfill checkpoint (see parse_last_cursor + the
-    download_service backfill lifecycle). It is consumed only by the native
-    Patreon ingester (the sole cursor-paged platform; gallery-dl's Patreon path
-    was removed at the #697 cutover): on a backfill/recovery run the ingester
-    resumes its newest→oldest walk from that pagination cursor instead of
-    restarting from the top. download_service threads it from
-    config_overrides["_backfill_cursor"]; SourceConfig deliberately does NOT read
-    it in from_dict (config_overrides also holds operator config, and
-    resume_cursor must only apply in backfill/recovery mode).
     """
     content_types: list[str] = field(default_factory=lambda: ["all"])
     sleep: float | None = None
@@ -122,7 +106,6 @@ class SourceConfig:
     filename_pattern: str | None = None
     save_metadata: bool = True
     timeout: int = _DEFAULT_GDL_TIMEOUT_SECONDS
-    resume_cursor: str | None = None
 
     @classmethod
     def from_dict(cls, data: dict) -> SourceConfig:
@@ -155,51 +138,6 @@ class DownloadResult:
     duration_seconds: float = 0.0
     started_at: str | None = None
     completed_at: str | None = None
-    # Plan #704 — structured fields the NATIVE ingester populates directly (None
-    # on the gallery-dl path, which keeps the regex-over-stdout route). When set,
-    # phase 3 reads these instead of scraping the text it would otherwise have to
-    # reconstruct: `run_stats` mirrors _compute_run_stats' shape; `cursor` is the
-    # backfill checkpoint the ingester knows exactly (no parse_last_cursor).
-    run_stats: dict | None = None
-    cursor: str | None = None
-    posts_processed: int = 0
-    # Plan #708 B1 — the server's Retry-After seconds on a RATE_LIMITED result, so
-    # the platform cooldown matches the hint instead of a flat default. None when
-    # unknown (no header, or not a rate-limit failure).
-    retry_after_seconds: float | None = None
-
-
-def extract_errors_warnings(stderr: str) -> str:
-    """Keep only the `[error]`/`[warning]` lines from a captured stderr blob.
-
-    A generic download-log helper (module-level so the native-ingester result
-    path can shape its logs without reaching through a GalleryDLService instance).
-    """
-    if not stderr:
-        return ""
-    kept = [
-        line for line in stderr.splitlines()
-        if "][error]" in line.lower() or "][warning]" in line.lower()
-    ]
-    return "\n".join(kept)
-
-
-def truncate_log(text: str, max_bytes: int = 500_000) -> str:
-    """Cap a captured log to `max_bytes`, eliding the middle (head + tail kept)."""
-    if not text:
-        return text
-    encoded = text.encode("utf-8")
-    if len(encoded) <= max_bytes:
-        return text
-    half = max_bytes // 2
-    head = encoded[:half].decode("utf-8", errors="ignore")
-    tail = encoded[-half:].decode("utf-8", errors="ignore")
-    head_lines = head.count("\n")
-    tail_lines = tail.count("\n")
-    total_lines = text.count("\n")
-    elided = max(0, total_lines - head_lines - tail_lines)
-    marker = f"\n\n... [{elided} lines elided, {len(encoded) - max_bytes} bytes] ...\n\n"
-    return head + marker + tail
 
 
 def _summarize_validation_failures(failures: list[dict]) -> str:
@@ -216,40 +154,6 @@ def _summarize_validation_failures(failures: list[dict]) -> str:
     return f"{n} files quarantined ({top_count}× {top_reason}, mixed)"
 
 
-# (parse_last_cursor was removed in plan #704: the native ingester now carries
-# its checkpoint cursor as a structured DownloadResult.cursor field, so there is
-# no log text to scrape — and gallery-dl platforms never had a cursor.)
-
-
-def make_run_stats(
-    *,
-    exit_code: int = 0,
-    downloaded_count: int = 0,
-    skipped_count: int = 0,
-    per_item_failures: int = 0,
-    warning_count: int = 0,
-    tier_gated_count: int = 0,
-    quarantined_count: int = 0,
-    dead_lettered_count: int = 0,
-) -> dict:
-    """The canonical `run_stats` dict shape, in ONE place.
-
-    Both result producers — gallery-dl's `_compute_run_stats` (log scrape) and the
-    native ingester's per-outcome tally (ingest_core) — build through this so the
-    key set can't drift between backends. Phase 3 + the Logs UI read these keys.
-    """
-    return {
-        "exit_code": exit_code,
-        "downloaded_count": downloaded_count,
-        "skipped_count": skipped_count,
-        "per_item_failures": per_item_failures,
-        "warning_count": warning_count,
-        "tier_gated_count": tier_gated_count,
-        "quarantined_count": quarantined_count,
-        "dead_lettered_count": dead_lettered_count,
-    }
-
-
 class GalleryDLService:
     """Service for executing gallery-dl downloads."""
 
@@ -283,10 +187,16 @@ class GalleryDLService:
         "permission denied", "tier required", "pledge required",
     ]
 
-    # Per-platform defaults for the gallery-dl-backed platforms. Patreon was
-    # removed at the plan-#697 cutover — it now uses the native ingester
-    # (services/patreon_ingester.py), not gallery-dl.
+    # Per-platform defaults. Lifted from GS — same six platforms FC supports.
     PLATFORM_DEFAULTS = {
+        "patreon": {
+            "content_types": ["images", "image_large", "attachments", "postfile", "content"],
+            "directory": ["{date:%Y-%m-%d}_{id}_{title[:40]}"],
+            "filename": "{num:>02}_{filename}.{extension}",
+            "videos": True,
+            "embeds": True,
+            "cursor": True,
+        },
         "subscribestar": {
             "content_types": ["all"],
             "directory": ["{date:%Y-%m-%d}_{id}_{title[:40]}"],
@@ -345,12 +255,7 @@ class GalleryDLService:
                 "skip": True,
                 "sleep": self._rate_limit,
                 "sleep-request": max(0.5, self._rate_limit / 4),
-                # 2 (not 3) retries — a stuck CDN host shouldn't burn a whole
-                # backfill chunk on one file. Anduo #40838 spent ~600s on a
-                # single image (4 extractor HEAD retries × 30s + 4 downloader
-                # GET retries × 120s); halving retries + the downloader
-                # timeout below caps a wedged file at ~1-2 min instead.
-                "retries": 2,
+                "retries": 3,
                 "timeout": 30.0,
                 "verify": True,
                 "postprocessors": [
@@ -365,17 +270,27 @@ class GalleryDLService:
             "downloader": {
                 "part": True,
                 "part-directory": str(self._config_dir / "temp"),
-                # See the extractor retries note above (Anduo #40838): 2
-                # retries + a 60s read-timeout (was 120) so a stalled
-                # connection fails fast. 60s is a per-read timeout, not a
-                # transfer cap — large GIFs that keep streaming are unaffected.
-                "retries": 2,
-                "timeout": 60.0,
-                # NOTE: the Patreon/Mux yt-dlp Referer/Origin forwarding lived
-                # here until the plan-#697 cutover. It was Patreon-specific (and
-                # would have wrongly tagged the other platforms' yt-dlp fetches);
-                # Patreon video is now handled by the native ingester's
-                # downloader (patreon_downloader._VIDEO_HEADERS), so it's gone.
+                "retries": 3,
+                "timeout": 120.0,
+                # Forward Patreon as Referer/Origin to yt-dlp when it
+                # fetches video manifests. Operator-flagged 2026-06-01
+                # (DaferQ patreon): video posts hosted on Mux carry a JWT
+                # playback restriction that checks Referer/Origin on every
+                # request — not just the token signature. gallery-dl's
+                # HEAD probe to stream.mux.com returns 200 (the token is
+                # valid), but yt-dlp's actual GET-with-Range to fetch the
+                # m3u8 manifest 403s because yt-dlp sends its own default
+                # Referer, which Mux's policy rejects. Forcing the right
+                # headers fixes the headers-only case; Mux IP-range
+                # restrictions are unfixable from here.
+                "ytdl": {
+                    "raw-options": {
+                        "http_headers": {
+                            "Referer": "https://www.patreon.com/",
+                            "Origin": "https://www.patreon.com",
+                        },
+                    },
+                },
             },
             "output": {"progress": True},
         }
@@ -427,7 +342,14 @@ class GalleryDLService:
 
         platform_section = config["extractor"].setdefault(platform, {})
 
-        if platform == "hentaifoundry":
+        if platform == "patreon":
+            if "all" in source_config.content_types:
+                platform_section["files"] = [
+                    "images", "image_large", "attachments", "postfile", "content",
+                ]
+            else:
+                platform_section["files"] = source_config.content_types
+        elif platform == "hentaifoundry":
             if "pictures" in source_config.content_types or "all" in source_config.content_types:
                 platform_section["include"] = "all"
 
@@ -587,6 +509,10 @@ class GalleryDLService:
         if not written_paths:
             return quarantined_relpaths, failures
 
+        quarantine_root = (
+            self.images_root / "_quarantine" / artist_slug / platform
+        )
+
         for path in written_paths:
             if not is_validatable(path):
                 continue
@@ -597,14 +523,34 @@ class GalleryDLService:
                 continue
             if result.ok:
                 continue
-            # Shared move+sidecar (file_validator.quarantine_file) — same impl the
-            # native ingester uses, so the layout + provenance sidecar can't drift.
-            dest = quarantine_file(
-                self.images_root, path, artist_slug, platform,
-                url=url, result=result,
-            )
-            if dest is None:
-                continue  # move failed → left in place, not counted
+            try:
+                quarantine_root.mkdir(parents=True, exist_ok=True)
+                dest = quarantine_root / path.name
+                counter = 1
+                while dest.exists():
+                    dest = quarantine_root / f"{path.stem}.{counter}{path.suffix}"
+                    counter += 1
+                path.rename(dest)
+                sidecar = dest.with_suffix(dest.suffix + ".quarantine.json")
+                sidecar.write_text(
+                    json.dumps(
+                        {
+                            "original_path": str(path),
+                            "source_url": url,
+                            "artist_slug": artist_slug,
+                            "platform": platform,
+                            "format": result.format,
+                            "reason": result.reason,
+                            "size": result.size,
+                            "quarantined_at": datetime.now(UTC).isoformat(),
+                        },
+                        indent=2,
+                    )
+                )
+            except OSError as exc:
+                log.error("Failed to quarantine %s: %s. File left in place.", path, exc)
+                continue
+
             log.warning(
                 "Quarantined corrupt file: %s → %s (%s: %s)",
                 path, dest, result.format, result.reason,
@@ -642,24 +588,41 @@ class GalleryDLService:
             if "][warning]" in line.lower() and "not allowed to view post" in line.lower()
         )
 
-        return make_run_stats(
-            exit_code=return_code,
-            downloaded_count=self._count_downloaded_files(stdout),
-            skipped_count=skipped_stdout + skipped_stderr,
-            per_item_failures=per_item_failures,
-            warning_count=warning_count,
-            tier_gated_count=tier_gated_count,
-        )
+        return {
+            "exit_code": return_code,
+            "downloaded_count": self._count_downloaded_files(stdout),
+            "skipped_count": skipped_stdout + skipped_stderr,
+            "per_item_failures": per_item_failures,
+            "warning_count": warning_count,
+            "tier_gated_count": tier_gated_count,
+        }
 
     @staticmethod
     def _extract_errors_warnings(stderr: str) -> str:
-        # Thin delegator to the module-level helper (kept for existing callers).
-        return extract_errors_warnings(stderr)
+        if not stderr:
+            return ""
+        kept = [
+            line for line in stderr.splitlines()
+            if "][error]" in line.lower() or "][warning]" in line.lower()
+        ]
+        return "\n".join(kept)
 
     @staticmethod
     def _truncate_log(text: str, max_bytes: int = 500_000) -> str:
-        # Thin delegator to the module-level helper (kept for existing callers).
-        return truncate_log(text, max_bytes)
+        if not text:
+            return text
+        encoded = text.encode("utf-8")
+        if len(encoded) <= max_bytes:
+            return text
+        half = max_bytes // 2
+        head = encoded[:half].decode("utf-8", errors="ignore")
+        tail = encoded[-half:].decode("utf-8", errors="ignore")
+        head_lines = head.count("\n")
+        tail_lines = tail.count("\n")
+        total_lines = text.count("\n")
+        elided = max(0, total_lines - head_lines - tail_lines)
+        marker = f"\n\n... [{elided} lines elided, {len(encoded) - max_bytes} bytes] ...\n\n"
+        return head + marker + tail
 
     async def download(
         self,

backend/app/services/importer.py

@@ -31,12 +31,7 @@ from ..models import (
     Source,
 )
 from ..utils import safe_probe
-from ..utils.paths import (
-    derive_subdir,
-    derive_top_level_artist,
-    hash_suffixed_name,
-    safe_ext,
-)
+from ..utils.paths import derive_subdir, derive_top_level_artist, hash_suffixed_name
 from ..utils.phash import compute_phash, find_similar
 from ..utils.sidecar import find_sidecar, parse_sidecar
 from ..utils.slug import slugify
@@ -87,9 +82,27 @@ def is_video(path: Path) -> bool:
 
 def _safe_ext(path: Path) -> str:
     """Conservatively extract a file extension for PostAttachment.ext
-    (varchar(32)). Thin wrapper over the shared `utils.paths.safe_ext` (kept for
-    the Path-typed call sites + the [[path_suffix_sanitize]] memory pointer)."""
-    return safe_ext(path)
+    (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:
@@ -430,17 +443,13 @@ class Importer:
             self._capture_attachment(
                 source, post=post, artist=artist_use, resolved=True,
             )
-            reason = f"archive probe rejected, captured unextracted: {probe.reason}"
-            log.warning("%s: %s", source.name, reason)
-            return ImportResult(status="attached", error=reason)
+            return ImportResult(status="attached")
 
         artist_use = artist if artist is not None else self._resolve_artist(source)
         post = self._post_for_sidecar(source, artist_use)
         member_ids: list[int] = []
-        member_total = 0
         with extract_archive(source) as members:
             for _name, member_path in members:
-                member_total += 1
                 if not is_supported(member_path):
                     continue  # non-media preserved via the stored archive
                 res = self._import_media(
@@ -457,19 +466,7 @@ class Importer:
                 status="imported", image_id=member_ids[0],
                 member_image_ids=member_ids,
             )
-        # No images landed — surface WHY so a post showing "no images" beside an
-        # archive is diagnosable instead of silent. Zero members usually means
-        # the extractor backend is missing/failed (unar for rar, py7zr for 7z)
-        # or the file is corrupt; non-zero-but-no-images means it held only
-        # non-media files.
-        reason = (
-            "archive extracted but held no supported image/video members"
-            if member_total
-            else "archive yielded no members (unsupported/corrupt, or the "
-            "extractor backend failed)"
-        )
-        log.warning("%s: %s", source.name, reason)
-        return ImportResult(status="attached", error=reason)
+        return ImportResult(status="attached")
 
     def _import_media(
         self, source: Path, attribution_path: Path,

backend/app/services/ingest_core.py

@@ -1,647 +0,0 @@
-"""Platform-agnostic native-ingest core (plan #706, build on #697/#703/#704/#705).
-
-The orchestration that drives a native subscription walk — page a feed →
-extract media → tiered skip (seen-ledger / on-disk / dead-letter) → download →
-mark-seen / record-failures / checkpoint-cursor → return a gallery-dl-shaped
-`DownloadResult`, across tick/backfill/recovery modes — is identical for every
-platform. Only four things are platform-specific, and they're INJECTED at
-construction by a thin adapter (e.g. `PatreonIngester`):
-
-  - `client`    — `.iter_posts(feed_id, cursor)` yielding `(post, included,
-                  page_cursor)` + `.extract_media(post, included) -> [media]`.
-  - `downloader`— `.download_post(post, media, artist_slug, is_seen,
-                  should_stop) -> [MediaOutcome]` (status in downloaded/
-                  skipped_seen/skipped_disk/quarantined/error;
-                  `.path`/`.error`/`.post_id`). `should_stop()` is polled
-                  between media so the time-box is honoured mid-post.
-  - ledger      — `seen_model` + `failed_model` SQLAlchemy models (+ their
-                  on-conflict UNIQUE constraint names) and a `ledger_key(media)`.
-  - failure map — the adapter overrides `_failure_result` (platform exception
-                  → DownloadResult.error_type) and supplies `error_base` (the
-                  exception type the walk catches) + `platform` (result label).
-
-Everything DB touches a SHORT-LIVED sync session from the injected sessionmaker —
-never held across a network fetch ([[db-connection-held-across-subprocess]]).
-Plain-HTTP homelab: no secure-context Web API.
-"""
-
-from __future__ import annotations
-
-import json
-import logging
-import time
-from collections.abc import Callable
-
-from sqlalchemy import delete, func, select, text
-from sqlalchemy.dialects.postgresql import insert as pg_insert
-
-from .gallery_dl import DownloadResult, ErrorType, make_run_stats
-
-log = logging.getLogger(__name__)
-
-# Stop a tick after this many CONTIGUOUS already-have-it media (seen-ledger or
-# on-disk) — the cheap native equivalent of gallery-dl's `exit:20`, now free of
-# per-file HEADs. Headroom against paywalled/undownloadable items interleaving.
-_TICK_SEEN_THRESHOLD = 20
-
-# plan #705 #7: after this many failed download/validate attempts a media is
-# "dead-lettered" and skipped on routine tick/backfill walks (recovery still
-# re-attempts it). Stops a permanently-broken media re-erroring forever.
-DEAD_LETTER_THRESHOLD = 3
-# last_error is Text but bound it so a giant traceback doesn't bloat the row.
-_ERROR_MAX = 1000
-
-# plan #709: throttle the live-progress write to the running DownloadEvent to one
-# every ~5s — a steady cadence for the Downloads view regardless of how big/slow a
-# page is (page boundaries can be minutes apart on image-dense backfills, so a
-# page-tied update would lurch). Trivial churn (~one single-row UPDATE / 5s).
-_LIVE_PROGRESS_INTERVAL = 5.0
-
-
-class Ingester:
-    """Generic native-ingest orchestration. Subclass with a platform adapter
-    (see the module docstring) — or construct directly with the keyword seams."""
-
-    def __init__(
-        self,
-        *,
-        client,
-        downloader,
-        session_factory: Callable[[], object],
-        seen_model,
-        failed_model,
-        seen_constraint: str,
-        failed_constraint: str,
-        ledger_key: Callable[[object], str],
-        platform: str,
-        error_base: type[Exception],
-    ):
-        self.client = client
-        self.downloader = downloader
-        self.session_factory = session_factory
-        self._seen_model = seen_model
-        self._failed_model = failed_model
-        self._seen_constraint = seen_constraint
-        self._failed_constraint = failed_constraint
-        self._ledger_key = ledger_key
-        self._platform = platform
-        self._error_base = error_base
-
-    # -- public ------------------------------------------------------------
-
-    def run(
-        self,
-        *,
-        source_id: int,
-        campaign_id: str,
-        artist_slug: str,
-        url: str,
-        mode: str,
-        resume_cursor: str | None = None,
-        time_budget_seconds: float = 870.0,
-        seen_threshold: int = _TICK_SEEN_THRESHOLD,
-        posts_base: int = 0,
-        event_id: int | None = None,
-    ) -> DownloadResult:
-        """Walk + download for one source, returning a gallery-dl-shaped result.
-
-        `mode` is "tick" | "backfill" | "recovery". Recovery bypasses the tier-1
-        seen-ledger AND the dead-letter ledger (tier-2 disk still skips kept
-        files). The walk stops on:
-          - budget exhaustion (time_budget_seconds)         → TIMEOUT / PARTIAL
-          - tick early-out (seen_threshold contiguous seen)  → success
-          - reaching the bottom of the feed                  → success (rc 0)
-        A client-level failure (drift / auth / network) fails the whole run loud.
-        """
-        bypass_seen = mode == "recovery"
-        # Only deep walks checkpoint their cursor mid-flight (plan #705 #6); a
-        # tick has no resumable backfill state.
-        checkpoint = mode in ("backfill", "recovery")
-        ledger_key = self._ledger_key
-        start = time.monotonic()
-        last_live = start  # plan #709: last live-progress write timestamp
-        log_lines: list[str] = []
-        written: list[str] = []
-        quarantined_paths: list[str] = []
-        downloaded = 0
-        errors = 0
-        quarantined = 0
-        dead_lettered = 0
-        skipped_count = 0
-        posts_processed = 0
-        # Net-new posts THIS chunk for the live progress badge (plan #704 #5);
-        # excludes the re-walked resume page so _backfill_posts stays a monotonic
-        # absolute across chunks instead of an inflating sum. posts_processed
-        # stays the gross per-chunk count used for the run summary.
-        chunk_new_posts = 0
-        consecutive_seen = 0
-        emitted_cursor: str | None = None
-        reached_bottom = False
-        budget_hit = False
-        early_out = False
-        stopped = False  # plan #708 B4: operator hit Stop mid-walk
-        cancel_armed = False  # latched once we observe a live "running" state
-
-        def _result(
-            *, success: bool, return_code: int,
-            error_type: ErrorType | None, error_message: str | None,
-        ) -> DownloadResult:
-            # plan #704: return STRUCTURED data — phase 3 reads run_stats/cursor
-            # directly instead of regex-scraping a reconstructed stdout. stdout
-            # stays a human-readable summary (no fake `Cursor:` lines).
-            return DownloadResult(
-                success=success,
-                url=url,
-                artist_slug=artist_slug,
-                platform=self._platform,
-                files_downloaded=downloaded,
-                files_quarantined=quarantined,
-                quarantined_paths=list(quarantined_paths),
-                written_paths=written,
-                stdout="\n".join(log_lines),
-                stderr="",
-                return_code=return_code,
-                error_type=error_type,
-                error_message=error_message,
-                duration_seconds=time.monotonic() - start,
-                cursor=emitted_cursor,
-                posts_processed=posts_processed,
-                run_stats=make_run_stats(
-                    exit_code=return_code,
-                    downloaded_count=downloaded,
-                    skipped_count=skipped_count,
-                    per_item_failures=errors,
-                    quarantined_count=quarantined,
-                    dead_lettered_count=dead_lettered,
-                ),
-            )
-
-        try:
-            for post, included, page_cursor in self.client.iter_posts(
-                campaign_id, cursor=resume_cursor
-            ):
-                # Checkpoint the cursor that FETCHED this page the moment we
-                # START it — so a chunk cut mid-page resumes the page, not the one
-                # after it. Carried as DownloadResult.cursor (plan #704).
-                if page_cursor and page_cursor != emitted_cursor:
-                    emitted_cursor = page_cursor
-                    # plan #705 #6: persist the cursor at each page boundary so a
-                    # worker SIGKILL mid-chunk resumes near the crash, not the
-                    # chunk start. (phase 3 still writes the final cursor — same
-                    # value; this is the crash-safety net.) plan #704 #5: persist
-                    # the live posts count alongside it so the badge climbs DURING
-                    # the chunk, not only when it ends.
-                    if checkpoint:
-                        # plan #708 B4: an operator Stop pops `_backfill_state` —
-                        # bail at the page boundary (progress already checkpointed)
-                        # before more network work, so the live chunk halts
-                        # promptly instead of running to its time-box. LATCH on the
-                        # first observed "running" state, so a run invoked WITHOUT a
-                        # running state (a unit test, or a stale call) never
-                        # spuriously self-cancels. A short SELECT, never held.
-                        if self._still_running(source_id):
-                            cancel_armed = True
-                        elif cancel_armed:
-                            stopped = True
-                            break
-                        self._checkpoint_cursor(source_id, emitted_cursor)
-                        self._checkpoint_posts(source_id, posts_base + chunk_new_posts)
-
-                # Time-box check at the post boundary (coarse, like a gallery-dl
-                # chunk). Backfill/recovery resume from emitted_cursor next chunk.
-                if time.monotonic() - start >= time_budget_seconds:
-                    budget_hit = True
-                    break
-
-                posts_processed += 1
-                # The resume page (its cursor == resume_cursor) was already
-                # counted by the chunk that checkpointed it — don't re-count it
-                # into the persisted badge (plan #704 #5). First chunk has
-                # resume_cursor None, so everything counts.
-                if not (resume_cursor and page_cursor == resume_cursor):
-                    chunk_new_posts += 1
-                media = self.client.extract_media(post, included)
-                if not media:
-                    continue
-
-                keys = [ledger_key(m) for m in media]
-                # Recovery bypasses BOTH the seen-ledger AND the dead-letter
-                # ledger (the operator's "try everything again"); routine walks
-                # skip seen + dead media (tier-1 + tier-1.5, plan #705 #7).
-                dead = set() if bypass_seen else self._dead_keys(source_id, keys)
-                seen = (
-                    set()
-                    if bypass_seen
-                    else self._seen_keys(source_id, keys)
-                )
-                skip = seen | dead
-
-                def _is_skip(m, _skip=skip) -> bool:
-                    return ledger_key(m) in _skip
-
-                # Honour the time-box DURING a media-dense post too, not only at
-                # the per-post boundary below — else one heavy post can blow the
-                # chunk budget out to the Celery soft limit (Pocketacer, 2026-06-07).
-                outcomes = self.downloader.download_post(
-                    post, media, artist_slug, is_seen=_is_skip,
-                    should_stop=lambda: time.monotonic() - start >= time_budget_seconds,
-                )
-
-                to_mark: list[tuple[str, str]] = []
-                to_clear: list[str] = []  # recovered → drop any dead-letter row
-                to_fail: list[tuple[str, str, str]] = []  # (key, post_id, error)
-                for media_item, outcome in zip(media, outcomes, strict=False):
-                    key = ledger_key(media_item)
-                    if key in dead:
-                        dead_lettered += 1  # skipped because previously dead
-                    if outcome.status == "downloaded":
-                        downloaded += 1
-                        if outcome.path is not None:
-                            written.append(str(outcome.path))
-                        to_mark.append((key, media_item.post_id))
-                        to_clear.append(key)
-                        consecutive_seen = 0
-                    elif outcome.status == "skipped_disk":
-                        # Already on disk (a prior run). Reconcile the ledger so a
-                        # later tick skips it at tier-1 without a disk stat, but
-                        # do NOT re-feed it to phase 3 — attach_in_place would see
-                        # the duplicate sha256 and unlink the on-disk copy.
-                        to_mark.append((key, media_item.post_id))
-                        to_clear.append(key)
-                        skipped_count += 1
-                        consecutive_seen += 1
-                    elif outcome.status == "skipped_seen":
-                        skipped_count += 1
-                        consecutive_seen += 1
-                    elif outcome.status == "quarantined":
-                        # New content that failed validation (corrupt) — counted
-                        # distinctly so the run surfaces a real quarantined total.
-                        # Not marked seen (a later walk may re-fetch a fixed file);
-                        # it IS new content, so it breaks the run-of-seen. Counts
-                        # toward the dead-letter ledger (plan #705 #7).
-                        quarantined += 1
-                        if outcome.path is not None:
-                            quarantined_paths.append(str(outcome.path))
-                        to_fail.append((key, media_item.post_id, outcome.error or "quarantined"))
-                        consecutive_seen = 0
-                    elif outcome.status == "error":
-                        errors += 1
-                        to_fail.append((key, media_item.post_id, outcome.error or "error"))
-                        # An error neither advances nor resets the run-of-seen.
-
-                    if mode == "tick" and consecutive_seen >= seen_threshold:
-                        early_out = True
-                        break
-
-                # Persist ledger changes AFTER the network fetch, on short
-                # sessions: mark downloaded/on-disk seen, clear any dead-letter
-                # for recovered media, and record failures (plan #705 #7).
-                if to_mark:
-                    self._mark_seen(source_id, to_mark)
-                if to_clear:
-                    self._clear_failures(source_id, to_clear)
-                if to_fail:
-                    self._record_failures(source_id, to_fail)
-
-                # plan #709: time-throttled live progress to the running event so
-                # the Downloads view ticks ~every 5s, independent of page size.
-                now = time.monotonic()
-                if event_id is not None and (now - last_live) >= _LIVE_PROGRESS_INTERVAL:
-                    last_live = now
-                    self._write_live_progress(event_id, {
-                        "downloaded": downloaded,
-                        "skipped": skipped_count,
-                        "errors": errors,
-                        "quarantined": quarantined,
-                        "posts": posts_processed,
-                    })
-
-                if early_out:
-                    break
-            else:
-                reached_bottom = True
-        except self._error_base as exc:
-            # The platform's client-error base — _failure_result (adapter)
-            # maps it to a typed error.
-            return self._failure_result(exc, _result)
-
-        # plan #708 B4: a Stop already popped the backfill state (incl. cursor +
-        # posts), so don't re-write them — return PARTIAL (reads as "ok/progress",
-        # the lifecycle no-ops since state is gone) instead of a false "complete".
-        if stopped:
-            return _result(
-                success=False, return_code=-1,
-                error_type=ErrorType.PARTIAL,
-                error_message=f"Stopped by operator: {downloaded} file(s) this chunk",
-            )
-
-        # Final authoritative posts count for the badge — captures the last page
-        # after the last boundary write and the time-box break (plan #704 #5).
-        if checkpoint:
-            self._checkpoint_posts(source_id, posts_base + chunk_new_posts)
-
-        if errors:
-            log_lines.append(f"{errors} media item(s) failed")
-        if quarantined:
-            log_lines.append(f"{quarantined} media item(s) quarantined (invalid)")
-        if dead_lettered:
-            log_lines.append(f"{dead_lettered} media item(s) skipped (dead-lettered)")
-        log_lines.append(
-            f"{self._platform} ingest ({mode}): {downloaded} downloaded, "
-            f"{skipped_count} skipped, {quarantined} quarantined, "
-            f"{dead_lettered} dead-lettered, {errors} error(s), "
-            f"{posts_processed} post(s)"
-            + (", reached end" if reached_bottom else "")
-            + (", time-boxed" if budget_hit else "")
-        )
-
-        if budget_hit:
-            # A chunk that hit its time-box but made forward progress is a
-            # NORMAL chunk boundary, not a failure (PARTIAL → status "ok"); the
-            # next chunk resumes from the emitted cursor. No progress → TIMEOUT,
-            # which feeds download_service's backfill stall-guard. rc<0 mirrors
-            # subprocess TimeoutExpired so completion detection stays false.
-            made_progress = downloaded > 0 or emitted_cursor != resume_cursor
-            if made_progress:
-                return _result(
-                    success=False, return_code=-1,
-                    error_type=ErrorType.PARTIAL,
-                    error_message=(
-                        f"Backfill chunk: {downloaded} file(s) — continuing"
-                    ),
-                )
-            return _result(
-                success=False, return_code=-1,
-                error_type=ErrorType.TIMEOUT,
-                error_message="Chunk timed out with no progress",
-            )
-
-        # Normal success: reached the bottom, or a tick that early-outed. rc 0 +
-        # error_type None is REQUIRED for a backfill/recovery walk that reached
-        # the bottom to be marked COMPLETE by
-        # download_service._apply_backfill_lifecycle — so we return None even
-        # when downloaded == 0 (a re-confirming walk that found nothing new still
-        # completed). success=True maps to status "ok" regardless. A tick that
-        # early-outed also returns here; ticks never set backfill state so the
-        # lifecycle is a no-op for them.
-        return _result(
-            success=True, return_code=0,
-            error_type=None, error_message=None,
-        )
-
-    # -- preview (dry-run) -------------------------------------------------
-
-    def preview(
-        self,
-        source_id: int,
-        campaign_id: str,
-        *,
-        page_limit: int = 3,
-        sample_size: int = 10,
-    ) -> dict:
-        """Dry-run (plan #708 B4): walk up to `page_limit` pages and count media
-        NOT already in the seen/dead ledgers, WITHOUT downloading anything.
-
-        Read-only — only the seen/dead SELECTs touch the DB (short sessions). Lets
-        an operator gauge "is this source worth a backfill?" cheaply. Returns:
-            {total_new, posts_scanned, pages_scanned, has_more,
-             sample: [{title, date, new}, ...]}  # sample = posts with new media
-        A client-level failure (auth/drift) propagates to the caller.
-        """
-        total_new = 0
-        posts_scanned = 0
-        pages_scanned = 0
-        has_more = False
-        sample: list[dict] = []
-        unset = object()
-        last_page: object = unset
-        for post, included, page_cursor in self.client.iter_posts(
-            campaign_id, cursor=None
-        ):
-            if page_cursor != last_page:
-                last_page = page_cursor
-                pages_scanned += 1
-                if pages_scanned > page_limit:
-                    has_more = True
-                    pages_scanned = page_limit
-                    break
-            posts_scanned += 1
-            media = self.client.extract_media(post, included)
-            if not media:
-                continue
-            keys = [self._ledger_key(m) for m in media]
-            skip = self._seen_keys(source_id, keys) | self._dead_keys(source_id, keys)
-            new_count = sum(1 for m in media if self._ledger_key(m) not in skip)
-            total_new += new_count
-            if new_count > 0 and len(sample) < sample_size:
-                meta = self.client.post_meta(post)
-                sample.append(
-                    {
-                        "title": meta.get("title") or "(untitled)",
-                        "date": meta.get("date"),
-                        "new": new_count,
-                    }
-                )
-        return {
-            "total_new": total_new,
-            "posts_scanned": posts_scanned,
-            "pages_scanned": pages_scanned,
-            "has_more": has_more,
-            "sample": sample,
-        }
-
-    # -- failure mapping (adapter overrides) -------------------------------
-
-    def _failure_result(self, exc: Exception, _result) -> DownloadResult:
-        """Map a platform client-error to a typed failed DownloadResult. The base
-        gives a safe default; adapters override with their exception taxonomy."""
-        log.warning("%s ingest failed: %s", self._platform, exc)
-        return _result(
-            success=False, return_code=1,
-            error_type=ErrorType.UNKNOWN_ERROR, error_message=str(exc),
-        )
-
-    # -- seen-ledger (short-lived sessions) --------------------------------
-
-    def _seen_keys(self, source_id: int, keys: list[str]) -> set[str]:
-        """Which of `keys` are already in the seen-ledger for this source.
-
-        One short SELECT on its own session — opened and closed without any
-        network in between (the GETs happen after, in download_post).
-        """
-        if not keys:
-            return set()
-        with self.session_factory() as session:
-            rows = session.execute(
-                select(self._seen_model.filehash).where(
-                    self._seen_model.source_id == source_id,
-                    self._seen_model.filehash.in_(keys),
-                )
-            ).scalars().all()
-        return set(rows)
-
-    def _checkpoint_cursor(self, source_id: int, cursor: str) -> None:
-        """Persist the in-progress backfill cursor mid-walk (plan #705 #6).
-
-        ATOMIC, single-key UPDATE: cast the JSON column to jsonb, set just
-        `_backfill_cursor`, cast back — so it never clobbers operator config or
-        the other backfill keys (no read-modify-write race). The in-flight guard
-        means only this source's one download runs at a time; a concurrent
-        operator stop is benign (a stray cursor with no `_backfill_state` is
-        ignored by tick mode and cleared on the next start).
-        """
-        with self.session_factory() as session:
-            session.execute(
-                text(
-                    "UPDATE source SET config_overrides = jsonb_set("
-                    "  coalesce(config_overrides::jsonb, '{}'::jsonb),"
-                    "  '{_backfill_cursor}', to_jsonb(cast(:cur AS text))"
-                    ")::json WHERE id = :sid"
-                ),
-                {"cur": cursor, "sid": source_id},
-            )
-            session.commit()
-
-    def _write_live_progress(self, event_id: int, counts: dict) -> None:
-        """Throttled mid-walk write of live counts to the RUNNING download_event
-        (plan #709) so the Downloads view shows progress before the chunk
-        finishes. A short session (never held across the walk); the `status =
-        'running'` guard avoids clobbering an event phase 3 already finalized.
-        `metadata` is JSONB — jsonb_set sets just the `live` key, leaving the rest
-        for phase 3 to overwrite with the final run_stats."""
-        with self.session_factory() as session:
-            session.execute(
-                text(
-                    "UPDATE download_event SET metadata = jsonb_set("
-                    "  coalesce(metadata, '{}'::jsonb), '{live}',"
-                    "  cast(:live AS jsonb)) "
-                    "WHERE id = :eid AND status = 'running'"
-                ),
-                {"live": json.dumps(counts), "eid": event_id},
-            )
-            session.commit()
-
-    def _still_running(self, source_id: int) -> bool:
-        """True while the source is armed for a deep walk (plan #708 B4).
-
-        An operator Stop (`source_service.stop_backfill`) pops `_backfill_state`,
-        so a False here means "cancel this chunk now". One short SELECT on its own
-        session — never held across the walk
-        ([[db-connection-held-across-subprocess]])."""
-        with self.session_factory() as session:
-            state = session.execute(
-                text(
-                    "SELECT config_overrides::jsonb ->> '_backfill_state' "
-                    "FROM source WHERE id = :sid"
-                ),
-                {"sid": source_id},
-            ).scalar_one_or_none()
-        return state == "running"
-
-    def _checkpoint_posts(self, source_id: int, posts: int) -> None:
-        """Persist the live backfill posts-processed count mid-walk (plan #704 #5).
-
-        Same atomic single-key jsonb_set dance as _checkpoint_cursor, on the
-        `_backfill_posts` key (cast to a JSON number) — so the progress badge
-        climbs DURING a chunk without clobbering operator config or the cursor.
-        The ingester OWNS this key now; download_service no longer accumulates it
-        post-chunk (which lagged a whole chunk and over-counted the resume page).
-        """
-        with self.session_factory() as session:
-            session.execute(
-                text(
-                    "UPDATE source SET config_overrides = jsonb_set("
-                    "  coalesce(config_overrides::jsonb, '{}'::jsonb),"
-                    "  '{_backfill_posts}', to_jsonb(cast(:posts AS int))"
-                    ")::json WHERE id = :sid"
-                ),
-                {"posts": posts, "sid": source_id},
-            )
-            session.commit()
-
-    def _mark_seen(self, source_id: int, items: list[tuple[str, str]]) -> None:
-        """Idempotent upsert of (filehash, post_id) seen-ledger rows for a page.
-
-        ON CONFLICT DO NOTHING against the (source_id, filehash) UNIQUE so a
-        re-sighting — or a concurrent walk — is a harmless no-op
-        ([[scalar_one_or_none-duplicates]]: never check-then-insert without the
-        DB constraint backing it). De-dup the batch locally first so a single
-        page can't present the same key twice to one INSERT.
-        """
-        seen_local: set[str] = set()
-        values = []
-        for key, post_id in items:
-            if key in seen_local:
-                continue
-            seen_local.add(key)
-            values.append(
-                {"source_id": source_id, "filehash": key, "post_id": post_id}
-            )
-        if not values:
-            return
-        with self.session_factory() as session:
-            stmt = pg_insert(self._seen_model).values(values)
-            stmt = stmt.on_conflict_do_nothing(constraint=self._seen_constraint)
-            session.execute(stmt)
-            session.commit()
-
-    # -- dead-letter ledger (plan #705 #7) ---------------------------------
-
-    def _dead_keys(self, source_id: int, keys: list[str]) -> set[str]:
-        """Which of `keys` have failed >= DEAD_LETTER_THRESHOLD times (dead).
-        One short SELECT; recovery never calls this (it re-attempts dead media)."""
-        if not keys:
-            return set()
-        with self.session_factory() as session:
-            rows = session.execute(
-                select(self._failed_model.filehash).where(
-                    self._failed_model.source_id == source_id,
-                    self._failed_model.filehash.in_(keys),
-                    self._failed_model.attempts >= DEAD_LETTER_THRESHOLD,
-                )
-            ).scalars().all()
-        return set(rows)
-
-    def _record_failures(
-        self, source_id: int, items: list[tuple[str, str, str]]
-    ) -> None:
-        """Upsert-increment the dead-letter ledger for failed media. On conflict
-        bump `attempts` and refresh last_error/last_failed_at (UNIQUE backs the
-        upsert — no check-then-insert). De-dup the batch (one row/key, last error
-        wins)."""
-        by_key: dict[str, str] = {}
-        for key, _post_id, err in items:
-            by_key[key] = (err or "")[:_ERROR_MAX]
-        if not by_key:
-            return
-        values = [
-            {"source_id": source_id, "filehash": k, "attempts": 1, "last_error": e}
-            for k, e in by_key.items()
-        ]
-        with self.session_factory() as session:
-            stmt = pg_insert(self._failed_model).values(values)
-            stmt = stmt.on_conflict_do_update(
-                constraint=self._failed_constraint,
-                set_={
-                    "attempts": self._failed_model.attempts + 1,
-                    "last_error": stmt.excluded.last_error,
-                    "last_failed_at": func.now(),
-                },
-            )
-            session.execute(stmt)
-            session.commit()
-
-    def _clear_failures(self, source_id: int, keys: list[str]) -> None:
-        """Drop dead-letter rows for media that just downloaded cleanly — they
-        recovered. A no-op DELETE for keys that were never failing."""
-        unique = list(dict.fromkeys(keys))
-        if not unique:
-            return
-        with self.session_factory() as session:
-            session.execute(
-                delete(self._failed_model).where(
-                    self._failed_model.source_id == source_id,
-                    self._failed_model.filehash.in_(unique),
-                )
-            )
-            session.commit()

backend/app/services/page_number_parser.py

@@ -1,41 +0,0 @@
-"""Parse a stated page number/range out of a post's title/description (FC-6.2).
-
-Artists often state where an installment sits in a series — "pages 9-12",
-"Page 5", "[3/8]". We use that to order chapters and flag missing-page gaps.
-This is best-effort: a confident match wins, otherwise we return None and the
-caller falls back to capture/post-date order. Keep it conservative — a wrong
-page number is worse than no page number — so matches require an explicit
-page keyword (page/pg/pp) or a bracketed N/M fraction, never a bare number.
-
-Supported forms (case-insensitive):
-  range    "pages 9-12", "pg 9–12", "pp. 9 - 12"        -> (9, 12)
-  fraction "page 3 of 8", "pg 3/8", "[3/8]", "(3/8)"     -> (3, 3)
-  single   "page 5", "pg 5", "pp 5"                       -> (5, 5)
-"""
-
-import re
-
-# Page keyword: page/pages/pg/pgs/pp/pp.  (NOT a bare "p" — too many false hits.)
-_KW = r"(?:pages?|pgs?|pp\.?)"
-_DASH = r"[-–—]"
-
-_RANGE = re.compile(rf"\b{_KW}\s*(\d{{1,4}})\s*{_DASH}\s*(\d{{1,4}})", re.I)
-_OF = re.compile(rf"\b{_KW}\s*(\d{{1,4}})\s*(?:of|/)\s*\d{{1,4}}\b", re.I)
-_BRACKET = re.compile(r"[\[(]\s*(\d{1,4})\s*/\s*\d{1,4}\s*[\])]")
-_SINGLE = re.compile(rf"\b{_KW}\s*(\d{{1,4}})\b", re.I)
-
-
-def parse_page_range(text: str | None) -> tuple[int, int] | None:
-    """Return (start, end) or None. start <= end; a single page yields (n, n)."""
-    if not text:
-        return None
-    m = _RANGE.search(text)
-    if m:
-        a, b = int(m.group(1)), int(m.group(2))
-        return (a, b) if a <= b else (b, a)
-    for rx in (_OF, _BRACKET, _SINGLE):
-        m = rx.search(text)
-        if m:
-            n = int(m.group(1))
-            return (n, n)
-    return None

backend/app/services/patreon_client.py

@@ -1,604 +0,0 @@
-"""Native Patreon JSON:API client (build step 1 of the native ingester).
-
-Clean-room reimplementation of the Patreon `/api/posts` read path. This is a
-plain, synchronous client over `requests` — the orchestrator already wraps
-sync importer calls, so nothing here needs to be async (mirrors
-patreon_resolver.py, which wraps its sync lookup in run_in_executor at the
-call site).
-
-Scope (build step 1): fetch + page + parse only. This module is NOT wired into
-download_service yet — that is a later step. The public surface here exists so
-the later step can drive it:
-  - PatreonClient(cookies_path).iter_posts(campaign_id)
-        → (post, included_index, page_cursor)
-  - extract_media(post, included_index) → list[MediaItem]
-  - parse_cursor_from_url(url) → cursor
-
-Drift detection is loud on purpose: Patreon ships JSON:API and the shapes we
-depend on (top-level `data`, media resources carrying `file_name`/`url`) are
-the contract. If a response comes back as an HTML login page or a media
-resource is missing the fields we resolve against, we raise PatreonDriftError
-rather than silently yielding empty media — so the later import step surfaces
-"Patreon changed something" instead of "creator has no posts".
-
-FC runs on a plain-HTTP homelab; nothing here uses a secure-context Web API.
-"""
-
-from __future__ import annotations
-
-import http.cookiejar
-import logging
-import os
-import re
-import time
-from collections.abc import Iterator
-from dataclasses import dataclass
-from html import unescape
-from pathlib import Path
-from urllib.parse import parse_qs, urlsplit
-
-import requests
-
-from ..utils.paths import safe_ext
-
-log = logging.getLogger(__name__)
-
-_POSTS_URL = "https://www.patreon.com/api/posts"
-_USER_AGENT = (
-    "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 "
-    "(KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36"
-)
-_TIMEOUT_SECONDS = 30.0
-
-# 429 backoff (plan #703): ride out a transient API rate-limit instead of
-# failing the whole walk (which would stamp RATE_LIMITED → platform-wide
-# cooldown → every Patreon source dark). Honor the server's `Retry-After`;
-# otherwise exponential base·2^(n-1), capped. Only after the retries are
-# exhausted does the 429 propagate as terminal RATE_LIMITED.
-_MAX_429_RETRIES = 3
-_BACKOFF_BASE_SECONDS = 2.0
-_BACKOFF_CAP_SECONDS = 30.0
-
-
-def _retry_after_seconds(
-    resp: requests.Response,
-    attempt: int,
-    *,
-    base: float = _BACKOFF_BASE_SECONDS,
-    cap: float = _BACKOFF_CAP_SECONDS,
-) -> float:
-    """Backoff delay for a 429: the `Retry-After` seconds header if present and
-    numeric, else exponential `base·2^(attempt-1)`, both capped. (HTTP-date form
-    of Retry-After is rare here and falls through to exponential.)"""
-    header = resp.headers.get("Retry-After")
-    if header:
-        try:
-            return min(float(header), cap)
-        except (TypeError, ValueError):
-            pass
-    return min(base * (2 ** max(0, attempt - 1)), cap)
-
-# JSON:API request contract (observed from real traffic — see module plan).
-_INCLUDE = (
-    "campaign,access_rules,attachments,attachments_media,audio,images,media,"
-    "native_video_insights,user,user_defined_tags,ti_checks"
-)
-_FIELDS_POST = (
-    "content,post_file,image,post_type,published_at,title,url,patreon_url,"
-    "current_user_can_view"
-)
-_FIELDS_MEDIA = "id,image_urls,download_url,metadata,file_name"
-_FIELDS_CAMPAIGN = "name,url"
-
-# A CDN download URL embeds a 32-char hex (MD5) path segment; that segment is
-# Patreon's stable per-file identity and is what we dedup + ledger against.
-# Same role gallery-dl's _filehash plays. Match the FIRST 32-hex run anywhere
-# in the URL (path or query); real Patreon CDN URLs carry exactly one.
-_FILEHASH_RE = re.compile(r"([0-9a-fA-F]{32})")
-
-# Inline post `content` is HTML; images are emitted as <img ... src="...">.
-# Pull every src; downstream dedup collapses any that duplicate a gallery item
-# by filehash. Tolerant of attribute ordering and single/double quotes.
-_CONTENT_IMG_RE = re.compile(r"<img\b[^>]*?\bsrc=[\"']([^\"']+)[\"']", re.IGNORECASE)
-
-
-class PatreonAPIError(Exception):
-    """Base for native Patreon client failures.
-
-    `status_code` carries the HTTP status when the failure was an HTTP response
-    (None for transport-level / parse failures), so the ingester can map it to a
-    DownloadResult.error_type (429 → rate_limited, 404 → not_found, …).
-    `retry_after` carries the server's `Retry-After` seconds on a terminal 429, so
-    the platform cooldown can match the server's hint instead of a flat default
-    (plan #708 B1).
-    """
-
-    def __init__(
-        self,
-        message: str,
-        *,
-        status_code: int | None = None,
-        retry_after: float | None = None,
-    ):
-        super().__init__(message)
-        self.status_code = status_code
-        self.retry_after = retry_after
-
-
-class PatreonAuthError(PatreonAPIError):
-    """Authentication / authorization failure — missing or expired session
-    cookies, an insufficient pledge tier, or an HTML login/challenge page served
-    where JSON was expected. DISTINCT from drift: the fix is rotating the
-    credential, not updating the ingester. Maps to error_type 'auth_error'.
-    """
-
-
-class PatreonDriftError(PatreonAPIError):
-    """A JSON response did not match the JSON:API shape we depend on.
-
-    Raised for: a missing top-level `data` list, `data` not a list, or a media
-    resource lacking the `file_name`/`url` fields we resolve against. Fail loud
-    so the import step flags API drift (ingester needs update) instead of
-    silently importing nothing. An HTML-login / non-JSON body is auth, not
-    drift — that raises PatreonAuthError.
-    """
-
-
-@dataclass
-class MediaItem:
-    """One resolved downloadable item belonging to a post.
-
-    Fields:
-      url       — the CDN/download URL to fetch.
-      filename  — media `file_name` when present; otherwise the URL basename
-                  (NEVER a network call). Bounded/sane extension.
-      kind      — one of: "images", "image_large", "attachments", "postfile",
-                  "content". Mirrors gallery-dl's `files` content-type names so
-                  the later step can honor the same per-source content_types.
-      filehash  — the 32-char hex (MD5) segment from the CDN URL, or None if
-                  the URL carries no such segment. Used for in-post dedup and
-                  the cross-run seen-ledger.
-      post_id   — the owning post's id (so a flattened media list stays
-                  traceable to its post).
-    """
-
-    url: str
-    filename: str
-    kind: str
-    filehash: str | None
-    post_id: str
-
-
-def _load_session(cookies_path: str | Path | None) -> requests.Session:
-    session = requests.Session()
-    session.headers.update(
-        {
-            "User-Agent": _USER_AGENT,
-            "Accept": "application/vnd.api+json",
-        }
-    )
-    if cookies_path and os.path.isfile(str(cookies_path)):
-        try:
-            jar = http.cookiejar.MozillaCookieJar(str(cookies_path))
-            jar.load(ignore_discard=True, ignore_expires=True)
-            session.cookies = jar  # type: ignore[assignment]
-        except (OSError, http.cookiejar.LoadError) as exc:
-            log.warning("Could not load Patreon cookies from %s: %s", cookies_path, exc)
-    return session
-
-
-def _filehash(url: str) -> str | None:
-    if not url:
-        return None
-    match = _FILEHASH_RE.search(url)
-    return match.group(1).lower() if match else None
-
-
-def _basename_from_url(url: str) -> str:
-    """Derive a sane filename from a URL when the media has no file_name.
-
-    Strips query/fragment, takes the path basename, and drops a junk
-    extension (the importer._safe_ext gotcha) so we never write base64 noise
-    as a name. Falls back to the filehash, then to "file".
-    """
-    path = urlsplit(url).path
-    base = os.path.basename(path)
-    if base:
-        ext = safe_ext(base)
-        stem = base[: -len(Path(base).suffix)] if Path(base).suffix else base
-        # Keep the stem bounded; URL-encoded stems can be enormous.
-        stem = stem[:120] or "file"
-        return f"{stem}{ext}"
-    fh = _filehash(url)
-    return fh or "file"
-
-
-def parse_cursor_from_url(url: str | None) -> str | None:
-    """Extract the `page[cursor]` query param from a links.next URL."""
-    if not url:
-        return None
-    query = urlsplit(url).query
-    values = parse_qs(query).get("page[cursor]")
-    if values and values[0]:
-        return values[0]
-    return None
-
-
-class PatreonClient:
-    """Synchronous Patreon JSON:API read client.
-
-    Construct with a path to a Netscape cookies.txt (the same file
-    CredentialService.get_cookies_path materializes). Cookies are loaded into a
-    requests.Session; no secure-context APIs are used.
-    """
-
-    def __init__(
-        self,
-        cookies_path: str | Path | None,
-        *,
-        request_sleep: float = 0.0,
-        max_retries: int = _MAX_429_RETRIES,
-    ):
-        self.cookies_path = str(cookies_path) if cookies_path else None
-        self._session = _load_session(cookies_path)
-        # Politeness: seconds to sleep before each /api/posts page fetch (paces
-        # the rate-limited API endpoint). 0 = no pacing. plan #703.
-        self._request_sleep = request_sleep or 0.0
-        self._max_retries = max_retries
-
-    # -- request -----------------------------------------------------------
-
-    def _params(self, campaign_id: str, cursor: str | None) -> dict[str, str]:
-        params = {
-            "include": _INCLUDE,
-            "fields[post]": _FIELDS_POST,
-            "fields[media]": _FIELDS_MEDIA,
-            "fields[campaign]": _FIELDS_CAMPAIGN,
-            "filter[campaign_id]": campaign_id,
-            "filter[contains_exclusive_posts]": "true",
-            "filter[is_draft]": "false",
-            "sort": "-published_at",
-            "json-api-version": "1.0",
-        }
-        if cursor:
-            params["page[cursor]"] = cursor
-        return params
-
-    def _fetch(self, campaign_id: str, cursor: str | None) -> dict:
-        if self._request_sleep > 0:
-            time.sleep(self._request_sleep)  # pace the API endpoint
-        attempt = 0
-        while True:
-            try:
-                resp = self._session.get(
-                    _POSTS_URL,
-                    params=self._params(campaign_id, cursor),
-                    timeout=_TIMEOUT_SECONDS,
-                )
-            except requests.RequestException as exc:
-                raise PatreonAPIError(
-                    f"Patreon posts request failed (campaign_id={campaign_id}): {exc}"
-                ) from exc
-
-            # Transient rate-limit: back off and retry rather than failing the
-            # whole walk. Only a PERSISTENT 429 (retries exhausted) falls
-            # through to the terminal RATE_LIMITED raise below.
-            if resp.status_code == 429 and attempt < self._max_retries:
-                attempt += 1
-                delay = _retry_after_seconds(resp, attempt)
-                log.warning(
-                    "Patreon 429 (campaign_id=%s) — backing off %.1fs (retry %d/%d)",
-                    campaign_id, delay, attempt, self._max_retries,
-                )
-                time.sleep(delay)
-                continue
-            break
-
-        if resp.status_code in (401, 403):
-            # Auth rejected — expired/missing cookies or an insufficient tier.
-            # Actionable as "rotate credentials", so it's auth, not drift/http.
-            raise PatreonAuthError(
-                f"Patreon posts API returned HTTP {resp.status_code} — auth "
-                f"rejected (cookies expired or tier insufficient; "
-                f"campaign_id={campaign_id})",
-                status_code=resp.status_code,
-            )
-        if resp.status_code != 200:
-            # A persistent 429 (retries exhausted) is terminal RATE_LIMITED — carry
-            # the server's raw Retry-After seconds so the cooldown matches its hint
-            # (plan #708 B1). Header is uncapped here; the cooldown clamps it.
-            retry_after = None
-            if resp.status_code == 429:
-                hdr = resp.headers.get("Retry-After")
-                if hdr:
-                    try:
-                        retry_after = float(hdr)
-                    except (TypeError, ValueError):
-                        retry_after = None
-            raise PatreonAPIError(
-                f"Patreon posts API returned HTTP {resp.status_code} "
-                f"(campaign_id={campaign_id})",
-                status_code=resp.status_code,
-                retry_after=retry_after,
-            )
-        try:
-            payload = resp.json()
-        except ValueError as exc:
-            # A non-JSON body here is almost always the HTML login/challenge
-            # page served when cookies are missing/expired — that is an AUTH
-            # failure (rotate cookies), not API drift (update the ingester) and
-            # not a transient network error.
-            raise PatreonAuthError(
-                "Patreon posts API returned a non-JSON response (likely an "
-                f"HTML login/challenge page — session expired; "
-                f"campaign_id={campaign_id}): {exc}"
-            ) from exc
-        return payload
-
-    # -- parsing -----------------------------------------------------------
-
-    @staticmethod
-    def _transform(response: dict) -> dict:
-        """Flatten the JSON:API `included` array for relationship resolution.
-
-        Returns a dict keyed by `(type, id)` → that resource's `attributes`
-        (so a post's relationships can be resolved in O(1)). Missing/oddly
-        shaped `included` entries are skipped rather than fatal — drift
-        detection for the top-level shape lives in _validate_response.
-        """
-        index: dict[tuple[str, str], dict] = {}
-        for inc in response.get("included") or []:
-            if not isinstance(inc, dict):
-                continue
-            rtype = inc.get("type")
-            rid = inc.get("id")
-            if rtype is None or rid is None:
-                continue
-            index[(str(rtype), str(rid))] = inc.get("attributes") or {}
-        return index
-
-    @staticmethod
-    def _validate_response(response: dict) -> None:
-        if not isinstance(response, dict):
-            raise PatreonDriftError("Patreon response was not a JSON object")
-        if "data" not in response:
-            raise PatreonDriftError("Patreon response missing top-level 'data' key")
-        data = response.get("data")
-        if not isinstance(data, list):
-            raise PatreonDriftError("Patreon response 'data' was not a list")
-
-    def _related_ids(self, post: dict, rel_name: str) -> list[str]:
-        rels = post.get("relationships") or {}
-        rel = rels.get(rel_name) or {}
-        data = rel.get("data")
-        if isinstance(data, dict):  # to-one relationship
-            data = [data]
-        if not isinstance(data, list):
-            return []
-        ids: list[str] = []
-        for ref in data:
-            if isinstance(ref, dict) and ref.get("id") is not None:
-                ids.append(str(ref["id"]))
-        return ids
-
-    @staticmethod
-    def _media_url(attrs: dict) -> str | None:
-        """Pick the best fetchable URL for a media resource.
-
-        Prefer the full-size `download_url`; fall back to the largest
-        `image_urls` size. gallery-dl prefers download_url too, only dipping
-        into image_urls when a smaller configured size is requested — FC
-        always wants the original, so download_url first.
-        """
-        download_url = attrs.get("download_url")
-        if isinstance(download_url, str) and download_url:
-            return download_url
-        image_urls = attrs.get("image_urls")
-        if isinstance(image_urls, dict):
-            for key in ("original", "full", "large", "default"):
-                candidate = image_urls.get(key)
-                if isinstance(candidate, str) and candidate:
-                    return candidate
-            # Otherwise take any non-empty string value.
-            for candidate in image_urls.values():
-                if isinstance(candidate, str) and candidate:
-                    return candidate
-        return None
-
-    def _media_item(self, attrs: dict, kind: str, post_id: str) -> MediaItem:
-        url = self._media_url(attrs)
-        if not url:
-            raise PatreonDriftError(
-                f"Patreon media (post {post_id}, kind={kind}) had no resolvable URL "
-                f"(no download_url / image_urls)"
-            )
-        # file_name is OPTIONAL: Patreon legitimately serves some gallery images
-        # without it (operator-flagged 2026-06-07, BlenderKnight post 73665615),
-        # and the URL basename is a fine fallback — the same thing gallery-dl
-        # uses. A genuine schema change shows up as no URL (above) or a media id
-        # absent from `included` (caller), not a missing name.
-        file_name = attrs.get("file_name")
-        filename = file_name if isinstance(file_name, str) and file_name else _basename_from_url(url)
-        return MediaItem(
-            url=url,
-            filename=filename,
-            kind=kind,
-            filehash=_filehash(url),
-            post_id=post_id,
-        )
-
-    def extract_media(self, post: dict, included_index: dict) -> list[MediaItem]:
-        """Resolve all downloadable media for one post.
-
-        Walks the kinds in the same order gallery-dl does — images,
-        image_large (post cover), attachments, postfile, content (inline
-        <img>) — and dedups within the post by filehash (first wins). The
-        image_large cover commonly duplicates a gallery image; deduping by
-        filehash collapses them to the gallery item (encountered first).
-        """
-        post_id = str(post.get("id") or "")
-        attrs = post.get("attributes") or {}
-        items: list[MediaItem] = []
-
-        def _resolve_rel(rel_name: str, kind: str) -> None:
-            for mid in self._related_ids(post, rel_name):
-                media_attrs = included_index.get(("media", mid))
-                if media_attrs is None:
-                    # Referenced but not in `included`: a media id with no
-                    # resource is drift (we asked for include=media).
-                    raise PatreonDriftError(
-                        f"Patreon post {post_id} references media {mid} "
-                        f"({rel_name}) not present in 'included'"
-                    )
-                items.append(self._media_item(media_attrs, kind, post_id))
-
-        # 1. gallery images
-        _resolve_rel("images", "images")
-
-        # 2. image_large — the post-level cover (`image.large_url`). Not a
-        #    media relationship; it lives on the post attributes.
-        image = attrs.get("image")
-        if isinstance(image, dict):
-            large_url = image.get("large_url") or image.get("url")
-            if isinstance(large_url, str) and large_url:
-                items.append(
-                    MediaItem(
-                        url=large_url,
-                        filename=_basename_from_url(large_url),
-                        kind="image_large",
-                        filehash=_filehash(large_url),
-                        post_id=post_id,
-                    )
-                )
-
-        # 3. attachments
-        _resolve_rel("attachments_media", "attachments")
-
-        # 4. postfile — the post's primary attached file (`post_file`).
-        post_file = attrs.get("post_file")
-        if isinstance(post_file, dict):
-            pf_url = post_file.get("url") or post_file.get("download_url")
-            if isinstance(pf_url, str) and pf_url:
-                pf_name = post_file.get("name")
-                filename = (
-                    pf_name
-                    if isinstance(pf_name, str) and pf_name
-                    else _basename_from_url(pf_url)
-                )
-                items.append(
-                    MediaItem(
-                        url=pf_url,
-                        filename=filename,
-                        kind="postfile",
-                        filehash=_filehash(pf_url),
-                        post_id=post_id,
-                    )
-                )
-
-        # 5. content — inline <img> in the post HTML body.
-        content = attrs.get("content")
-        if isinstance(content, str) and content:
-            for raw_src in _CONTENT_IMG_RE.findall(content):
-                src = unescape(raw_src)
-                if not src:
-                    continue
-                items.append(
-                    MediaItem(
-                        url=src,
-                        filename=_basename_from_url(src),
-                        kind="content",
-                        filehash=_filehash(src),
-                        post_id=post_id,
-                    )
-                )
-
-        return _dedup_by_filehash(items)
-
-    @staticmethod
-    def post_meta(post: dict) -> dict:
-        """Title + published date for a post — for the preview sample (plan #708
-        B4). Part of the client contract `ingest_core.Ingester.preview` calls."""
-        attrs = post.get("attributes") or {}
-        title = attrs.get("title")
-        published = attrs.get("published_at")
-        return {
-            "title": title if isinstance(title, str) else None,
-            "date": published if isinstance(published, str) else None,
-        }
-
-    # -- iteration ---------------------------------------------------------
-
-    def iter_posts(
-        self, campaign_id: str, cursor: str | None = None
-    ) -> Iterator[tuple[dict, dict, str | None]]:
-        """Yield (post, included_index, page_cursor) for every post in the feed.
-
-        Pages newest→oldest via `links.next`, validating each response for
-        drift before yielding. The triple gives a caller everything it needs to
-        resolve and checkpoint without re-fetching:
-          - post           — the raw post resource.
-          - included_index — the page's flattened `included` (the same object
-                             for every post on a page), to pass straight to
-                             extract_media(post, included_index).
-          - page_cursor    — the cursor that FETCHED this post's page (None for
-                             the first page). The caller checkpoints THIS value,
-                             matching the existing backfill cursor logic where
-                             the saved cursor re-fetches the page being
-                             processed (so a chunk cut mid-page resumes the page,
-                             not the one after it).
-        """
-        current_cursor = cursor
-        while True:
-            response = self._fetch(campaign_id, current_cursor)
-            self._validate_response(response)
-            page_cursor = current_cursor
-            included_index = self._transform(response)
-            for post in response.get("data") or []:
-                if isinstance(post, dict):
-                    yield post, included_index, page_cursor
-            next_url = (response.get("links") or {}).get("next")
-            next_cursor = parse_cursor_from_url(next_url)
-            if not next_cursor:
-                return
-            current_cursor = next_cursor
-
-    # -- verify ------------------------------------------------------------
-
-    def verify_auth(self, campaign_id: str) -> tuple[bool | None, str]:
-        """Cheap auth probe: fetch the first `/api/posts` page and report whether
-        the credential authenticated, WITHOUT downloading anything.
-
-        Returns `(ok, message)` matching the credential-verify contract:
-          - True  — authenticated (the feed returned a valid JSON:API page).
-          - False — the credential was rejected (PatreonAuthError: 401/403, or an
-                    HTML login page → cookies expired / tier insufficient).
-          - None  — inconclusive: API drift (our parser is stale, not a cred
-                    problem) or a transient network/HTTP error.
-        """
-        try:
-            response = self._fetch(campaign_id, None)
-            self._validate_response(response)
-        except PatreonAuthError as exc:
-            return False, f"Patreon rejected the credential — {exc}"
-        except PatreonDriftError as exc:
-            return None, f"Couldn't verify — Patreon's API shape changed: {exc}"
-        except PatreonAPIError as exc:
-            return None, f"Couldn't verify (network/HTTP issue): {exc}"
-        return True, "Credentials valid — the Patreon feed authenticated."
-
-
-def _dedup_by_filehash(items: list[MediaItem]) -> list[MediaItem]:
-    """Drop later items sharing a filehash with an earlier one (first wins).
-
-    Items with no filehash (None) are never deduped against each other — we
-    can't prove they're the same file, so keep them all.
-    """
-    seen: set[str] = set()
-    out: list[MediaItem] = []
-    for item in items:
-        if item.filehash is not None:
-            if item.filehash in seen:
-                continue
-            seen.add(item.filehash)
-        out.append(item)
-    return out

backend/app/services/patreon_downloader.py

@@ -1,510 +0,0 @@
-"""Native Patreon media downloader (build step 2b of the native ingester).
-
-Given a Patreon post and its already-resolved `MediaItem`s (from
-patreon_client.extract_media), download the media to the EXACT on-disk layout
-gallery-dl produces, write a sidecar JSON the existing importer consumes, and
-report per-media outcomes.
-
-This module is PURE: no DB. The cross-run seen-ledger and the iter_posts
-orchestration are a LATER step. The tier-1 (seen) skip is an INJECTED predicate
-(`is_seen`), so this module needs no DB and is unit-testable without network.
-
-On-disk layout (matches gallery-dl):
-  <images_root>/<artist_slug>/patreon/<DIR>/<NN>_<filename>
-  where <DIR> = "<YYYY-MM-DD>_<post_id>_<title40>" (date prefix omitted when the
-  post's published_at is unparseable), <NN> is the 1-based index of the item in
-  the post zero-padded to 2, and <filename> is MediaItem.filename. The sidecar
-  is written next to the media as <NN>_<stem>.json (media_path.with_suffix).
-
-Video: a MediaItem whose URL is a Mux/HLS stream (host stream.mux.com or path
-endswith .m3u8) is fetched with yt-dlp (subprocess), passing the same
-Referer/Origin headers gallery-dl forwards for Mux playback (see gallery_dl.py
-_get_default_config). yt-dlp may remux to a container of its own choosing, so we
-accept the actual output extension and record the real path.
-
-FC runs on a plain-HTTP homelab; nothing here uses a secure-context Web API.
-"""
-
-from __future__ import annotations
-
-import contextlib
-import json
-import logging
-import os
-import subprocess
-import time
-from collections.abc import Callable
-from dataclasses import dataclass
-from datetime import datetime
-from pathlib import Path
-from urllib.parse import urlsplit
-
-import requests
-
-from .file_validator import is_validatable, quarantine_file, validate_file
-from .patreon_client import (
-    _BACKOFF_CAP_SECONDS,
-    _load_session,
-    _retry_after_seconds,
-)
-
-log = logging.getLogger(__name__)
-
-_TITLE_MAX = 40
-_TIMEOUT_SECONDS = 120.0
-_CHUNK = 1 << 16
-# Retry a media GET that hits a TRANSIENT failure within the same pass (plan
-# #705 #8): a transport blip (connection reset / timeout / truncated stream), a
-# 429, or a 5xx. PERMANENT failures (404 gone, 403 forbidden) fail fast straight
-# to the error/dead-letter path — no point re-fetching them. Keeps a momentary
-# network hiccup from becoming a per-item error that waits for the next walk.
-_MAX_MEDIA_RETRIES = 3
-# requests transport errors worth retrying (vs. an HTTPError, which is a real
-# server response and is classified by status code).
-_TRANSIENT_TRANSPORT_EXC = (
-    requests.ConnectionError,
-    requests.Timeout,
-    requests.exceptions.ChunkedEncodingError,
-)
-
-# Referer/Origin yt-dlp must send for Mux-hosted Patreon video. Mux's JWT
-# playback policy checks Referer/Origin on every request, so yt-dlp must send
-# Patreon's, not its own default. (gallery-dl forwarded the same headers before
-# the #697 cutover removed its Patreon path; this is now the only place they
-# live.)
-_VIDEO_HEADERS = {
-    "Referer": "https://www.patreon.com/",
-    "Origin": "https://www.patreon.com",
-}
-
-# Characters Windows/gallery-dl path-restrict forbids, plus path separators.
-_FORBIDDEN = set('<>:"/\\|?*')
-
-
-def _sanitize(name: str) -> str:
-    """Make `name` safe for a single filesystem path segment.
-
-    Replaces path separators, the Windows-forbidden set <>:"/\\|?* and control
-    characters with `_`, then strips trailing dots/spaces (gallery-dl
-    path-restrict behavior). Never returns empty (falls back to "_").
-    """
-    out = []
-    for ch in name:
-        if ch in _FORBIDDEN or ord(ch) < 32:
-            out.append("_")
-        else:
-            out.append(ch)
-    cleaned = "".join(out).rstrip(". ")
-    return cleaned or "_"
-
-
-def _is_video_url(url: str) -> bool:
-    parts = urlsplit(url)
-    if parts.hostname and parts.hostname.lower() == "stream.mux.com":
-        return True
-    return parts.path.lower().endswith(".m3u8")
-
-
-def _post_dir_name(post: dict) -> str:
-    """Build the post directory name matching gallery-dl's layout."""
-    post_id = str(post.get("id") or "")
-    attrs = post.get("attributes") or {}
-    title = attrs.get("title")
-    title = title if isinstance(title, str) else ""
-    title40 = title[:_TITLE_MAX]
-
-    published = attrs.get("published_at")
-    date_prefix = None
-    if isinstance(published, str) and published:
-        s = published.strip()
-        if s.endswith("Z"):
-            s = s[:-1] + "+00:00"
-        try:
-            dt = datetime.fromisoformat(s)
-        except ValueError:
-            dt = None
-        if dt is not None:
-            date_prefix = f"{dt:%Y-%m-%d}"
-
-    if date_prefix:
-        raw = f"{date_prefix}_{post_id}_{title40}"
-    else:
-        raw = f"{post_id}_{title40}"
-    return _sanitize(raw)
-
-
-@dataclass
-class MediaOutcome:
-    """Per-media result of a download_post pass.
-
-    status is one of: "downloaded", "skipped_seen", "skipped_disk",
-    "quarantined", "error". `path` is the final on-disk path for "downloaded"
-    (the actual yt-dlp output for video), the path that already existed for
-    "skipped_disk", or the _quarantine destination for "quarantined"; None for
-    "skipped_seen" and (usually) "error". `error` carries the failure/validation
-    reason for "error"/"quarantined", else None.
-    """
-
-    media: object  # MediaItem (avoid importing the name for a bare annotation)
-    status: str
-    path: Path | None
-    error: str | None
-
-
-class PatreonDownloader:
-    """Download resolved Patreon media to gallery-dl's on-disk layout.
-
-    PURE: no DB. The HTTP session and the yt-dlp invocation are injectable seams
-    so tests run without network or a real subprocess:
-      - pass `session=` to stub `session.get`, or monkeypatch `_fetch_to_file`.
-      - monkeypatch `_run_ytdlp` to avoid spawning yt-dlp.
-    """
-
-    def __init__(
-        self,
-        images_root: Path,
-        cookies_path: str | None = None,
-        *,
-        validate: bool = True,
-        rate_limit: float = 0.0,
-        session: requests.Session | None = None,
-    ):
-        self.images_root = Path(images_root)
-        self.cookies_path = str(cookies_path) if cookies_path else None
-        self._validate = validate
-        # Politeness: seconds to sleep before each actual media download (paces
-        # the CDN; honors ImportSettings.download_rate_limit_seconds, the same
-        # value gallery-dl used as its between-downloads `sleep`). 0 = no pacing.
-        # Applied only to real downloads, not to seen/disk skips. plan #703.
-        self._rate_limit = rate_limit or 0.0
-        # Build a cookie-loaded session the same way patreon_client does, so the
-        # CDN GETs carry the creator's auth.
-        self.session = session if session is not None else _load_session(cookies_path)
-
-    # -- public ------------------------------------------------------------
-
-    def download_post(
-        self,
-        post: dict,
-        media_items: list,
-        artist_slug: str,
-        *,
-        is_seen: Callable[[object], bool] = lambda m: False,
-        should_stop: Callable[[], bool] = lambda: False,
-    ) -> list[MediaOutcome]:
-        """Download every media item of one post; return per-item outcomes.
-
-        Builds the post directory, iterates media (1-based NN), applies the
-        two-tier skip (injected is_seen, then disk), downloads (plain GET or
-        yt-dlp for video), writes the sidecar for each freshly-downloaded item,
-        validates, and returns outcomes. Resilient: one media's failure yields
-        an "error" outcome for that item; the rest proceed.
-
-        `should_stop()` is polled BEFORE each media item: a media-dense post can
-        otherwise run a backfill chunk far past its time-box (the engine only
-        re-checks the budget between posts), so we honour the deadline mid-post
-        and return the items done so far — the rest re-fetch next chunk (they
-        were never marked seen). Bounds chunk overrun to one media download.
-        """
-        post_dir = self.images_root / artist_slug / "patreon" / _post_dir_name(post)
-        outcomes: list[MediaOutcome] = []
-
-        for i, media in enumerate(media_items, start=1):
-            if should_stop():
-                break
-            try:
-                outcomes.append(
-                    self._download_one(post, media, post_dir, artist_slug, i, is_seen)
-                )
-            except Exception as exc:  # resilient: isolate one item's failure
-                log.warning(
-                    "Patreon media failed (post %s, item %d): %s",
-                    post.get("id"), i, exc,
-                )
-                outcomes.append(
-                    MediaOutcome(media=media, status="error", path=None, error=str(exc))
-                )
-        return outcomes
-
-    # -- per-item ----------------------------------------------------------
-
-    def _download_one(
-        self,
-        post: dict,
-        media,
-        post_dir: Path,
-        artist_slug: str,
-        index: int,
-        is_seen: Callable[[object], bool],
-    ) -> MediaOutcome:
-        # tier-1: seen ledger (injected; no DB here).
-        if is_seen(media):
-            return MediaOutcome(media=media, status="skipped_seen", path=None, error=None)
-
-        nn = f"{index:02d}"
-        final_name = _sanitize(f"{nn}_{media.filename}")
-        media_path = post_dir / final_name
-
-        # tier-2: already on disk.
-        if media_path.exists():
-            return MediaOutcome(
-                media=media, status="skipped_disk", path=media_path, error=None
-            )
-        # Video may land at a different extension; honor a pre-existing remux.
-        if _is_video_url(media.url):
-            existing = self._existing_video_output(media_path)
-            if existing is not None:
-                return MediaOutcome(
-                    media=media, status="skipped_disk", path=existing, error=None
-                )
-
-        post_dir.mkdir(parents=True, exist_ok=True)
-
-        # Pace real downloads only (the skips above already returned). plan #703.
-        if self._rate_limit > 0:
-            time.sleep(self._rate_limit)
-
-        if _is_video_url(media.url):
-            out_path = self._run_ytdlp(media.url, media_path, _VIDEO_HEADERS)
-            if out_path is None or not Path(out_path).exists():
-                return MediaOutcome(
-                    media=media,
-                    status="error",
-                    path=None,
-                    error="yt-dlp produced no output",
-                )
-            out_path = Path(out_path)
-        else:
-            out_path = self._fetch_get(media.url, media_path)
-            reason, quarantine_dest = self._validate_path(
-                out_path, artist_slug, media.url
-            )
-            if reason is not None:
-                # Quarantined (corrupt/invalid) — distinct from a download error
-                # so the run can report a real files_quarantined count + paths.
-                return MediaOutcome(
-                    media=media, status="quarantined",
-                    path=quarantine_dest, error=reason,
-                )
-
-        self._write_sidecar(post, out_path)
-        return MediaOutcome(media=media, status="downloaded", path=out_path, error=None)
-
-    # -- download seams ----------------------------------------------------
-
-    def _fetch_get(self, url: str, dest: Path) -> Path:
-        """Stream `url` to a .part file then atomic-rename to `dest`.
-
-        Thin wrapper over `_fetch_to_file` so tests can stub either the whole
-        GET path (`_fetch_to_file`) or just `session.get`.
-        """
-        part = dest.with_name(dest.name + ".part")
-        try:
-            self._fetch_to_file(url, part)
-        except Exception:
-            with contextlib.suppress(OSError):
-                part.unlink()
-            raise
-        os.replace(part, dest)
-        return dest
-
-    def _fetch_to_file(self, url: str, dest: Path) -> None:
-        """Stream a non-video URL to `dest` via the (stubbable) session, retrying
-        TRANSIENT failures within the same pass (plan #705 #8) and RESUMING from
-        the bytes already on disk via a Range request when a retry follows a
-        mid-download cut (plan #708 B5).
-
-        Retried (backoff): transport blips (connection reset / timeout /
-        truncated stream — incl. mid-download), HTTP 429 (honoring Retry-After),
-        and 5xx. Failed fast (no retry → HTTPError → per-item error → dead-letter
-        path): 4xx other than 429 (404 gone, 403 forbidden) — re-fetching a
-        permanent failure is pointless.
-
-        Resume: on a retry, if bytes already landed in `dest`, ask for the rest
-        with `Range: bytes=<have>-`. A 206 means the server honored it → append; a
-        200 means it ignored it (served the whole file) → start clean. The caller
-        (_fetch_get) stages into a `.part`, so a non-range server never corrupts
-        the output — the worst case is re-downloading from zero, as before.
-        """
-        attempt = 0
-        while True:
-            have = dest.stat().st_size if dest.exists() else 0
-            headers = {"Range": f"bytes={have}-"} if have > 0 else None
-            try:
-                resp = self.session.get(
-                    url, stream=True, timeout=_TIMEOUT_SECONDS, headers=headers,
-                )
-                if (resp.status_code == 429 or resp.status_code >= 500) \
-                        and attempt < _MAX_MEDIA_RETRIES:
-                    attempt += 1
-                    delay = _retry_after_seconds(resp, attempt)
-                    log.warning(
-                        "Patreon media transient HTTP %d (%s) — backing off "
-                        "%.1fs (retry %d/%d)",
-                        resp.status_code, url, delay, attempt, _MAX_MEDIA_RETRIES,
-                    )
-                    time.sleep(delay)
-                    continue
-                # A Range that starts at/past EOF (we already have the whole file)
-                # comes back 416 — the bytes we kept ARE the file.
-                if have > 0 and resp.status_code == 416:
-                    return
-                # 2xx → ok; 4xx-non-429 (or an exhausted 429/5xx) → HTTPError
-                # (permanent for this pass) → not caught below → per-item error.
-                resp.raise_for_status()
-                # 206 → server honored the Range; append after the kept bytes.
-                # Anything else (200) → it served the whole file → start clean.
-                mode = "ab" if (have > 0 and resp.status_code == 206) else "wb"
-                with open(dest, mode) as fh:
-                    for chunk in resp.iter_content(chunk_size=_CHUNK):
-                        if chunk:
-                            fh.write(chunk)
-                return
-            except _TRANSIENT_TRANSPORT_EXC as exc:
-                if attempt >= _MAX_MEDIA_RETRIES:
-                    raise  # exhausted → terminal error outcome
-                attempt += 1
-                delay = min(2.0 * (2 ** (attempt - 1)), _BACKOFF_CAP_SECONDS)
-                log.warning(
-                    "Patreon media transport error (%s) — backing off %.1fs "
-                    "(retry %d/%d): %s",
-                    url, delay, attempt, _MAX_MEDIA_RETRIES, exc,
-                )
-                time.sleep(delay)
-
-    def _run_ytdlp(self, url: str, dest: Path, headers: dict) -> Path | None:
-        """Invoke yt-dlp to fetch a Mux/HLS stream to (around) `dest`.
-
-        Returns the actual output path (yt-dlp may remux to a different
-        container, so we resolve the real file afterward). Overridden/
-        monkeypatched in tests to avoid spawning a real subprocess.
-
-        The output template uses `dest` without its extension; yt-dlp appends
-        the chosen container extension. We pass Referer/Origin (Mux JWT policy)
-        and the cookies file.
-
-        Mirrors the GET path's transient/permanent split (plan #705 #8): a hung
-        fetch (TimeoutExpired) or a spawn failure (OSError) is TRANSIENT — back
-        off and retry. A non-zero yt-dlp exit (CalledProcessError) is treated as
-        PERMANENT for this pass — yt-dlp already does its OWN internal network
-        retries, so a non-zero exit is effectively a real failure (private/gone/
-        geo-blocked), like a 4xx on the GET path: fail fast to the per-item error
-        → dead-letter path.
-        """
-        dest = Path(dest)
-        out_template = str(dest.with_suffix("")) + ".%(ext)s"
-        cmd = ["yt-dlp", "--no-progress", "-o", out_template]
-        for key, value in headers.items():
-            cmd += ["--add-header", f"{key}:{value}"]
-        if self.cookies_path and os.path.isfile(self.cookies_path):
-            cmd += ["--cookies", self.cookies_path]
-        cmd.append(url)
-        attempt = 0
-        while True:
-            try:
-                subprocess.run(
-                    cmd,
-                    check=True,
-                    capture_output=True,
-                    text=True,
-                    timeout=_TIMEOUT_SECONDS,
-                )
-                break
-            except subprocess.CalledProcessError as exc:
-                # Permanent for this pass — fail fast (no retry).
-                log.warning(
-                    "yt-dlp failed (exit %s) for %s: %s",
-                    exc.returncode, url, (exc.stderr or "").strip() or exc,
-                )
-                return None
-            except (OSError, subprocess.TimeoutExpired) as exc:
-                # Transient — back off and retry, like a transport blip on a GET.
-                if attempt >= _MAX_MEDIA_RETRIES:
-                    log.warning(
-                        "yt-dlp transient failure exhausted for %s: %s", url, exc
-                    )
-                    return None
-                attempt += 1
-                delay = min(2.0 * (2 ** (attempt - 1)), _BACKOFF_CAP_SECONDS)
-                log.warning(
-                    "yt-dlp transient failure (%s) — backing off %.1fs "
-                    "(retry %d/%d): %s",
-                    url, delay, attempt, _MAX_MEDIA_RETRIES, exc,
-                )
-                time.sleep(delay)
-        return self._existing_video_output(dest)
-
-    def _existing_video_output(self, dest: Path) -> Path | None:
-        """Find a yt-dlp output for `dest` regardless of chosen extension.
-
-        Returns `dest` itself if present, else any sibling sharing the same
-        stem (the remuxed container). None if nothing matched.
-        """
-        dest = Path(dest)
-        if dest.exists():
-            return dest
-        stem = dest.with_suffix("").name
-        parent = dest.parent
-        if not parent.is_dir():
-            return None
-        for cand in sorted(parent.iterdir()):
-            if cand.is_file() and cand.stem == stem and cand.name != dest.name:
-                return cand
-        return None
-
-    # -- validation --------------------------------------------------------
-
-    def _validate_path(
-        self, path: Path, artist_slug: str, source_url: str | None = None
-    ) -> tuple[str | None, Path | None]:
-        """Validate a freshly-written file; quarantine if bad.
-
-        Uses the shared `file_validator.quarantine_file` — same move + provenance
-        sidecar gallery-dl writes (the native path used to skip the sidecar; that
-        parity gap is closed here). Returns `(reason, quarantine_dest)` when
-        quarantined (dest is the original path if the move itself failed), else
-        `(None, None)` (ok / not validatable / disabled). plan #704: the dest is
-        surfaced so the run reports a real quarantined-paths list.
-        """
-        if not self._validate or not is_validatable(path):
-            return None, None
-        try:
-            result = validate_file(path)
-        except Exception as exc:
-            log.warning("Validator raised on %s: %s", path, exc)
-            return None, None
-        if result.ok:
-            return None, None
-        dest = quarantine_file(
-            self.images_root, path, artist_slug, "patreon",
-            url=source_url, result=result,
-        )
-        return (result.reason or "validation failed"), (dest or path)
-
-    # -- sidecar -----------------------------------------------------------
-
-    def _write_sidecar(self, post: dict, media_path: Path) -> Path:
-        """Write the importer-consumed sidecar next to `media_path`.
-
-        Patreon uses base-default sidecar keys (parse_sidecar maps
-        category->platform, id->external_post_id, title->post_title,
-        content->description, published_at->post_date, url->post_url). Patreon
-        registers no derive_post_url, so `url` is trusted as the permalink — we
-        pass the post's attributes.url.
-        """
-        attrs = post.get("attributes") or {}
-        title = attrs.get("title")
-        content = attrs.get("content")
-        published = attrs.get("published_at")
-        url = attrs.get("url")
-        data = {
-            "category": "patreon",
-            "id": str(post.get("id") or ""),
-            "title": title if isinstance(title, str) else "",
-            "content": content if isinstance(content, str) else "",
-            "published_at": published if isinstance(published, str) else None,
-            "url": url if isinstance(url, str) else None,
-        }
-        sidecar_path = media_path.with_suffix(".json")
-        sidecar_path.write_text(json.dumps(data, indent=2))
-        return sidecar_path

backend/app/services/patreon_ingester.py

@@ -1,197 +0,0 @@
-"""Native Patreon ingester — the Patreon ADAPTER over the platform-agnostic core.
-
-The orchestration that drives a native subscription walk (page a feed → extract
-media → tiered skip → download → mark-seen / record-failures / checkpoint-cursor
-→ return a gallery-dl-shaped `DownloadResult`, across tick/backfill/recovery)
-now lives in `ingest_core.Ingester` — it's identical for every platform. This
-module is the thin Patreon adapter: it wires the Patreon `client`/`downloader`/
-ledger models/constraints/key into the core and supplies the Patreon-specific
-failure mapping. `download_service.download_source` calls `PatreonIngester.run`
-exactly as before; the public surface (this class, `_ledger_key`,
-`DEAD_LETTER_THRESHOLD`, `verify_patreon_credential`) is unchanged.
-
-Three modes (selected by `download_service` from `config_overrides` state):
-  - tick      — newest→oldest, skip seen (tier-1 ledger + tier-2 disk), early-out
-                after N contiguous already-have-it items (the cheap native
-                equivalent of gallery-dl's `exit:20`, now free of per-file HEADs).
-  - backfill  — full-history walk in a time-boxed chunk, resuming from the
-                pagination cursor checkpoint; reaches the bottom → "complete".
-  - recovery  — like backfill but BYPASSES the tier-1 seen-ledger AND the
-                dead-letter ledger, so deliberately-dropped-and-deleted near-dups
-                get re-fetched and re-evaluated under the current pHash threshold
-                (tier-2 disk skip still spares files we kept).
-
-The seen/dead-letter ledgers live in Postgres (`patreon_seen_media` /
-`patreon_failed_media`); the core opens SHORT-LIVED sync sessions per page batch
-— never held across a network fetch ([[db-connection-held-across-subprocess]]).
-FC runs on a plain-HTTP homelab; nothing here uses a secure-context Web API.
-"""
-
-from __future__ import annotations
-
-import asyncio
-import logging
-from collections.abc import Callable
-from pathlib import Path
-
-from ..models import PatreonFailedMedia, PatreonSeenMedia
-from .gallery_dl import DownloadResult, ErrorType
-from .ingest_core import DEAD_LETTER_THRESHOLD, Ingester
-from .patreon_client import (
-    MediaItem,
-    PatreonAPIError,
-    PatreonAuthError,
-    PatreonClient,
-    PatreonDriftError,
-)
-from .patreon_downloader import PatreonDownloader
-from .patreon_resolver import extract_vanity, resolve_campaign_id_for_source
-
-__all__ = [
-    "DEAD_LETTER_THRESHOLD",
-    "PatreonIngester",
-    "_ledger_key",
-    "verify_patreon_credential",
-]
-
-log = logging.getLogger(__name__)
-
-# Ledger keys are stored in patreon_seen_media.filehash VARCHAR(128); bound any
-# synthesized key so a pathologically long file_name can't overflow the column.
-_LEDGER_KEY_MAX = 128
-
-
-def _ledger_key(media: MediaItem) -> str:
-    """Stable per-media identity for the cross-run seen-ledger.
-
-    A Patreon CDN URL carries a 32-char MD5 (`media.filehash`) — that is the
-    natural key. Some media have none: Mux/HLS video (`stream.mux.com`, no
-    content hash at discovery) and the odd inline-content `<img>` pointing at a
-    hashless URL. The plan calls the video case the ``video:<post_id>:<media_id>``
-    sentinel; `MediaItem` carries no media_id, so the post-scoped filename is the
-    stable proxy. Bounded to the column width.
-    """
-    if media.filehash:
-        return media.filehash
-    return f"{media.post_id}:{media.filename}"[:_LEDGER_KEY_MAX]
-
-
-class PatreonIngester(Ingester):
-    """Walk a Patreon campaign's posts, download unseen media, return a
-    `DownloadResult`. A thin adapter over `ingest_core.Ingester`.
-
-    Construct with the per-source `cookies_path` and a sync sessionmaker for the
-    ledgers. `client` / `downloader` are injectable seams so unit tests run
-    without network, subprocess, or a real CDN.
-    """
-
-    def __init__(
-        self,
-        images_root: Path,
-        cookies_path: str | None,
-        session_factory: Callable[[], object],
-        *,
-        validate: bool = True,
-        rate_limit: float = 0.0,
-        request_sleep: float = 0.0,
-        client: PatreonClient | None = None,
-        downloader: PatreonDownloader | None = None,
-    ):
-        self.images_root = Path(images_root)
-        self.cookies_path = str(cookies_path) if cookies_path else None
-        # Pacing (plan #703): request_sleep paces the API page fetches,
-        # rate_limit paces the media downloads. Injected client/downloader (in
-        # tests) already carry their own pacing, so these only apply to the
-        # default-constructed ones.
-        resolved_client = (
-            client
-            if client is not None
-            else PatreonClient(cookies_path, request_sleep=request_sleep)
-        )
-        resolved_downloader = (
-            downloader
-            if downloader is not None
-            else PatreonDownloader(
-                self.images_root, cookies_path, validate=validate, rate_limit=rate_limit
-            )
-        )
-        super().__init__(
-            client=resolved_client,
-            downloader=resolved_downloader,
-            session_factory=session_factory,
-            seen_model=PatreonSeenMedia,
-            failed_model=PatreonFailedMedia,
-            seen_constraint="uq_patreon_seen_media_source_id",
-            failed_constraint="uq_patreon_failed_media_source_id",
-            ledger_key=_ledger_key,
-            platform="patreon",
-            error_base=PatreonAPIError,
-        )
-
-    # -- failure mapping (Patreon exception taxonomy) ----------------------
-
-    def _failure_result(self, exc: Exception, _result) -> DownloadResult:
-        """Map a client-level exception to a loud, typed failed DownloadResult.
-
-        We NEVER return a silent zero-download "success" — the whole point of the
-        native ingester is to fail RED when Patreon's API shape or our auth
-        changes. The typed mapping lets FailingSourcesCard render the right chip
-        and tells the operator what to do:
-          - PatreonAuthError  → AUTH_ERROR  (rotate cookies)
-          - PatreonDriftError → API_DRIFT   (ingester field-set/parser needs update)
-          - HTTP 429 / 404    → RATE_LIMITED / NOT_FOUND
-          - other HTTP status → HTTP_ERROR;  transport failure → NETWORK_ERROR
-
-        PatreonAuthError and PatreonDriftError both subclass PatreonAPIError, so
-        they must be matched before the generic HTTP/transport fallthrough.
-        """
-        message = str(exc)
-        if isinstance(exc, PatreonAuthError):
-            error_type = ErrorType.AUTH_ERROR
-        elif isinstance(exc, PatreonDriftError):
-            error_type = ErrorType.API_DRIFT
-            message = f"Patreon API changed — ingester needs update: {message}"
-        else:  # generic PatreonAPIError: HTTP non-2xx (status_code set) or transport
-            status = getattr(exc, "status_code", None)
-            if status == 429:
-                error_type = ErrorType.RATE_LIMITED
-            elif status == 404:
-                error_type = ErrorType.NOT_FOUND
-            elif status is not None:
-                error_type = ErrorType.HTTP_ERROR
-            else:
-                error_type = ErrorType.NETWORK_ERROR
-        log.warning("Patreon ingest failed (%s): %s", error_type.value, message)
-        result = _result(
-            success=False, return_code=1,
-            error_type=error_type, error_message=message,
-        )
-        # plan #708 B1: carry the server's Retry-After up to the cooldown.
-        if error_type == ErrorType.RATE_LIMITED:
-            result.retry_after_seconds = getattr(exc, "retry_after", None)
-        return result
-
-
-async def verify_patreon_credential(
-    url: str,
-    cookies_path: str | None,
-    overrides: dict | None,
-) -> tuple[bool | None, str]:
-    """Native Patreon credential probe — the verify counterpart to the ingester's
-    download path, sharing its campaign-id resolution. Resolves the campaign id
-    (override / id: URL / vanity) then does ONE authenticated `/api/posts` fetch
-    via PatreonClient.verify_auth. Returns the uniform `(ok, message)` contract
-    (True / False / None) so download_backends.verify_credential can treat it
-    interchangeably with the gallery-dl probe. No download, no DB.
-    """
-    campaign_id, _ = await resolve_campaign_id_for_source(url, cookies_path, overrides)
-    if not campaign_id:
-        vanity = extract_vanity(url)
-        return None, (
-            f"Couldn't resolve the Patreon campaign id — can't verify. "
-            f"source_url={url!r}; vanity={vanity!r} "
-            "(cookies expired, or the creator moved/renamed?)."
-        )
-    client = PatreonClient(cookies_path)
-    loop = asyncio.get_running_loop()
-    return await loop.run_in_executor(None, client.verify_auth, campaign_id)

backend/app/services/patreon_resolver.py

@@ -19,56 +19,18 @@ import asyncio
 import http.cookiejar
 import logging
 import os
-import re
 
 import requests
 
 log = logging.getLogger(__name__)
 
 _CAMPAIGNS_URL = "https://www.patreon.com/api/campaigns"
-_POSTS_API = "https://www.patreon.com/api/posts"
-
-# A source URL of the form `.../id:<digits>` already carries the campaign id
-# (no lookup needed). The vanity regex deliberately EXCLUDES the id: form so the
-# two paths don't overlap.
-#
-# Patreon serves the creator vanity under several path prefixes — bare
-# (`patreon.com/Atole`), `c/` (`patreon.com/c/Atole`), and `cw/`
-# (`patreon.com/cw/Atole`, its current "creator workspace" URL). The optional
-# prefix group must list `cw/` BEFORE `c/` so the longer prefix wins — otherwise
-# `cw/Atole` matches the bare branch and yields vanity="cw" (operator-flagged
-# 2026-06-07: every `/cw/` source failed resolution on vanity="cw").
-_ID_URL_RE = re.compile(r"/id:(\d+)")
-# A single-post permalink — patreon.com/posts/<slug>-<post_id> (or bare
-# /posts/<post_id>). The trailing digits are the post id; the creator's
-# campaign is resolved from the post itself (operator-flagged 2026-06-07: a
-# /posts/ source resolved vanity="posts"). `posts/` is excluded from the vanity
-# regex so it never masquerades as a creator slug.
-_POST_URL_RE = re.compile(r"/posts/(?:[^/?#]*-)?(\d+)(?:[/?#]|$)")
-_VANITY_PREFIXES = ("cw/", "c/")
-_VANITY_RE = re.compile(
-    r"^https?://(?:www\.)?patreon\.com/(?:cw/|c/)?(?!(?:id:|posts/))([^/?#]+)",
-    re.IGNORECASE,
-)
 _USER_AGENT = (
     "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 "
     "(KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36"
 )
 _TIMEOUT_SECONDS = 10.0
 
-# Fallback resolution: Patreon's `/api/campaigns?filter[vanity]=` lookup has
-# proven unreliable (returns empty `data` for creators that clearly exist —
-# operator-flagged 2026-06-06). gallery-dl never used that endpoint; it scrapes
-# the campaign id out of the creator page's bootstrap JSON. We do the same as a
-# fallback: fetch the creator page HTML and pull the first campaign id out of
-# any of these embeddings (ordered most- to least-specific).
-_PAGE_CAMPAIGN_ID_PATTERNS = (
-    re.compile(r'"id":\s*"(\d+)",\s*"type":\s*"campaign"'),
-    re.compile(r'"campaign":\s*\{\s*"data":\s*\{\s*"id":\s*"(\d+)"'),
-    re.compile(r"/api/campaigns/(\d+)"),
-    re.compile(r'"campaign_id":\s*"?(\d+)'),
-)
-
 
 def _load_cookie_jar(cookies_path: str | None) -> http.cookiejar.MozillaCookieJar | None:
     if not cookies_path or not os.path.isfile(cookies_path):
@@ -83,15 +45,6 @@ def _load_cookie_jar(cookies_path: str | None) -> http.cookiejar.MozillaCookieJa
 
 
 def _sync_lookup(vanity: str, cookies_path: str | None) -> str | None:
-    """Resolve a vanity to a campaign id: try the campaigns API first (cheap,
-    structured), then fall back to scraping the creator page (robust against the
-    API's empty-data failures). Returns None only when both miss."""
-    return _lookup_via_api(vanity, cookies_path) or _lookup_via_page(
-        vanity, cookies_path
-    )
-
-
-def _lookup_via_api(vanity: str, cookies_path: str | None) -> str | None:
     jar = _load_cookie_jar(cookies_path)
     headers = {
         "User-Agent": _USER_AGENT,
@@ -139,87 +92,6 @@ def _lookup_via_api(vanity: str, cookies_path: str | None) -> str | None:
     return campaign_id
 
 
-def _scrape_campaign_id(html: str) -> str | None:
-    """First campaign id found in creator-page HTML via the known embeddings."""
-    if not isinstance(html, str):
-        return None
-    for pat in _PAGE_CAMPAIGN_ID_PATTERNS:
-        m = pat.search(html)
-        if m:
-            return m.group(1)
-    return None
-
-
-def _lookup_via_page(vanity: str, cookies_path: str | None) -> str | None:
-    """Fallback: GET the creator page and scrape the campaign id from the page
-    bootstrap (gallery-dl's method). Tries both the bare and `/c/` vanity paths
-    Patreon redirects between. Never raises."""
-    jar = _load_cookie_jar(cookies_path)
-    headers = {"User-Agent": _USER_AGENT, "Accept": "text/html"}
-    # Try the bare vanity and every known creator-path prefix Patreon
-    # redirects between (c/, cw/) — the one the source used isn't known here
-    # (extract_vanity already stripped it).
-    page_urls = [f"https://www.patreon.com/{vanity}"]
-    page_urls += [f"https://www.patreon.com/{p}{vanity}" for p in _VANITY_PREFIXES]
-    for page_url in page_urls:
-        try:
-            resp = requests.get(
-                page_url,
-                headers=headers,
-                cookies=jar,
-                timeout=_TIMEOUT_SECONDS,
-                allow_redirects=True,
-            )
-        except requests.RequestException as exc:
-            log.warning("Patreon creator-page fetch failed for %s: %s", page_url, exc)
-            continue
-        if resp.status_code != 200:
-            continue
-        campaign_id = _scrape_campaign_id(resp.text)
-        if campaign_id:
-            log.info(
-                "Resolved Patreon vanity=%s → campaign_id=%s via creator page",
-                vanity, campaign_id,
-            )
-            return campaign_id
-    return None
-
-
-def _lookup_campaign_from_post(post_id: str, cookies_path: str | None) -> str | None:
-    """Resolve the owning campaign id from a single post id via the Patreon
-    post API (`/api/posts/<id>?include=campaign`). A `/posts/` source URL points
-    at one post, but a subscription walks the whole creator — so we follow the
-    post to its campaign. Never raises."""
-    jar = _load_cookie_jar(cookies_path)
-    headers = {"User-Agent": _USER_AGENT, "Accept": "application/vnd.api+json"}
-    params = {"include": "campaign", "fields[campaign]": "name"}
-    url = f"{_POSTS_API}/{post_id}"
-    try:
-        resp = requests.get(
-            url, params=params, headers=headers, cookies=jar,
-            timeout=_TIMEOUT_SECONDS,
-        )
-    except requests.RequestException as exc:
-        log.warning("Patreon post lookup failed for post=%s: %s", post_id, exc)
-        return None
-    if resp.status_code != 200:
-        log.warning("Patreon post API returned HTTP %d for post=%s", resp.status_code, post_id)
-        return None
-    try:
-        payload = resp.json()
-    except ValueError:
-        return None
-    data = payload.get("data") if isinstance(payload, dict) else None
-    if not isinstance(data, dict):
-        return None
-    campaign = ((data.get("relationships") or {}).get("campaign") or {}).get("data") or {}
-    campaign_id = campaign.get("id")
-    if isinstance(campaign_id, str) and campaign_id:
-        log.info("Resolved Patreon post=%s → campaign_id=%s", post_id, campaign_id)
-        return campaign_id
-    return None
-
-
 async def resolve_campaign_id(
     vanity: str,
     cookies_path: str | None,
@@ -228,55 +100,3 @@ async def resolve_campaign_id(
     Never raises."""
     loop = asyncio.get_running_loop()
     return await loop.run_in_executor(None, _sync_lookup, vanity, cookies_path)
-
-
-async def resolve_campaign_from_post(
-    post_id: str, cookies_path: str | None,
-) -> str | None:
-    """Async wrapper around _lookup_campaign_from_post. Never raises."""
-    loop = asyncio.get_running_loop()
-    return await loop.run_in_executor(
-        None, _lookup_campaign_from_post, post_id, cookies_path
-    )
-
-
-def extract_vanity(url: str) -> str | None:
-    """The vanity slug from a Patreon creator URL, or None for an `id:` URL."""
-    m = _VANITY_RE.match(url or "")
-    return m.group(1) if m else None
-
-
-async def resolve_campaign_id_for_source(
-    url: str,
-    cookies_path: str | None,
-    overrides: dict | None,
-) -> tuple[str | None, str | None]:
-    """Resolve a Patreon source to its campaign id — the single resolution path
-    shared by the download ingester and the credential-verify probe.
-
-    Order: cached `patreon_campaign_id` override → an `id:<digits>` URL → a
-    `/posts/<id>` permalink (resolve the owning campaign from the post) → a
-    vanity lookup against the campaigns API. Returns
-    `(campaign_id, newly_resolved_id)`: `newly_resolved_id` is non-None whenever
-    a lookup actually ran, so the caller caches it on the source (the
-    override/id: paths needed no lookup). `(None, None)` when unresolvable.
-    Never raises.
-    """
-    overrides = overrides or {}
-    cached = overrides.get("patreon_campaign_id")
-    if cached:
-        return cached, None
-    id_match = _ID_URL_RE.search(url or "")
-    if id_match:
-        return id_match.group(1), None
-    # A single-post URL → follow the post to its creator's campaign so the
-    # source subscribes to the whole feed (a subscription isn't one post).
-    post_match = _POST_URL_RE.search(url or "")
-    if post_match:
-        resolved = await resolve_campaign_from_post(post_match.group(1), cookies_path)
-        return resolved, resolved
-    vanity = extract_vanity(url)
-    if vanity:
-        resolved = await resolve_campaign_id(vanity, cookies_path)
-        return resolved, resolved
-    return None, None

backend/app/services/platform_lock.py

@@ -1,59 +0,0 @@
-"""Per-platform download concurrency cap.
-
-Some platforms (Patreon) are API-rate-sensitive enough that two *simultaneous*
-walks can trip the server's rate limit even with each source pacing its own
-requests. The platform-cooldown handles the AFTERMATH of a 429; this is the
-preventive half — it serializes downloads PER PLATFORM to one at a time.
-
-Different platforms still run concurrently up to the worker's concurrency; only
-a second walk on the SAME serialized platform waits. The lock lives in Redis
-(the Celery broker) with a TTL, so a SIGKILL'd worker can't wedge a platform —
-the lock auto-expires shortly after the download hard time limit.
-"""
-from __future__ import annotations
-
-import logging
-
-import redis
-
-from ..config import get_config
-
-log = logging.getLogger(__name__)
-
-# Platforms walked one-at-a-time. gallery-dl platforms are intentionally NOT
-# here: each runs as a self-pacing subprocess and they're lower-volume. Add a
-# platform here to cap it to a single concurrent walk.
-SERIALIZED_PLATFORMS = frozenset({"patreon"})
-
-_LOCK_PREFIX = "fc:download_lock:"
-
-_client: redis.Redis | None = None
-
-
-def _redis() -> redis.Redis:
-    # One client per worker process (Celery prefork forks before tasks run, so
-    # each process lazily builds its own). redis-py pools connections.
-    global _client
-    if _client is None:
-        _client = redis.from_url(get_config().celery_broker_url)
-    return _client
-
-
-def platform_lock(platform: str, *, ttl_seconds: int):
-    """A non-blocking Redis lock for `platform`, or None when the platform is
-    not serialized. Caller does `.acquire(blocking=False)` / `.release()`.
-
-    Returns None (rather than raising) on any Redis error so a broker hiccup
-    degrades to the prior behaviour (uncapped) instead of stalling downloads.
-    """
-    if platform not in SERIALIZED_PLATFORMS:
-        return None
-    try:
-        return _redis().lock(
-            f"{_LOCK_PREFIX}{platform}",
-            timeout=ttl_seconds,
-            blocking=False,
-        )
-    except redis.RedisError as exc:  # pragma: no cover - broker outage
-        log.warning("platform_lock unavailable for %s: %s", platform, exc)
-        return None

backend/app/services/series_match_service.py

@@ -1,337 +0,0 @@
-"""FC-6.3 — assisted continuation matcher.
-
-Scores a (post, candidate series) pair from several weighted signals; above the
-configured threshold it records a *pending* SeriesSuggestion. Confirm-only — the
-operator accepts (post becomes a chapter) or dismisses. No single signal gates;
-the score is an additive weighted sum, each signal a 0..1 strength.
-
-The learned "title pattern" isn't persisted — it's derived on the fly from the
-post titles already in a series, so it sharpens automatically as more posts are
-confirmed into the series.
-"""
-
-import re
-import time
-
-from sqlalchemy import and_, func, select
-from sqlalchemy.dialects.postgresql import insert as pg_insert
-from sqlalchemy.ext.asyncio import AsyncSession
-
-from ..models import ImageRecord, Post, Tag, TagKind
-from ..models.series_chapter import SeriesChapter
-from ..models.series_page import SeriesPage
-from ..models.series_suggestion import SeriesSuggestion
-from ..models.tag import image_tag
-from .page_number_parser import parse_page_range
-from .series_service import SeriesError, SeriesService
-
-# Additive signal weights (sum to 1.0 → max score 1.0). Kept as constants;
-# only the on/off + threshold are operator-tunable (sensitivity is the knob
-# that matters; per-signal weights are an over-tune for v1).
-WEIGHTS = {"title": 0.40, "artist": 0.20, "pages": 0.25, "tags": 0.15}
-
-_DISTINCTIVE_KINDS = (TagKind.character, TagKind.series, TagKind.fandom)
-_MAX_CANDIDATES = 50
-
-# Strip installment markers so titles collapse to their stable stem.
-_PAGE_TOKEN = re.compile(r"\b(?:pages?|pgs?|pp\.?)\s*\d+(?:\s*[-–—/]\s*\d+)?", re.I)
-_BRACKET_NUM = re.compile(r"[\[(]\s*\d+\s*(?:/\s*\d+)?\s*[\])]")
-_TRAILING_NUM = re.compile(r"[\s\-_#]*\d+\s*$")
-
-
-def normalize_title(title: str | None) -> str:
-    if not title:
-        return ""
-    t = _PAGE_TOKEN.sub("", title)
-    t = _BRACKET_NUM.sub("", t)
-    t = _TRAILING_NUM.sub("", t)
-    return " ".join(t.split()).strip().lower()
-
-
-def _common_prefix(strings: list[str]) -> str:
-    strings = [s for s in strings if s]
-    if not strings:
-        return ""
-    pre = strings[0]
-    for s in strings[1:]:
-        i = 0
-        while i < len(pre) and i < len(s) and pre[i] == s[i]:
-            i += 1
-        pre = pre[:i]
-        if not pre:
-            break
-    return pre.strip()
-
-
-def title_signal(series_titles: list[str], post_title: str | None) -> float:
-    """Overlap of the post title against the series' title stem (the longest
-    common prefix of its known titles, page/installment markers removed)."""
-    norm = [normalize_title(t) for t in series_titles]
-    norm = [t for t in norm if t]
-    pt = normalize_title(post_title)
-    if not norm or not pt:
-        return 0.0
-    stem = _common_prefix(norm)
-    if len(stem) < 3:
-        stem = max(norm, key=len)
-    i = 0
-    while i < len(stem) and i < len(pt) and stem[i] == pt[i]:
-        i += 1
-    return min(1.0, i / max(len(stem), 4))
-
-
-def pages_signal(series_max_stated_end: int | None, post_start: int | None) -> float:
-    """1.0 when the post's first page continues right after the series' last
-    stated page; partial for a near-continuation; 0 otherwise."""
-    if series_max_stated_end is None or post_start is None:
-        return 0.0
-    diff = post_start - series_max_stated_end
-    if diff == 1:
-        return 1.0
-    if 2 <= diff <= 3:
-        return 0.5
-    return 0.0
-
-
-def tags_signal(shared_distinctive: int) -> float:
-    if shared_distinctive <= 0:
-        return 0.0
-    return min(1.0, shared_distinctive / 3.0)
-
-
-def weighted_score(signals: dict) -> float:
-    return round(sum(WEIGHTS[k] * signals.get(k, 0.0) for k in WEIGHTS), 4)
-
-
-class SeriesMatchService:
-    def __init__(self, session: AsyncSession):
-        self.session = session
-
-    async def _post_image_ids(self, post_id: int) -> list[int]:
-        rows = (
-            await self.session.execute(
-                select(ImageRecord.id).where(ImageRecord.primary_post_id == post_id)
-            )
-        ).scalars().all()
-        return list(rows)
-
-    async def _series_image_ids(self, series_tag_id: int) -> set[int]:
-        rows = (
-            await self.session.execute(
-                select(SeriesPage.image_id).where(
-                    SeriesPage.series_tag_id == series_tag_id
-                )
-            )
-        ).scalars().all()
-        return set(rows)
-
-    async def _series_post_titles(self, series_tag_id: int) -> list[str]:
-        rows = (
-            await self.session.execute(
-                select(Post.post_title)
-                .select_from(SeriesPage)
-                .join(ImageRecord, ImageRecord.id == SeriesPage.image_id)
-                .join(Post, Post.id == ImageRecord.primary_post_id)
-                .where(
-                    and_(
-                        SeriesPage.series_tag_id == series_tag_id,
-                        Post.post_title.isnot(None),
-                    )
-                )
-                .distinct()
-            )
-        ).scalars().all()
-        return [t for t in rows if t]
-
-    async def _series_max_stated_end(self, series_tag_id: int) -> int | None:
-        return await self.session.scalar(
-            select(func.max(SeriesChapter.stated_page_end)).where(
-                SeriesChapter.series_tag_id == series_tag_id
-            )
-        )
-
-    async def _distinctive_tags(self, image_ids: list[int] | set[int]) -> set[int]:
-        ids = list(image_ids)
-        if not ids:
-            return set()
-        rows = (
-            await self.session.execute(
-                select(image_tag.c.tag_id)
-                .join(Tag, Tag.id == image_tag.c.tag_id)
-                .where(
-                    and_(
-                        image_tag.c.image_record_id.in_(ids),
-                        Tag.kind.in_(_DISTINCTIVE_KINDS),
-                    )
-                )
-                .distinct()
-            )
-        ).scalars().all()
-        return set(rows)
-
-    async def _candidate_series(self, artist_id: int) -> list[int]:
-        """Series that already contain a page by this artist — cross-artist
-        series are rare, so same-artist is the candidate bound."""
-        rows = (
-            await self.session.execute(
-                select(SeriesPage.series_tag_id)
-                .join(ImageRecord, ImageRecord.id == SeriesPage.image_id)
-                .where(ImageRecord.artist_id == artist_id)
-                .distinct()
-            )
-        ).scalars().all()
-        return list(rows)
-
-    async def _decided_series(self, post_id: int) -> set[int]:
-        rows = (
-            await self.session.execute(
-                select(SeriesSuggestion.series_tag_id).where(
-                    and_(
-                        SeriesSuggestion.post_id == post_id,
-                        SeriesSuggestion.status.in_(["added", "dismissed"]),
-                    )
-                )
-            )
-        ).scalars().all()
-        return set(rows)
-
-    async def match_post(self, post_id: int, *, threshold: float) -> int:
-        """Score a post against its artist's series; upsert pending suggestions
-        for those at/above threshold. Returns the number written."""
-        post = await self.session.get(Post, post_id)
-        if post is None or post.artist_id is None:
-            return 0
-        post_images = await self._post_image_ids(post_id)
-        if not post_images:
-            return 0
-        post_image_set = set(post_images)
-        rng = parse_page_range(f"{post.post_title or ''} {post.description or ''}")
-        post_start = rng[0] if rng else None
-        post_dtags = await self._distinctive_tags(post_images)
-
-        decided = await self._decided_series(post_id)
-        candidates = await self._candidate_series(post.artist_id)
-        made = 0
-        for sid in candidates[:_MAX_CANDIDATES]:
-            if sid in decided:
-                continue
-            s_images = await self._series_image_ids(sid)
-            if post_image_set & s_images:
-                continue  # the post is already (partly) in this series
-            signals = {
-                "title": title_signal(
-                    await self._series_post_titles(sid), post.post_title
-                ),
-                "artist": 1.0,  # candidates are same-artist by construction
-                "pages": pages_signal(
-                    await self._series_max_stated_end(sid), post_start
-                ),
-                "tags": tags_signal(
-                    len(post_dtags & await self._distinctive_tags(s_images))
-                ),
-            }
-            score = weighted_score(signals)
-            if score < threshold:
-                continue
-            await self.session.execute(
-                pg_insert(SeriesSuggestion)
-                .values(
-                    post_id=post_id, series_tag_id=sid,
-                    score=score, signals=signals, status="pending",
-                )
-                .on_conflict_do_update(
-                    constraint="uq_series_suggestion_post_series",
-                    set_={"score": score, "signals": signals, "status": "pending"},
-                    where=SeriesSuggestion.status == "pending",
-                )
-            )
-            made += 1
-        return made
-
-    # ---- queue ops --------------------------------------------------------
-
-    async def list_pending(self) -> list[dict]:
-        rows = (
-            await self.session.execute(
-                select(
-                    SeriesSuggestion.id,
-                    SeriesSuggestion.post_id,
-                    SeriesSuggestion.series_tag_id,
-                    SeriesSuggestion.score,
-                    SeriesSuggestion.signals,
-                    Post.post_title,
-                    Post.external_post_id,
-                    Tag.name.label("series_name"),
-                )
-                .join(Post, Post.id == SeriesSuggestion.post_id)
-                .join(Tag, Tag.id == SeriesSuggestion.series_tag_id)
-                .where(SeriesSuggestion.status == "pending")
-                .order_by(SeriesSuggestion.score.desc())
-            )
-        ).all()
-        return [
-            {
-                "id": r.id,
-                "post_id": r.post_id,
-                "series_tag_id": r.series_tag_id,
-                "series_name": r.series_name,
-                "post_title": r.post_title or f"Post {r.external_post_id}",
-                "score": r.score,
-                "signals": r.signals or {},
-            }
-            for r in rows
-        ]
-
-    async def accept(self, suggestion_id: int) -> dict:
-        s = await self.session.get(SeriesSuggestion, suggestion_id)
-        if s is None:
-            raise SeriesError(f"suggestion {suggestion_id} not found")
-        if s.status != "pending":
-            raise SeriesError(f"suggestion {suggestion_id} is already {s.status}")
-        out = await SeriesService(self.session).add_post_as_chapter(
-            s.series_tag_id, s.post_id
-        )
-        s.status = "added"
-        self.session.add(s)
-        return out
-
-    async def dismiss(self, suggestion_id: int) -> None:
-        s = await self.session.get(SeriesSuggestion, suggestion_id)
-        if s is None:
-            raise SeriesError(f"suggestion {suggestion_id} not found")
-        if s.status == "pending":
-            s.status = "dismissed"
-            self.session.add(s)
-
-    async def rescan(
-        self, *, threshold: float, time_budget_seconds: float | None = None,
-        after_post_id: int = 0,
-    ) -> dict:
-        """Score every post (id > after_post_id) against its artist's series,
-        time-boxed + resumable like the other long maintenance sweeps."""
-        post_ids = (
-            await self.session.execute(
-                select(Post.id)
-                .where(Post.id > after_post_id)
-                .order_by(Post.id.asc())
-            )
-        ).scalars().all()
-        summary = {
-            "scanned": 0, "suggested": 0,
-            "partial": False, "resume_after_id": after_post_id,
-        }
-        start = time.monotonic()
-        for pid in post_ids:
-            summary["scanned"] += 1
-            summary["resume_after_id"] = pid
-            summary["suggested"] += await self.match_post(pid, threshold=threshold)
-            await self.session.commit()  # commit per post so progress survives
-            if (
-                time_budget_seconds is not None
-                and time.monotonic() - start >= time_budget_seconds
-            ):
-                summary["partial"] = True
-                break
-        else:
-            summary["partial"] = False
-        return summary

backend/app/services/series_service.py

@@ -1,8 +1,4 @@
-"""Series = a series-kind Tag + ordered chapters, each holding ordered pages.
-
-Reading order is (series_chapter.chapter_number, series_page.page_number). A
-chapter may be a placeholder (no pages) reserving a slot. An image lives in at
-most one series ⇒ one chapter (series_page.image_id is UNIQUE).
+"""Series = a series-kind Tag + ordered series_page membership.
 
 All mutations are Core/set-based and run in the request transaction.
 """
@@ -11,11 +7,9 @@ from sqlalchemy import and_, func, select, update
 from sqlalchemy.dialects.postgresql import insert as pg_insert
 from sqlalchemy.ext.asyncio import AsyncSession
 
-from ..models import Artist, ImageRecord, Post, Tag, TagKind
-from ..models.series_chapter import SeriesChapter
+from ..models import ImageRecord, Tag, TagKind
 from ..models.series_page import SeriesPage
 from .gallery_service import thumbnail_url
-from .page_number_parser import parse_page_range
 
 
 class SeriesError(ValueError):
@@ -27,8 +21,6 @@ class SeriesService:
     def __init__(self, session: AsyncSession):
         self.session = session
 
-    # ---- guards / helpers -------------------------------------------------
-
     async def _require_series(self, series_tag_id: int) -> Tag:
         tag = await self.session.get(Tag, series_tag_id)
         if tag is None:
@@ -40,7 +32,7 @@ class SeriesService:
     @staticmethod
     def _clean_ids(ids: list[int]) -> list[int]:
         if not ids:
-            raise SeriesError("ids must be a non-empty list")
+            raise SeriesError("image_ids must be a non-empty list")
         if len(ids) > 500:
             raise SeriesError("selection too large (max 500)")
         seen: set[int] = set()
@@ -51,217 +43,51 @@ class SeriesService:
                 out.append(int(x))
         return out
 
-    async def _require_chapter(self, series_tag_id: int, chapter_id: int):
-        row = (
-            await self.session.execute(
-                select(SeriesChapter).where(
-                    and_(
-                        SeriesChapter.id == chapter_id,
-                        SeriesChapter.series_tag_id == series_tag_id,
-                    )
-                )
-            )
-        ).scalar_one_or_none()
-        if row is None:
-            raise SeriesError(
-                f"chapter {chapter_id} is not in series {series_tag_id}"
-            )
-        return row
-
-    async def _ensure_default_chapter(self, series_tag_id: int) -> int:
-        """Lowest-numbered chapter of the series, creating a first chapter if
-        the series has none (a fresh series, or the legacy single-chapter path)."""
-        existing = await self.session.scalar(
-            select(SeriesChapter.id)
-            .where(SeriesChapter.series_tag_id == series_tag_id)
-            .order_by(SeriesChapter.chapter_number.asc())
-            .limit(1)
-        )
-        if existing is not None:
-            return existing
-        res = await self.session.execute(
-            pg_insert(SeriesChapter)
-            .values(series_tag_id=series_tag_id, chapter_number=1)
-            .returning(SeriesChapter.id)
-        )
-        return res.scalar_one()
-
-    async def _chapter_page_order(self, chapter_id: int) -> list[int]:
+    async def _member_order(self, series_tag_id: int) -> list[int]:
         rows = (
             await self.session.execute(
                 select(SeriesPage.image_id)
-                .where(SeriesPage.chapter_id == chapter_id)
+                .where(SeriesPage.series_tag_id == series_tag_id)
                 .order_by(SeriesPage.page_number.asc())
             )
         ).scalars().all()
         return list(rows)
 
-    async def _chapter_order(self, series_tag_id: int) -> list[int]:
-        rows = (
-            await self.session.execute(
-                select(SeriesChapter.id)
-                .where(SeriesChapter.series_tag_id == series_tag_id)
-                .order_by(SeriesChapter.chapter_number.asc())
-            )
-        ).scalars().all()
-        return list(rows)
-
-    # ---- read -------------------------------------------------------------
-
-    @staticmethod
-    def _gaps(chapters: list[dict]) -> list[dict]:
-        """Missing-page gaps between consecutive chapters with stated ranges."""
-        out: list[dict] = []
-        prev = None
-        for ch in chapters:
-            start = ch["stated_page_start"]
-            if (
-                prev is not None
-                and prev["stated_page_end"] is not None
-                and start is not None
-                and start > prev["stated_page_end"] + 1
-            ):
-                out.append(
-                    {
-                        "after_chapter_id": prev["id"],
-                        "start": prev["stated_page_end"] + 1,
-                        "end": start - 1,
-                    }
-                )
-            prev = ch
-        return out
-
-    @staticmethod
-    def _part_gaps(chapters: list[dict]) -> list[dict]:
-        """Missing-Part gaps between consecutive chapters whose stated_part
-        numbers jump by more than 1 (e.g. a series with Part 1 and Part 3, or one
-        authored straight from a Part 2 post). Mirrors _gaps but on stated_part —
-        only chapters that actually carry a stated_part participate."""
-        out: list[dict] = []
-        prev = None
-        for ch in chapters:
-            cur = ch["stated_part"]
-            if cur is None:
-                continue
-            if prev is not None and cur > prev["stated_part"] + 1:
-                out.append(
-                    {
-                        "after_chapter_id": prev["id"],
-                        "start": prev["stated_part"] + 1,
-                        "end": cur - 1,
-                    }
-                )
-            prev = ch
-        return out
-
     async def list_pages(self, series_tag_id: int) -> dict:
         tag = await self._require_series(series_tag_id)
         rows = (
             await self.session.execute(
                 select(
-                    SeriesChapter.id.label("chapter_id"),
-                    SeriesChapter.chapter_number,
-                    SeriesChapter.stated_part,
-                    SeriesChapter.title,
-                    SeriesChapter.is_placeholder,
-                    SeriesChapter.stated_page_start,
-                    SeriesChapter.stated_page_end,
                     SeriesPage.image_id,
                     SeriesPage.page_number,
-                    SeriesPage.stated_page,
                     ImageRecord.sha256,
                     ImageRecord.mime,
                     ImageRecord.path,
                     ImageRecord.thumbnail_path,
-                    ImageRecord.primary_post_id,
-                    Post.post_title,
-                )
-                .select_from(SeriesChapter)
-                .outerjoin(SeriesPage, SeriesPage.chapter_id == SeriesChapter.id)
-                .outerjoin(ImageRecord, ImageRecord.id == SeriesPage.image_id)
-                .outerjoin(Post, Post.id == ImageRecord.primary_post_id)
-                .where(SeriesChapter.series_tag_id == series_tag_id)
-                .order_by(
-                    SeriesChapter.chapter_number.asc(),
-                    SeriesPage.page_number.asc(),
                 )
+                .join(ImageRecord, ImageRecord.id == SeriesPage.image_id)
+                .where(SeriesPage.series_tag_id == series_tag_id)
+                .order_by(SeriesPage.page_number.asc())
             )
         ).all()
-
-        chapters: list[dict] = []
-        flat: list[dict] = []
-        by_id: dict[int, dict] = {}
-        # chapter_id -> {post_id: title} seen across its pages, so we can label a
-        # chapter with its source post when all its pages come from one post.
-        posts_seen: dict[int, dict[int, str | None]] = {}
-        for r in rows:
-            ch = by_id.get(r.chapter_id)
-            if ch is None:
-                ch = {
-                    "id": r.chapter_id,
-                    "chapter_number": r.chapter_number,
-                    "stated_part": r.stated_part,
-                    "title": r.title,
-                    "is_placeholder": r.is_placeholder,
-                    "stated_page_start": r.stated_page_start,
-                    "stated_page_end": r.stated_page_end,
-                    "source_post": None,
-                    "pages": [],
-                }
-                by_id[r.chapter_id] = ch
-                posts_seen[r.chapter_id] = {}
-                chapters.append(ch)
-            if r.image_id is None:
-                continue  # placeholder / empty chapter
-            if r.primary_post_id is not None:
-                posts_seen[r.chapter_id][r.primary_post_id] = r.post_title
-            page = {
-                "image_id": r.image_id,
-                "chapter_id": r.chapter_id,
-                "page_number": r.page_number,
-                "stated_page": r.stated_page,
-                "thumbnail_url": thumbnail_url(r.thumbnail_path, r.sha256, r.mime),
-                "image_url": f"/images/{r.path.split('/images/', 1)[-1]}",
-            }
-            ch["pages"].append(page)
-            flat.append(page)
-
-        # A chapter's source_post is set only when every page shares one post —
-        # the common case (a series authored from a post). Mixed chapters stay
-        # null rather than guessing.
-        for ch in chapters:
-            seen = posts_seen.get(ch["id"], {})
-            if len(seen) == 1:
-                pid, title = next(iter(seen.items()))
-                ch["source_post"] = {"id": pid, "title": title}
-
         return {
             "series": {"id": tag.id, "name": tag.name},
-            "chapters": chapters,
-            "pages": flat,  # back-compat: flat reading order across chapters
-            "gaps": self._gaps(chapters),
-            "part_gaps": self._part_gaps(chapters),
+            "pages": [
+                {
+                    "image_id": r.image_id,
+                    "page_number": r.page_number,
+                    "thumbnail_url": thumbnail_url(r.thumbnail_path, r.sha256, r.mime),
+                    "image_url": f"/images/{r.path.split('/images/', 1)[-1]}",
+                }
+                for r in rows
+            ],
         }
 
-    # ---- pages ------------------------------------------------------------
-
     async def add_images(
-        self,
-        series_tag_id: int,
-        image_ids: list[int],
-        chapter_id: int | None = None,
-        stated_pages: dict[int, int] | None = None,
+        self, series_tag_id: int, image_ids: list[int]
     ) -> int:
-        """Append images as pages of a chapter (the series' default chapter when
-        chapter_id is None). Images already in THIS series are left untouched;
-        images in another series are moved here (image_id is UNIQUE)."""
         await self._require_series(series_tag_id)
         ids = self._clean_ids(image_ids)
-        if chapter_id is None:
-            chapter_id = await self._ensure_default_chapter(series_tag_id)
-        else:
-            await self._require_chapter(series_tag_id, chapter_id)
-
         existing = dict(
             (
                 await self.session.execute(
@@ -270,30 +96,29 @@ class SeriesService:
                 )
             ).all()
         )
+        # Already in THIS series → leave untouched (no churn).
         to_add = [i for i in ids if existing.get(i) != series_tag_id]
         if not to_add:
             return 0
-        # Move: drop any prior membership for these images (UNIQUE image_id).
+        # Remove any prior membership for the to_add images (the "move").
         await self.session.execute(
-            SeriesPage.__table__.delete().where(SeriesPage.image_id.in_(to_add))
+            SeriesPage.__table__.delete().where(
+                SeriesPage.image_id.in_(to_add)
+            )
         )
         max_pn = (
             await self.session.scalar(
-                select(func.coalesce(func.max(SeriesPage.page_number), 0)).where(
-                    SeriesPage.chapter_id == chapter_id
-                )
+                select(func.coalesce(func.max(SeriesPage.page_number), 0))
+                .where(SeriesPage.series_tag_id == series_tag_id)
             )
         ) or 0
-        sp = stated_pages or {}
         await self.session.execute(
             pg_insert(SeriesPage).values(
                 [
                     {
                         "series_tag_id": series_tag_id,
-                        "chapter_id": chapter_id,
                         "image_id": iid,
                         "page_number": max_pn + offset,
-                        "stated_page": sp.get(iid),
                     }
                     for offset, iid in enumerate(to_add, start=1)
                 ]
@@ -316,469 +141,32 @@ class SeriesService:
         )
         return res.rowcount or 0
 
-    async def reorder_pages(
-        self, series_tag_id: int, chapter_id: int, ordered_image_ids: list[int]
+    async def reorder(
+        self, series_tag_id: int, ordered_image_ids: list[int]
     ) -> None:
         await self._require_series(series_tag_id)
-        await self._require_chapter(series_tag_id, chapter_id)
         ordered = self._clean_ids(ordered_image_ids)
-        current = set(await self._chapter_page_order(chapter_id))
+        current = set(await self._member_order(series_tag_id))
         if set(ordered) != current or len(ordered) != len(current):
             raise SeriesError(
-                "ordered image_ids must exactly match the chapter's pages"
+                "ordered image_ids must exactly match series membership"
             )
         for idx, iid in enumerate(ordered, start=1):
             await self.session.execute(
                 update(SeriesPage)
                 .where(
                     and_(
-                        SeriesPage.chapter_id == chapter_id,
+                        SeriesPage.series_tag_id == series_tag_id,
                         SeriesPage.image_id == iid,
                     )
                 )
                 .values(page_number=idx)
             )
 
-    async def reorder(
-        self, series_tag_id: int, ordered_image_ids: list[int]
-    ) -> None:
-        """Legacy series-wide reorder — reorders the default chapter. Valid only
-        for single-chapter series (the only shape the pre-chapter UI produced);
-        multi-chapter series must use reorder_pages with an explicit chapter."""
-        await self._require_series(series_tag_id)
-        chapters = await self._chapter_order(series_tag_id)
-        if len(chapters) > 1:
-            raise SeriesError(
-                "series has multiple chapters; reorder a specific chapter"
-            )
-        chapter_id = (
-            chapters[0]
-            if chapters
-            else await self._ensure_default_chapter(series_tag_id)
-        )
-        await self.reorder_pages(series_tag_id, chapter_id, ordered_image_ids)
-
-    # ---- chapters ---------------------------------------------------------
-
-    async def create_chapter(
-        self,
-        series_tag_id: int,
-        *,
-        title: str | None = None,
-        is_placeholder: bool = False,
-        stated_page_start: int | None = None,
-        stated_page_end: int | None = None,
-    ) -> dict:
-        await self._require_series(series_tag_id)
-        max_cn = (
-            await self.session.scalar(
-                select(
-                    func.coalesce(func.max(SeriesChapter.chapter_number), 0)
-                ).where(SeriesChapter.series_tag_id == series_tag_id)
-            )
-        ) or 0
-        res = await self.session.execute(
-            pg_insert(SeriesChapter)
-            .values(
-                series_tag_id=series_tag_id,
-                chapter_number=max_cn + 1,
-                title=title,
-                is_placeholder=is_placeholder,
-                stated_page_start=stated_page_start,
-                stated_page_end=stated_page_end,
-            )
-            .returning(SeriesChapter.id, SeriesChapter.chapter_number)
-        )
-        row = res.one()
-        return {"id": row.id, "chapter_number": row.chapter_number}
-
-    async def update_chapter(
-        self,
-        series_tag_id: int,
-        chapter_id: int,
-        *,
-        title: str | None = None,
-        stated_part: int | None = None,
-        stated_page_start: int | None = None,
-        stated_page_end: int | None = None,
-        set_title: bool = False,
-        set_part: bool = False,
-        set_start: bool = False,
-        set_end: bool = False,
-    ) -> None:
-        """Partial chapter edit. The set_* flags say which fields to write (so
-        None can be written explicitly, e.g. clearing a stated page or part)."""
-        await self._require_series(series_tag_id)
-        await self._require_chapter(series_tag_id, chapter_id)
-        values: dict = {}
-        if set_title:
-            values["title"] = title
-        if set_part:
-            values["stated_part"] = stated_part
-        if set_start:
-            values["stated_page_start"] = stated_page_start
-        if set_end:
-            values["stated_page_end"] = stated_page_end
-        if not values:
-            return
-        await self.session.execute(
-            update(SeriesChapter)
-            .where(SeriesChapter.id == chapter_id)
-            .values(**values)
-        )
-
-    async def _renumber_chapters(self, series_tag_id: int) -> None:
-        for idx, cid in enumerate(
-            await self._chapter_order(series_tag_id), start=1
-        ):
-            await self.session.execute(
-                update(SeriesChapter)
-                .where(SeriesChapter.id == cid)
-                .values(chapter_number=idx)
-            )
-
-    async def reorder_chapters(
-        self, series_tag_id: int, ordered_chapter_ids: list[int]
-    ) -> None:
-        await self._require_series(series_tag_id)
-        ordered = self._clean_ids(ordered_chapter_ids)
-        current = set(await self._chapter_order(series_tag_id))
-        if set(ordered) != current or len(ordered) != len(current):
-            raise SeriesError(
-                "ordered chapter_ids must exactly match the series' chapters"
-            )
-        for idx, cid in enumerate(ordered, start=1):
-            await self.session.execute(
-                update(SeriesChapter)
-                .where(SeriesChapter.id == cid)
-                .values(chapter_number=idx)
-            )
-
-    async def delete_chapter(self, series_tag_id: int, chapter_id: int) -> None:
-        await self._require_series(series_tag_id)
-        await self._require_chapter(series_tag_id, chapter_id)
-        # Pages cascade-delete with the chapter (FK ondelete=CASCADE).
-        await self.session.execute(
-            SeriesChapter.__table__.delete().where(
-                SeriesChapter.id == chapter_id
-            )
-        )
-        await self._renumber_chapters(series_tag_id)
-
-    async def merge_chapter(
-        self, series_tag_id: int, source_chapter_id: int, target_chapter_id: int
-    ) -> int:
-        """Move source chapter's pages onto the end of the target chapter, then
-        delete the now-empty source chapter. Returns the number of pages moved."""
-        await self._require_series(series_tag_id)
-        if source_chapter_id == target_chapter_id:
-            raise SeriesError("cannot merge a chapter into itself")
-        await self._require_chapter(series_tag_id, source_chapter_id)
-        await self._require_chapter(series_tag_id, target_chapter_id)
-        moving = await self._chapter_page_order(source_chapter_id)
-        max_pn = (
-            await self.session.scalar(
-                select(func.coalesce(func.max(SeriesPage.page_number), 0)).where(
-                    SeriesPage.chapter_id == target_chapter_id
-                )
-            )
-        ) or 0
-        for offset, iid in enumerate(moving, start=1):
-            await self.session.execute(
-                update(SeriesPage)
-                .where(
-                    and_(
-                        SeriesPage.chapter_id == source_chapter_id,
-                        SeriesPage.image_id == iid,
-                    )
-                )
-                .values(chapter_id=target_chapter_id, page_number=max_pn + offset)
-            )
-        await self.session.execute(
-            SeriesChapter.__table__.delete().where(
-                SeriesChapter.id == source_chapter_id
-            )
-        )
-        await self._renumber_chapters(series_tag_id)
-        return len(moving)
-
     async def set_cover(self, series_tag_id: int, image_id: int) -> None:
-        """Cover = the (chapter_number=1, page_number=1) image. Bring the image's
-        chapter to the front and the image to the front of its chapter."""
         await self._require_series(series_tag_id)
-        row = (
-            await self.session.execute(
-                select(SeriesPage.chapter_id)
-                .where(
-                    and_(
-                        SeriesPage.series_tag_id == series_tag_id,
-                        SeriesPage.image_id == image_id,
-                    )
-                )
-            )
-        ).scalar_one_or_none()
-        if row is None:
+        order = await self._member_order(series_tag_id)
+        if image_id not in order:
             raise SeriesError(f"image {image_id} is not in this series")
-        chapter_id = row
-        chapters = await self._chapter_order(series_tag_id)
-        if chapters and chapters[0] != chapter_id:
-            new_chapters = [chapter_id] + [c for c in chapters if c != chapter_id]
-            await self.reorder_chapters(series_tag_id, new_chapters)
-        pages = await self._chapter_page_order(chapter_id)
-        if pages and pages[0] != image_id:
-            new_pages = [image_id] + [p for p in pages if p != image_id]
-            await self.reorder_pages(series_tag_id, chapter_id, new_pages)
-
-    # ---- post → series (FC-6.2) ------------------------------------------
-
-    async def _post_images_ordered(self, post_id: int) -> list[int]:
-        """A post's images in capture order (ImageRecord.primary_post_id), the
-        same ordering the posts feed renders thumbnails in."""
-        rows = (
-            await self.session.execute(
-                select(ImageRecord.id)
-                .where(ImageRecord.primary_post_id == post_id)
-                .order_by(ImageRecord.id.asc())
-            )
-        ).scalars().all()
-        return list(rows)
-
-    @staticmethod
-    def _stated_map(image_ids: list[int], start: int | None) -> dict | None:
-        """Per-image stated_page = start, start+1, ... when the post stated a
-        starting page; None when it didn't (fall back to capture order)."""
-        if start is None:
-            return None
-        return {iid: start + i for i, iid in enumerate(image_ids)}
-
-    async def promote_post_to_series(self, post_id: int) -> dict:
-        """Case 1: a self-contained multi-image post becomes its own series.
-        Creates a series tag named after the post, one chapter, the post's
-        images as its pages (stated pages parsed from the post when present)."""
-        from .tag_service import TagService
-
-        post = await self.session.get(Post, post_id)
-        if post is None:
-            raise SeriesError(f"post {post_id} not found")
-        image_ids = await self._post_images_ordered(post_id)
-        if not image_ids:
-            raise SeriesError("post has no images to make a series from")
-        name = (post.post_title or f"Series from post {post_id}").strip()[:200]
-        tag = await TagService(self.session).find_or_create(name, TagKind.series)
-        rng = parse_page_range(f"{post.post_title or ''} {post.description or ''}")
-        start, end = rng if rng else (None, None)
-        ch = await self.create_chapter(
-            tag.id, stated_page_start=start, stated_page_end=end,
-        )
-        added = await self.add_images(
-            tag.id, image_ids, chapter_id=ch["id"],
-            stated_pages=self._stated_map(image_ids, start),
-        )
-        return {
-            "series_tag_id": tag.id, "name": tag.name,
-            "chapter_id": ch["id"], "added": added,
-        }
-
-    async def add_post_as_chapter(
-        self, series_tag_id: int, post_id: int
-    ) -> dict:
-        """Case 2: append a post as the next chapter of an existing series. The
-        chapter is titled after the post and slotted by parsed page number when
-        present (else appended at the end)."""
-        await self._require_series(series_tag_id)
-        post = await self.session.get(Post, post_id)
-        if post is None:
-            raise SeriesError(f"post {post_id} not found")
-        image_ids = await self._post_images_ordered(post_id)
-        if not image_ids:
-            raise SeriesError("post has no images to add")
-        rng = parse_page_range(f"{post.post_title or ''} {post.description or ''}")
-        start, end = rng if rng else (None, None)
-        ch = await self.create_chapter(
-            series_tag_id, title=post.post_title or None,
-            stated_page_start=start, stated_page_end=end,
-        )
-        added = await self.add_images(
-            series_tag_id, image_ids, chapter_id=ch["id"],
-            stated_pages=self._stated_map(image_ids, start),
-        )
-        if start is not None:
-            await self._place_chapter_by_stated(series_tag_id, ch["id"], start)
-        return {
-            "series_tag_id": series_tag_id, "chapter_id": ch["id"],
-            "added": added,
-        }
-
-    async def _place_chapter_by_stated(
-        self, series_tag_id: int, chapter_id: int, start: int
-    ) -> None:
-        """Move `chapter_id` to sit before the first chapter whose stated start
-        is higher — so a post stating pages 5-8 lands ahead of the 9-12 chapter."""
-        rows = (
-            await self.session.execute(
-                select(SeriesChapter.id, SeriesChapter.stated_page_start)
-                .where(SeriesChapter.series_tag_id == series_tag_id)
-                .order_by(SeriesChapter.chapter_number.asc())
-            )
-        ).all()
-        current = [r.id for r in rows]
-        others = [r for r in rows if r.id != chapter_id]
-        insert_at = len(others)
-        for idx, r in enumerate(others):
-            if r.stated_page_start is not None and r.stated_page_start > start:
-                insert_at = idx
-                break
-        new_order = [r.id for r in others]
-        new_order.insert(insert_at, chapter_id)
-        if new_order != current:
-            await self.reorder_chapters(series_tag_id, new_order)
-
-    # ---- browse list (FC-6.2) --------------------------------------------
-
-    async def list_series(
-        self, *, sort: str = "recent", artist_id: int | None = None
-    ) -> list[dict]:
-        """Series cards for the browse view: cover thumb, name, artist, chapter
-        + page counts, a gap flag, and last-updated. sort ∈ recent|name|size."""
-        # Page counts + most-recent activity per series.
-        page_rows = (
-            await self.session.execute(
-                select(
-                    SeriesPage.series_tag_id,
-                    func.count().label("pages"),
-                ).group_by(SeriesPage.series_tag_id)
-            )
-        ).all()
-        page_count = {r.series_tag_id: r.pages for r in page_rows}
-
-        ch_rows = (
-            await self.session.execute(
-                select(
-                    SeriesChapter.series_tag_id,
-                    SeriesChapter.id,
-                    SeriesChapter.chapter_number,
-                    SeriesChapter.stated_page_start,
-                    SeriesChapter.stated_page_end,
-                    SeriesChapter.updated_at,
-                )
-                .order_by(
-                    SeriesChapter.series_tag_id,
-                    SeriesChapter.chapter_number.asc(),
-                )
-            )
-        ).all()
-        chapters_by_series: dict[int, list] = {}
-        updated_by_series: dict[int, object] = {}
-        for r in ch_rows:
-            chapters_by_series.setdefault(r.series_tag_id, []).append(r)
-            prev = updated_by_series.get(r.series_tag_id)
-            if prev is None or (r.updated_at and r.updated_at > prev):
-                updated_by_series[r.series_tag_id] = r.updated_at
-
-        # Cover = the (chapter_number, page_number)-minimum page per series.
-        cover_q = (
-            select(
-                SeriesPage.series_tag_id,
-                ImageRecord.sha256,
-                ImageRecord.mime,
-                ImageRecord.thumbnail_path,
-                ImageRecord.artist_id,
-                func.row_number()
-                .over(
-                    partition_by=SeriesPage.series_tag_id,
-                    order_by=(
-                        SeriesChapter.chapter_number.asc(),
-                        SeriesPage.page_number.asc(),
-                    ),
-                )
-                .label("rn"),
-            )
-            .join(SeriesChapter, SeriesChapter.id == SeriesPage.chapter_id)
-            .join(ImageRecord, ImageRecord.id == SeriesPage.image_id)
-            .subquery()
-        )
-        cover_rows = (
-            await self.session.execute(
-                select(
-                    cover_q.c.series_tag_id,
-                    cover_q.c.sha256,
-                    cover_q.c.mime,
-                    cover_q.c.thumbnail_path,
-                    cover_q.c.artist_id,
-                ).where(cover_q.c.rn == 1)
-            )
-        ).all()
-        cover_by_series = {r.series_tag_id: r for r in cover_rows}
-
-        # Artist names for the covers.
-        artist_ids = {
-            r.artist_id for r in cover_rows if r.artist_id is not None
-        }
-        artist_by_id: dict[int, object] = {}
-        if artist_ids:
-            arows = (
-                await self.session.execute(
-                    select(Artist.id, Artist.name, Artist.slug)
-                    .where(Artist.id.in_(artist_ids))
-                )
-            ).all()
-            artist_by_id = {a.id: a for a in arows}
-
-        tags = (
-            await self.session.execute(
-                select(Tag.id, Tag.name).where(Tag.kind == TagKind.series)
-            )
-        ).all()
-
-        out: list[dict] = []
-        for t in tags:
-            cover = cover_by_series.get(t.id)
-            if artist_id is not None and (
-                cover is None or cover.artist_id != artist_id
-            ):
-                continue
-            chs = chapters_by_series.get(t.id, [])
-            gap = bool(
-                self._gaps(
-                    [
-                        {
-                            "id": c.id,
-                            "stated_page_start": c.stated_page_start,
-                            "stated_page_end": c.stated_page_end,
-                        }
-                        for c in chs
-                    ]
-                )
-            )
-            artist = artist_by_id.get(cover.artist_id) if cover else None
-            out.append(
-                {
-                    "id": t.id,
-                    "name": t.name,
-                    "cover_thumbnail_url": (
-                        thumbnail_url(
-                            cover.thumbnail_path, cover.sha256, cover.mime
-                        )
-                        if cover
-                        else None
-                    ),
-                    "artist_name": artist.name if artist else None,
-                    "artist_slug": artist.slug if artist else None,
-                    "chapter_count": len(chs),
-                    "page_count": page_count.get(t.id, 0),
-                    "has_gap": gap,
-                    "updated_at": (
-                        updated_by_series.get(t.id).isoformat()
-                        if updated_by_series.get(t.id)
-                        else None
-                    ),
-                }
-            )
-
-        if sort == "name":
-            out.sort(key=lambda s: s["name"].lower())
-        elif sort == "size":
-            out.sort(key=lambda s: s["page_count"], reverse=True)
-        else:  # recent
-            out.sort(key=lambda s: s["updated_at"] or "", reverse=True)
-        return out
+        new_order = [image_id] + [i for i in order if i != image_id]
+        await self.reorder(series_tag_id, new_order)

backend/app/services/showcase_service.py

@@ -20,20 +20,11 @@ class ShowcaseService:
     async def random_sample(self, limit: int = 60) -> list[dict]:
         if limit < 1 or limit > 200:
             raise ValueError("limit must be between 1 and 200")
-        # Over-sample then random-order (#699): SYSTEM_ROWS reads CONTIGUOUS rows
-        # from each sampled page, so sequentially-imported near-duplicates
-        # (multi-image posts, variant sets) come back adjacent and cluster in the
-        # showcase ("three near-identical in a row"). Sampling a multiple of
-        # `limit` spans more pages, and ORDER BY random() before taking `limit`
-        # breaks the physical adjacency — far better spread, still cheap
-        # (random() over a few hundred rows, not the whole table).
-        oversample = min(limit * 5, 1000)
         stmt = select(ImageRecord).from_statement(
             text(
-                "SELECT * FROM ("
-                "  SELECT * FROM image_record TABLESAMPLE SYSTEM_ROWS(:o)"
-                ") sub ORDER BY random() LIMIT :n"
-            ).bindparams(o=oversample, n=limit)
+                "SELECT * FROM image_record "
+                "TABLESAMPLE SYSTEM_ROWS(:n)"
+            ).bindparams(n=limit)
         )
         rows = (await self.session.execute(stmt)).scalars().all()
         return [

backend/app/services/source_service.py

@@ -64,15 +64,6 @@ class SourceRecord:
     consecutive_failures: int
     next_check_at: str | None
     backfill_runs_remaining: int
-    # plan #693: derived from config_overrides for the UI badge.
-    backfill_state: str | None  # "running" | "complete" | "stalled" | None (idle)
-    backfill_chunks: int
-    # plan #697: a running deep-walk that bypasses the Patreon seen-ledger
-    # (recovery) vs. a normal backfill. Lets the badge label it "Recovering".
-    backfill_bypass_seen: bool
-    # plan #704: cumulative posts processed across the walk's chunks — live
-    # progress for the badge.
-    backfill_posts: int
 
     def to_dict(self) -> dict:
         return {
@@ -91,10 +82,6 @@ class SourceRecord:
             "consecutive_failures": self.consecutive_failures,
             "next_check_at": self.next_check_at,
             "backfill_runs_remaining": self.backfill_runs_remaining,
-            "backfill_state": self.backfill_state,
-            "backfill_chunks": self.backfill_chunks,
-            "backfill_bypass_seen": self.backfill_bypass_seen,
-            "backfill_posts": self.backfill_posts,
         }
 
 
@@ -102,13 +89,10 @@ class SourceRecord:
 
 _EDITABLE = {"enabled", "url", "config_overrides", "check_interval_override", "platform"}
 
-# Plan #693: backfill safety cap. "Start backfill" (and a newly created
-# enabled source) arms a run-until-done walk; this caps how many time-boxed
-# chunks it may spend before pausing as "stalled", so a pathological walk that
-# never reaches the bottom can't run forever. Generous on purpose — at
-# BACKFILL_CHUNK_SECONDS (600s) per chunk this is ~33h of cumulative walk, far
-# beyond any real catalog; the cursor stall-guard is the real terminator.
-BACKFILL_MAX_CHUNKS = 200
+# Plan #544 follow-up: newly created enabled sources pre-arm backfill so
+# their first N polls walk gallery-dl's full post history with the longer
+# timeout (matches the manual "Deep scan" button's default).
+NEW_SOURCE_BACKFILL_RUNS = 3
 
 
 class SourceService:
@@ -151,7 +135,6 @@ class SourceService:
         self, source: Source, artist: Artist, settings: ImportSettings,
     ) -> SourceRecord:
         nxt = compute_next_check_at(source, artist, settings)
-        co = source.config_overrides or {}
         return SourceRecord(
             id=source.id,
             artist_id=source.artist_id,
@@ -168,10 +151,6 @@ class SourceService:
             consecutive_failures=source.consecutive_failures or 0,
             next_check_at=nxt.isoformat() if nxt else None,
             backfill_runs_remaining=source.backfill_runs_remaining or 0,
-            backfill_state=co.get("_backfill_state"),
-            backfill_chunks=int(co.get("_backfill_chunks", 0)),
-            backfill_bypass_seen=bool(co.get("_backfill_bypass_seen")),
-            backfill_posts=int(co.get("_backfill_posts", 0)),
         )
 
     async def _row_to_record(self, source: Source) -> SourceRecord:
@@ -233,17 +212,16 @@ class SourceService:
             select(func.count(Source.id)).where(Source.artist_id == artist_id)
         )).scalar_one()
 
-        # Plan #693: a freshly added subscription has no archive yet, so it
-        # should walk its full post history once. Arm run-until-done backfill
-        # (state="running" + the chunk cap); the time-boxed chunks march to the
-        # bottom across ticks, then flip to "complete" and tick mode takes over.
-        # Disabled sources (incl. sidecar synthetics, url='sidecar:...') are
-        # never polled, so leave them idle.
-        if enabled:
-            config_overrides = {**(config_overrides or {}), "_backfill_state": "running"}
-            backfill_runs = BACKFILL_MAX_CHUNKS
-        else:
-            backfill_runs = 0
+        # Plan #544 follow-up: a freshly added subscription has no archive
+        # yet, so the first few polls would walk the full post history in
+        # tick mode and trip exit:20 after ~20 contiguous archive hits —
+        # except there are none yet, so tick mode would walk forever and
+        # blow the wall-clock cap. Pre-arm backfill so the initial syncs
+        # use the longer timeout + skip:True walk. Tick mode resumes once
+        # the budget is spent or the queue drains.
+        # Disabled sources (incl. sidecar synthetics, url='sidecar:...')
+        # are never polled, so leave their counter at 0.
+        backfill_runs = NEW_SOURCE_BACKFILL_RUNS if enabled else 0
         source = Source(
             artist_id=artist_id, platform=platform, url=url,
             enabled=enabled, config_overrides=config_overrides,
@@ -308,71 +286,22 @@ class SourceService:
         await self.session.commit()
         return await self._row_to_record(source)
 
-    async def start_backfill(self, source_id: int) -> SourceRecord:
-        """Plan #693: arm a run-until-done backfill. Sets state="running" and
-        the chunk cap; download runs then walk the full post history in
-        time-boxed chunks (skip:True + BACKFILL_CHUNK_SECONDS), resuming from
-        the cursor each chunk, until gallery-dl reaches the bottom (→ state
-        "complete") or the cap/stall-guard pauses it (→ "stalled"). Clears any
-        prior cursor/chunk/stall state so a re-start walks fresh from the top."""
+    async def set_backfill_runs(
+        self, source_id: int, runs: int,
+    ) -> SourceRecord:
+        """Plan #544: arm a source for backfill mode. The next `runs`
+        download runs will use gallery-dl's full-walk config (skip: True
+        + 30-min timeout) instead of the catch-up default. Runs must be
+        1..10 — bigger is rejected to keep the operator from accidentally
+        setting a runaway budget."""
+        if not isinstance(runs, int) or runs < 1 or runs > 10:
+            raise ValueError("runs must be an integer in [1, 10]")
         source = (await self.session.execute(
             select(Source).where(Source.id == source_id)
         )).scalar_one_or_none()
         if source is None:
             raise LookupError(f"source id={source_id} not found")
-        co = dict(source.config_overrides or {})
-        co["_backfill_state"] = "running"
-        for k in ("_backfill_cursor", "_backfill_cursor_stalls", "_backfill_chunks",
-                  "_backfill_posts"):
-            co.pop(k, None)
-        source.config_overrides = co
-        source.backfill_runs_remaining = BACKFILL_MAX_CHUNKS
-        await self.session.commit()
-        return await self._row_to_record(source)
-
-    async def start_recovery(self, source_id: int) -> SourceRecord:
-        """Plan #697: arm a RECOVERY walk — a backfill that bypasses the Patreon
-        seen-ledger so deliberately-dropped-and-deleted near-dups get re-fetched
-        and re-evaluated under the CURRENT pHash threshold (tier-2 disk still
-        spares files we kept). Reuses the entire #693 backfill state machine
-        (time-boxed chunks, cursor checkpoint, complete/stall lifecycle) plus the
-        `_backfill_bypass_seen` flag that flips download mode to recovery. Clears
-        any prior cursor/chunk/stall state so it walks fresh from the top. The
-        flag is cleared on completion (download_service) and on stop.
-
-        Recovery is Patreon-only (the seen-ledger is Patreon's); for other
-        platforms the flag is inert (download_service ignores it) and the walk
-        runs as a plain backfill. The UI gates the action to Patreon sources."""
-        source = (await self.session.execute(
-            select(Source).where(Source.id == source_id)
-        )).scalar_one_or_none()
-        if source is None:
-            raise LookupError(f"source id={source_id} not found")
-        co = dict(source.config_overrides or {})
-        co["_backfill_state"] = "running"
-        co["_backfill_bypass_seen"] = True
-        for k in ("_backfill_cursor", "_backfill_cursor_stalls", "_backfill_chunks",
-                  "_backfill_posts"):
-            co.pop(k, None)
-        source.config_overrides = co
-        source.backfill_runs_remaining = BACKFILL_MAX_CHUNKS
-        await self.session.commit()
-        return await self._row_to_record(source)
-
-    async def stop_backfill(self, source_id: int) -> SourceRecord:
-        """Plan #693: cancel an in-progress backfill — back to idle/tick mode.
-        Clears the running state + cursor/chunk/stall bookkeeping."""
-        source = (await self.session.execute(
-            select(Source).where(Source.id == source_id)
-        )).scalar_one_or_none()
-        if source is None:
-            raise LookupError(f"source id={source_id} not found")
-        co = dict(source.config_overrides or {})
-        for k in ("_backfill_state", "_backfill_cursor", "_backfill_cursor_stalls",
-                  "_backfill_chunks", "_backfill_bypass_seen", "_backfill_posts"):
-            co.pop(k, None)
-        source.config_overrides = co
-        source.backfill_runs_remaining = 0
+        source.backfill_runs_remaining = runs
         await self.session.commit()
         return await self._row_to_record(source)
 

backend/app/services/tag_service.py

@@ -1,7 +1,5 @@
 """Tag CRUD + autocomplete + image-tag association."""
 
-import logging
-import time
 from collections.abc import Sequence
 from dataclasses import dataclass
 
@@ -14,25 +12,6 @@ from ..models import Tag, TagKind, image_tag
 from ..models.tag_allowlist import TagAllowlist
 from ..models.tag_reference_embedding import TagReferenceEmbedding
 
-log = logging.getLogger(__name__)
-
-
-def normalize_tag_name(name: str) -> str:
-    """Canonical tag form (#701): collapse whitespace + capitalize the first
-    letter of each word, PRESERVING the rest of the word.
-
-    Per-word capitalize (NOT str.title(), which mangles apostrophes:
-    don't → Don'T). The word tail is left untouched so acronyms survive
-    (DC stays DC, NSFW stays NSFW) — operator-revised 2026-06-06: protecting
-    acronyms matters more than folding ALL-CAPS input. This MATCHES
-    ml/tag_name._title_word, so a tag suggested by the Camie tagger keeps the
-    exact casing the suggestion UI showed when it round-trips through the
-    create endpoint on Accept.
-    """
-    return " ".join(
-        w[:1].upper() + w[1:] for w in (name or "").split()
-    )
-
 
 class TagValidationError(ValueError):
     """Raised when tag construction breaks the kind/fandom rules."""
@@ -92,10 +71,6 @@ class TagService:
         - if fandom_id is set, the referenced tag must exist and have
           kind == TagKind.fandom
         """
-        # NOTE: case is NOT normalized here — find_or_create is the shared path
-        # the ML tagger / allowlist also use, and Title-Casing the booru
-        # vocabulary breaks allowlist matching. Display-casing (Title Case) is
-        # applied at the user-entry layer (api/tags create_tag) only (#701).
         name = name.strip()
         if not name:
             raise TagValidationError("Tag name cannot be empty")
@@ -118,17 +93,13 @@ class TagService:
         # inserts; without the savepoint the outer transaction would
         # poison and the calling request crashes. Mirrors
         # importer._get_or_create.
-        # Case-insensitive match (#701): a normalized (Title Case) input must
-        # find an existing differently-cased tag instead of forking a duplicate.
         stmt = (
             select(Tag)
-            .where(func.lower(Tag.name) == name.lower())
+            .where(Tag.name == name)
             .where(Tag.kind == kind)
             .where(
                 Tag.fandom_id.is_(None) if fandom_id is None else Tag.fandom_id == fandom_id
             )
-            .order_by(Tag.id)
-            .limit(1)
         )
         existing = (await self.session.execute(stmt)).scalar_one_or_none()
         if existing:
@@ -278,11 +249,9 @@ class TagService:
         if tag is None:
             raise TagValidationError(f"Tag {tag_id} not found")
 
-        # Case-insensitive clash (#701) — renaming onto a differently-cased tag
-        # is still a merge, not a silent fork.
         clash_stmt = (
             select(Tag)
-            .where(func.lower(Tag.name) == new_name.lower())
+            .where(Tag.name == new_name)
             .where(Tag.kind == tag.kind)
             .where(
                 Tag.fandom_id.is_(None)
@@ -290,8 +259,6 @@ class TagService:
                 else Tag.fandom_id == tag.fandom_id
             )
             .where(Tag.id != tag_id)
-            .order_by(Tag.id)
-            .limit(1)
         )
         clash = (await self.session.execute(clash_stmt)).scalar_one_or_none()
         if clash is not None:
@@ -515,18 +482,8 @@ class TagService:
         )
 
     async def _repoint_series_pages(self, src: int, tgt: int) -> None:
-        from ..models.series_chapter import SeriesChapter
         from ..models.series_page import SeriesPage
 
-        # Move the chapters first so the pages' chapter_id FK stays valid: a
-        # chapter left pointing at src would cascade-delete (with its pages) when
-        # src is removed. chapter_number may now collide across the merged set —
-        # acceptable (it's an ordering key, not unique).
-        await self.session.execute(
-            update(SeriesChapter)
-            .where(SeriesChapter.series_tag_id == src)
-            .values(series_tag_id=tgt)
-        )
         # image_id is UNIQUE across series_page and src != tgt, so an
         # image in src's series cannot already be in tgt's — no collision.
         await self.session.execute(
@@ -544,21 +501,6 @@ class TagService:
             update(Tag).where(Tag.fandom_id == src).values(fandom_id=tgt)
         )
 
-    async def _image_assoc_counts(self, tag_ids: list[int]) -> dict[int, int]:
-        """image_tag row counts keyed by tag_id, for the survivor heuristic
-        (the best-connected tag in a collision group survives → fewest
-        FK repoints). Tags with zero associations are absent from the map."""
-        if not tag_ids:
-            return {}
-        rows = (
-            await self.session.execute(
-                select(image_tag.c.tag_id, func.count())
-                .where(image_tag.c.tag_id.in_(tag_ids))
-                .group_by(image_tag.c.tag_id)
-            )
-        ).all()
-        return {tid: int(n) for tid, n in rows}
-
     async def _create_protective_aliases(
         self, src_name: str, src_kind: TagKind, tgt: int
     ) -> bool:
@@ -606,189 +548,3 @@ class TagService:
             if res.rowcount:
                 created = True
         return created
-
-
-# ---------------------------------------------------------------------------
-# #714: retro-normalize existing tags to the #701 canonical (Title Case +
-# collapsed whitespace) and merge case/whitespace-variant duplicates.
-# ---------------------------------------------------------------------------
-
-_NORMALIZE_SAMPLE_CAP = 50
-
-
-def _group_existing_tags(
-    rows,
-) -> dict[tuple, list[tuple[int, str]]]:
-    """Group (id, name, kind, fandom_id) rows by their post-normalization
-    identity: (kind, COALESCE(fandom_id, -1), canonical_name). Every tag that
-    would collapse to the same canonical lives in ONE group, so the canonical
-    form is unique within (kind, fandom) once each group is resolved."""
-    groups: dict[tuple, list[tuple[int, str]]] = {}
-    for tag_id, name, kind, fandom_id in rows:
-        canonical = normalize_tag_name(name)
-        key = (kind, fandom_id if fandom_id is not None else -1, canonical)
-        groups.setdefault(key, []).append((tag_id, name))
-    return groups
-
-
-def _group_needs_change(canonical: str, members: list[tuple[int, str]]) -> bool:
-    """A group is already canonical iff it's a single member whose name equals
-    the canonical form. Anything else (a collision, or a lone mis-cased tag)
-    needs work."""
-    if len(members) > 1:
-        return True
-    return members[0][1] != canonical
-
-
-def _best_connected(tag_ids: list[int], counts: dict[int, int]) -> int:
-    """The tag with the most image associations (→ fewest FK repoints when it
-    survives), tie-broken to the lowest id for determinism. Module-level so the
-    key closes over its parameter, not a loop variable (ruff B023)."""
-    return max(tag_ids, key=lambda tid: (counts.get(tid, 0), -tid))
-
-
-async def normalize_existing_tags(
-    session: AsyncSession,
-    *,
-    dry_run: bool = False,
-    time_budget_seconds: float | None = None,
-) -> dict:
-    """Convert the back-catalog to the #701 canonical tag form.
-
-    For each (kind, fandom, canonical) group: pick a survivor, merge any
-    case/whitespace-variant siblings INTO it via the tested merge path
-    (TagService._do_merge — FK repoints + protective aliases), then rename the
-    survivor to the canonical form. Idempotent: a group that is already a lone
-    canonical tag is a no-op, so re-running is safe.
-
-    A first run over a fresh back-catalog can touch tens of thousands of tags
-    (the whole booru-derived vocabulary needs recasing) and won't finish inside
-    one Celery time limit — it timed out at 40 min (operator-flagged 2026-06-07).
-    `time_budget_seconds` time-boxes the live run: it stops cleanly at the budget
-    and reports `partial`/`remaining` so the caller can re-enqueue and continue.
-    Because it commits per group and is idempotent, the next run just picks up
-    the groups still needing change.
-
-    dry_run=True returns a projection (counts + a sample of the changes) with no
-    mutations. Live runs commit per group and isolate failures per group so one
-    bad group can't strand the rest.
-
-    Returns (dry_run):
-      {"groups": N, "collisions": M, "tags_to_merge": K, "tags_to_rename": R,
-       "total_changes": T, "sample": [{"to", "from": [...], "kind", "merge"}]}
-    Returns (live):
-      {"groups_processed", "merged", "renamed", "aliases_created", "errors",
-       "total_changes", "remaining", "partial", "sample": [...]}
-    """
-    rows = (
-        await session.execute(
-            select(Tag.id, Tag.name, Tag.kind, Tag.fandom_id)
-        )
-    ).all()
-    groups = _group_existing_tags(rows)
-
-    # Deterministic sample/ordering: by kind then canonical name.
-    touched = sorted(
-        (
-            (key, members)
-            for key, members in groups.items()
-            if _group_needs_change(key[2], members)
-        ),
-        key=lambda km: (
-            km[0][0].value if hasattr(km[0][0], "value") else str(km[0][0]),
-            km[0][2].lower(),
-        ),
-    )
-
-    sample = [
-        {
-            "to": key[2],
-            "from": [name for _id, name in members],
-            "kind": key[0].value if hasattr(key[0], "value") else str(key[0]),
-            "merge": len(members) > 1,
-        }
-        for key, members in touched[:_NORMALIZE_SAMPLE_CAP]
-    ]
-
-    if dry_run:
-        collisions = sum(1 for key, m in touched if len(m) > 1)
-        tags_to_merge = sum(len(m) - 1 for key, m in touched if len(m) > 1)
-        # A group renames iff the canonical form isn't already one of its
-        # members' exact names (else that member is picked as survivor → no
-        # rename, the rest merge into it).
-        tags_to_rename = sum(
-            1
-            for key, m in touched
-            if key[2] not in {name for _id, name in m}
-        )
-        return {
-            "groups": len(groups),
-            "collisions": collisions,
-            "tags_to_merge": tags_to_merge,
-            "tags_to_rename": tags_to_rename,
-            "total_changes": len(touched),
-            "sample": sample,
-        }
-
-    svc = TagService(session)
-    summary = {
-        "groups_processed": 0,
-        "merged": 0,
-        "renamed": 0,
-        "aliases_created": 0,
-        "errors": 0,
-        "total_changes": len(touched),
-        "remaining": len(touched),
-        "partial": False,
-        "sample": sample,
-    }
-    start = time.monotonic()
-    for done, (key, members) in enumerate(touched):
-        # Time-box: stop cleanly before the Celery limit kills us mid-group and
-        # strands the run as a timeout. The caller re-enqueues to finish the
-        # rest (idempotent — already-canonical groups are skipped next pass).
-        if (
-            time_budget_seconds is not None
-            and time.monotonic() - start >= time_budget_seconds
-        ):
-            summary["partial"] = True
-            summary["remaining"] = len(touched) - done
-            break
-        canonical = key[2]
-        names_by_id = dict(members)
-        # Survivor: prefer a member already named canonically (no rename, no
-        # self-alias); else the best-connected (fewest FK repoints); else
-        # lowest id for determinism.
-        survivor_id = next(
-            (tid for tid, name in members if name == canonical), None
-        )
-        if survivor_id is None:
-            counts = await svc._image_assoc_counts(list(names_by_id))
-            survivor_id = _best_connected(list(names_by_id), counts)
-        loser_ids = [tid for tid in names_by_id if tid != survivor_id]
-        try:
-            survivor = await session.get(Tag, survivor_id)
-            for loser_id in loser_ids:
-                loser = await session.get(Tag, loser_id)
-                if loser is None:
-                    continue
-                result = await svc._do_merge(loser, survivor)
-                if result.alias_created:
-                    summary["aliases_created"] += 1
-            if survivor.name != canonical:
-                survivor.name = canonical
-                await session.flush()
-                summary["renamed"] += 1
-            await session.commit()
-            summary["groups_processed"] += 1
-            summary["merged"] += len(loser_ids)
-        except Exception as exc:  # one bad group must not strand the rest
-            await session.rollback()
-            summary["errors"] += 1
-            log.warning(
-                "tag normalize failed for group %r: %s", canonical, exc
-            )
-    else:
-        # Loop finished without hitting the time budget — nothing left to do.
-        summary["remaining"] = 0
-    return summary

backend/app/tasks/admin.py

@@ -55,146 +55,3 @@ def bulk_delete_images_task(self, *, image_ids: list[int]) -> dict:
         return cleanup_service.delete_images(
             session, image_ids=image_ids, images_root=IMAGES_ROOT,
         )
-
-
-# Time-box one chunk well under the soft limit so a large archive back-catalog
-# can't run the task into the Celery time limit (or hog the maintenance_long
-# lane). The task re-enqueues itself with the resume cursor until the scan is
-# exhausted — mirrors normalize_tags_task (operator-asked 2026-06-07: reasonable
-# timeout, then re-queue so other work keeps flowing).
-_REEXTRACT_CHUNK_SECONDS = 600
-
-
-@celery.task(
-    name="backend.app.tasks.admin.reextract_archive_attachments_task",
-    bind=True,
-    autoretry_for=(OperationalError, DBAPIError),
-    retry_backoff=15, retry_backoff_max=180, max_retries=1,
-    soft_time_limit=1800, time_limit=2400,  # 30 min / 40 min
-)
-def reextract_archive_attachments_task(self, after_id: int = 0) -> dict:
-    """Wraps cleanup_service.reextract_archive_attachments (#713 part 2):
-    re-extract PostAttachments that are actually archives but were filed
-    opaquely before the magic-byte gate, and link their members to the post.
-
-    Time-boxed + self-resuming: scans attachments after ``after_id`` and, on a
-    chunk cut, re-enqueues from where it stopped so a big backlog finishes across
-    chunks instead of dying at the soft limit."""
-    SessionLocal = _sync_session_factory()
-    with SessionLocal() as session:
-        summary = cleanup_service.reextract_archive_attachments(
-            session, images_root=IMAGES_ROOT,
-            time_budget_seconds=_REEXTRACT_CHUNK_SECONDS, after_id=after_id,
-        )
-    # More attachments past this chunk's cursor — continue in the next.
-    if summary.get("partial") and summary.get("resume_after_id", 0) > after_id:
-        log.info(
-            "reextract chunk done (%d scanned, %d archives, resume after id %s) "
-            "— re-enqueuing to continue",
-            summary.get("scanned", 0), summary.get("archives", 0),
-            summary["resume_after_id"],
-        )
-        reextract_archive_attachments_task.delay(summary["resume_after_id"])
-    return summary
-
-
-# Time-box one chunk well under the soft limit so a large back-catalog (the
-# first run recases the whole booru vocabulary) can't run the task into the
-# Celery time limit — it timed out at 40 min, operator-flagged 2026-06-07. The
-# task re-enqueues itself until nothing remains (idempotent — already-canonical
-# groups are skipped). 600s keeps each chunk short enough that the recovery
-# sweep and other maintenance tasks interleave on the concurrency-1 queue.
-_NORMALIZE_CHUNK_SECONDS = 600
-
-
-@celery.task(
-    name="backend.app.tasks.admin.normalize_tags_task",
-    bind=True,
-    autoretry_for=(OperationalError, DBAPIError),
-    retry_backoff=15, retry_backoff_max=180, max_retries=1,
-    soft_time_limit=1800, time_limit=2400,  # 30 min / 40 min
-)
-def normalize_tags_task(self) -> dict:
-    """Wraps tag_service.normalize_existing_tags (#714): Title-Case the
-    back-catalog and merge case/whitespace-variant duplicate tags via the
-    tested async merge path. Time-boxed + self-resuming so a huge first run
-    finishes across chunks instead of timing out. Runs under its own asyncio
-    loop + per-task async engine (NullPool), mirroring download_source."""
-    import asyncio
-
-    from ..services.tag_service import normalize_existing_tags
-    from ._async_session import async_session_factory
-
-    async def _run() -> dict:
-        async_factory, async_engine = async_session_factory()
-        try:
-            async with async_factory() as session:
-                # normalize_existing_tags commits per group internally.
-                return await normalize_existing_tags(
-                    session, dry_run=False,
-                    time_budget_seconds=_NORMALIZE_CHUNK_SECONDS,
-                )
-        finally:
-            await async_engine.dispose()
-
-    summary = asyncio.run(_run())
-    # More groups to canonicalize than fit this chunk — continue in the next.
-    if summary.get("partial") and summary.get("remaining", 0) > 0:
-        log.info(
-            "normalize_tags_task chunk done (%d processed, %d remaining) — "
-            "re-enqueuing to continue",
-            summary.get("groups_processed", 0), summary["remaining"],
-        )
-        normalize_tags_task.delay()
-    return summary
-
-
-# Time-box one rescan chunk well under the soft limit and re-enqueue from the
-# cursor — scoring every post against its artist's series is O(posts) and grows
-# with the library (FC-6.3). Mirrors normalize_tags_task.
-_SERIES_RESCAN_CHUNK_SECONDS = 600
-
-
-@celery.task(
-    name="backend.app.tasks.admin.rescan_series_suggestions_task",
-    bind=True,
-    autoretry_for=(OperationalError, DBAPIError),
-    retry_backoff=15, retry_backoff_max=180, max_retries=1,
-    soft_time_limit=1800, time_limit=2400,  # 30 min / 40 min
-)
-def rescan_series_suggestions_task(self, after_post_id: int = 0) -> dict:
-    """Score posts against their artist's series and write pending suggestions
-    (FC-6.3). Settings-gated; time-boxed + self-resuming from a post-id cursor.
-    Per-task async engine (NullPool) under its own asyncio loop, like normalize."""
-    import asyncio
-
-    from ..models import ImportSettings
-    from ..services.series_match_service import SeriesMatchService
-    from ._async_session import async_session_factory
-
-    async def _run() -> dict:
-        async_factory, async_engine = async_session_factory()
-        try:
-            async with async_factory() as session:
-                settings = await ImportSettings.load(session)
-                if not settings.series_suggest_enabled:
-                    return {"skipped": "series suggestions disabled"}
-                threshold = settings.series_suggest_threshold
-                return await SeriesMatchService(session).rescan(
-                    threshold=threshold,
-                    time_budget_seconds=_SERIES_RESCAN_CHUNK_SECONDS,
-                    after_post_id=after_post_id,
-                )
-        finally:
-            await async_engine.dispose()
-
-    summary = asyncio.run(_run())
-    if summary.get("partial") and summary.get("resume_after_id", 0) > after_post_id:
-        log.info(
-            "rescan_series_suggestions chunk done (%d scanned, %d suggested, "
-            "resume after %s) — re-enqueuing",
-            summary.get("scanned", 0), summary.get("suggested", 0),
-            summary["resume_after_id"],
-        )
-        rescan_series_suggestions_task.delay(summary["resume_after_id"])
-    return summary

backend/app/tasks/backup.py

@@ -41,11 +41,7 @@ def _mark_failed(session, row: BackupRun, exc: BaseException) -> None:
     bind=True,
     autoretry_for=(OperationalError, DBAPIError),
     retry_backoff=10, retry_backoff_max=120, max_retries=2,
-    # A pg_dump can't be chunked; the 12-min limit timed out once the DB grew
-    # (operator-flagged 2026-06-07). 30/35 min gives real headroom. (A long
-    # backup still briefly holds the concurrency-1 maintenance lane — the
-    # structural fix is a dedicated lane for the long one-shots.)
-    soft_time_limit=1800, time_limit=2100,
+    soft_time_limit=600, time_limit=720,
 )
 def backup_db_task(self, *, tag: str | None = None,
                    triggered_by: str = "manual") -> dict:

backend/app/tasks/download.py

@@ -31,7 +31,7 @@ _KEY_PATH = IMAGES_ROOT / "secrets" / "credential_key.b64"
 # SoftTimeLimitExceeded in-process, whereas the HARD limit SIGKILLs the
 # worker (no chance to finalize). Both gallery-dl subprocess budgets
 # (gallery_dl.py: _DEFAULT_GDL_TIMEOUT_SECONDS=870 tick,
-# BACKFILL_CHUNK_SECONDS=600 per backfill chunk, plan #693) MUST sit below the soft limit
+# BACKFILL_TIMEOUT_SECONDS=1170 backfill) MUST sit below the soft limit
 # so subprocess.run raises its own TimeoutExpired first — that path
 # captures partial stdout/stderr and finalizes the DownloadEvent. soft is
 # max-subprocess (1170) + ~180s phase-3 persist headroom; hard is soft +
@@ -44,25 +44,6 @@ _KEY_PATH = IMAGES_ROOT / "secrets" / "credential_key.b64"
 DOWNLOAD_SOFT_TIME_LIMIT = 1350
 DOWNLOAD_HARD_TIME_LIMIT = 1500
 
-# Per-platform serialization (plan: concurrency cap). When a serialized
-# platform (Patreon) is already walking, defer this run by re-enqueuing it a
-# little later rather than holding a worker slot or bowling into the same rate
-# limit. Bounded so a wedged platform eventually runs anyway. The lock TTL sits
-# just past the hard kill so a SIGKILL'd worker's lock auto-expires; a backfill
-# chunk only holds it ~10 min, so the bounded wait stays well under the 30-min
-# DownloadEvent recovery sweep.
-_PLATFORM_LOCK_TTL = DOWNLOAD_HARD_TIME_LIMIT + 120
-_SERIALIZE_COUNTDOWN = 20
-_MAX_SERIALIZE_WAITS = 45  # ~15 min ceiling, then run uncapped as a safety valve
-
-
-def _peek_platform(source_id: int) -> str | None:
-    SyncFactory = _sync_session_factory()
-    with SyncFactory() as session:
-        return session.execute(
-            select(Source.platform).where(Source.id == source_id)
-        ).scalar_one_or_none()
-
 
 def _finalize_soft_limited(session: SyncSession, source_id: int) -> None:
     """Defense in depth for the soft-time-limit kill path.
@@ -94,11 +75,10 @@ def _finalize_soft_limited(session: SyncSession, source_id: int) -> None:
     ev.finished_at = now
     ev.error = (
         f"killed by Celery soft time limit ({DOWNLOAD_SOFT_TIME_LIMIT}s) "
-        "before the download finished — the run exceeded its time budget. "
-        "Progress is checkpointed per page, so the next tick resumes near "
-        "the cut; the backfill budget was decremented so it walks less. If "
-        "this recurs, a single post is heavy enough to overrun the chunk "
-        "time-box on its own."
+        "before the gallery-dl subprocess returned — the run exceeded its "
+        "budget and its stdout/stderr were lost with the worker thread. "
+        "If this recurs, the source is too large for one run; the backfill "
+        "budget was decremented so the next tick walks less."
     )
     ev.metadata_ = {
         **(ev.metadata_ or {}),
@@ -127,48 +107,9 @@ def _finalize_soft_limited(session: SyncSession, source_id: int) -> None:
     soft_time_limit=DOWNLOAD_SOFT_TIME_LIMIT,
     time_limit=DOWNLOAD_HARD_TIME_LIMIT,
 )
-def download_source(self, source_id: int, _serialize_waits: int = 0) -> int:
+def download_source(self, source_id: int) -> int:
     """Returns the DownloadEvent.id."""
 
-    # Per-platform concurrency cap: only one Patreon walk runs at a time.
-    from ..services.platform_lock import platform_lock
-
-    platform = _peek_platform(source_id)
-    lock = (
-        platform_lock(platform, ttl_seconds=_PLATFORM_LOCK_TTL)
-        if platform is not None
-        else None
-    )
-    if lock is not None:
-        try:
-            got_lock = bool(lock.acquire(blocking=False))
-        except Exception:  # noqa: BLE001 — broker hiccup → degrade to uncapped
-            log.warning("platform lock acquire failed for %s; running uncapped", platform)
-            lock = None
-            got_lock = True
-        if lock is not None and not got_lock:
-            if _serialize_waits < _MAX_SERIALIZE_WAITS:
-                # Another walk on this platform holds the lock — re-enqueue
-                # ourselves shortly (the pending DownloadEvent stays; no new
-                # event, no log spam) rather than running concurrently into
-                # the rate limit.
-                download_source.apply_async(
-                    (source_id,),
-                    {"_serialize_waits": _serialize_waits + 1},
-                    countdown=_SERIALIZE_COUNTDOWN,
-                )
-                log.info(
-                    "download_source(%s) deferred — %s already walking (wait %d/%d)",
-                    source_id, platform, _serialize_waits + 1, _MAX_SERIALIZE_WAITS,
-                )
-                return -1
-            log.warning(
-                "download_source(%s) running WITHOUT the %s lock — max serialize "
-                "waits hit, proceeding uncapped",
-                source_id, platform,
-            )
-            lock = None
-
     async def _run():
         async_factory, async_engine = async_session_factory()
         SyncFactory = _sync_session_factory()
@@ -202,11 +143,6 @@ def download_source(self, source_id: int, _serialize_waits: int = 0) -> int:
                         gdl=gdl,
                         importer=importer,
                         cred_service=cred_service,
-                        # The native Patreon ingester opens its own short-lived
-                        # sync sessions for the seen-ledger (never held across
-                        # the walk). Same factory the importer's sync session
-                        # comes from — a different DB connection per checkout.
-                        sync_session_factory=SyncFactory,
                     )
                     return await svc.download_source(source_id)
         finally:
@@ -228,12 +164,3 @@ def download_source(self, source_id: int, _serialize_waits: int = 0) -> int:
         except Exception:  # noqa: BLE001 — cleanup must not swallow the kill
             log.exception("soft-limit finalize failed for source %s", source_id)
         raise
-    finally:
-        # Release the per-platform lock so the next walk can proceed. The TTL is
-        # a backstop for a SIGKILL; here we free it the instant the run ends. A
-        # late release after TTL expiry raises (lock already gone) — ignore it.
-        if lock is not None:
-            try:
-                lock.release()
-            except Exception:  # noqa: BLE001 — TTL may have already freed it
-                pass

backend/app/tasks/library_audit.py

@@ -13,7 +13,6 @@ State machine:
 """
 
 import logging
-import time
 import traceback
 from datetime import UTC, datetime
 
@@ -32,12 +31,6 @@ log = logging.getLogger(__name__)
 _BATCH = 500
 _PROGRESS_TICK = 100
 _MAX_MATCHED = 50_000
-# One chunk's wall-clock budget. Was a single 2h pass that timed out on large
-# libraries and held the concurrency-1 maintenance queue the whole time
-# (operator-flagged 2026-06-07). Now: scan ~10 min, persist the keyset cursor +
-# matches, re-enqueue to continue — so backups/vacuum/normalize chunks can
-# interleave. soft/hard limits sit just above so the budget fires first.
-_CHUNK_SECONDS = 600
 
 _RULES = {
     "transparency": transparency.evaluate,
@@ -53,16 +46,13 @@ _RULES = {
     retry_backoff_max=60,
     retry_jitter=True,
     max_retries=3,
-    soft_time_limit=900,
-    time_limit=1000,
+    soft_time_limit=7200,
+    time_limit=7500,
 )
 def scan_library_for_rule(self, audit_id: int) -> dict:
-    """See module docstring. Time-boxed + self-resuming: one call scans a
-    ~10-min chunk, persists the resume cursor + matches, and re-enqueues itself
-    until the library is exhausted. Returns a small summary dict for eager-mode
+    """See module docstring. Returns a small summary dict for eager-mode
     test assertions (real workers ignore the return value)."""
     SessionLocal = _sync_session_factory()
-    start = time.monotonic()
     try:
         with SessionLocal() as session:
             audit = session.get(LibraryAuditRun, audit_id)
@@ -73,10 +63,9 @@ def scan_library_for_rule(self, audit_id: int) -> dict:
                 _mark_error(session, audit_id, f"unknown rule {audit.rule!r}")
                 return {"audit_id": audit_id, "status": "error"}
             params = dict(audit.params or {})
-            # Resume from the previous chunk's persisted state.
-            matched: list[int] = list(audit.matched_ids or [])
-            scanned = audit.scanned_count or 0
-            last_id = audit.resume_after_id or 0
+            matched: list[int] = []
+            scanned = 0
+            last_id = 0
             while True:
                 # Cancellation check between batches.
                 current_status = session.execute(
@@ -85,15 +74,6 @@ def scan_library_for_rule(self, audit_id: int) -> dict:
                 ).scalar_one()
                 if current_status == "cancelled":
                     return {"audit_id": audit_id, "status": "cancelled"}
-                # Time-box: persist the cursor + matches and re-enqueue so the
-                # queue is freed between chunks. The next call resumes here.
-                if time.monotonic() - start >= _CHUNK_SECONDS:
-                    _persist_chunk(session, audit_id, scanned, matched, last_id)
-                    scan_library_for_rule.delay(audit_id)
-                    return {
-                        "audit_id": audit_id, "status": "running",
-                        "partial": True, "scanned": scanned,
-                    }
                 rows = session.execute(
                     select(ImageRecord.id, ImageRecord.path)
                     .where(ImageRecord.id > last_id)
@@ -134,16 +114,10 @@ def scan_library_for_rule(self, audit_id: int) -> dict:
                         )
                         return {"audit_id": audit_id, "status": "error"}
                     if scanned % _PROGRESS_TICK == 0:
-                        # Cheap heartbeat: scanned_count + last_progress_at so the
-                        # recovery sweep sees the multi-chunk audit is alive. The
-                        # cursor + matches are persisted at chunk boundaries.
                         session.execute(
                             update(LibraryAuditRun)
                             .where(LibraryAuditRun.id == audit_id)
-                            .values(
-                                scanned_count=scanned,
-                                last_progress_at=datetime.now(UTC),
-                            )
+                            .values(scanned_count=scanned)
                         )
                         session.commit()
             # Final state.
@@ -154,10 +128,8 @@ def scan_library_for_rule(self, audit_id: int) -> dict:
                     scanned_count=scanned,
                     matched_count=len(matched),
                     matched_ids=matched,
-                    resume_after_id=last_id,
                     status="ready",
                     finished_at=datetime.now(UTC),
-                    last_progress_at=datetime.now(UTC),
                 )
             )
             session.commit()
@@ -168,14 +140,9 @@ def scan_library_for_rule(self, audit_id: int) -> dict:
                 "matched": len(matched),
             }
     except SoftTimeLimitExceeded:
-        # Backstop (the in-chunk budget should fire first): the audit stays
-        # 'running' with its last committed cursor; re-enqueue to continue from
-        # there rather than marking the whole run an error.
-        log.warning(
-            "audit %s: soft time limit hit — re-enqueuing to resume", audit_id,
-        )
-        scan_library_for_rule.delay(audit_id)
-        return {"audit_id": audit_id, "status": "running", "partial": True}
+        with SessionLocal() as session:
+            _mark_error(session, audit_id, "soft_time_limit exceeded (>7200s)")
+        raise
     except (OperationalError, DBAPIError):
         # Retryable per the decorator; leave row in 'running' and let
         # autoretry try again. Recovery sweep catches if all retries fail.
@@ -187,23 +154,6 @@ def scan_library_for_rule(self, audit_id: int) -> dict:
         raise
 
 
-def _persist_chunk(session, audit_id, scanned, matched, last_id) -> None:
-    """Persist a chunk boundary: scanned count, matches so far, and the keyset
-    cursor the next chunk resumes from. Keeps status='running'."""
-    session.execute(
-        update(LibraryAuditRun)
-        .where(LibraryAuditRun.id == audit_id)
-        .values(
-            scanned_count=scanned,
-            matched_count=len(matched),
-            matched_ids=list(matched),
-            resume_after_id=last_id,
-            last_progress_at=datetime.now(UTC),
-        )
-    )
-    session.commit()
-
-
 def _mark_error(session, audit_id: int, error_msg: str) -> None:
     session.execute(
         update(LibraryAuditRun)

backend/app/tasks/maintenance.py

@@ -83,13 +83,8 @@ TASK_RUN_KEEP_FAILURE_SECONDS = 7 * 24 * 3600  # 7 days
 # small buffer) so the sweep never flags in-flight work.
 #
 # Backups: images backup has time_limit=23400s (6.5h). 7h covers it
-# with a 30-min buffer.
+# with a 30-min buffer; db backup at 12 min hard limit fits trivially.
 BACKUP_STALL_THRESHOLD_MINUTES = 7 * 60
-# DB backup/restore is seconds-to-minutes (35-min hard limit). It must NOT share
-# the images' 7h window — a DB backup wedged on NFS would otherwise sit "running"
-# for 7 hours holding the concurrency-1 maintenance_long lane (operator-flagged
-# 2026-06-07). 40 min gives a small buffer over the hard limit.
-BACKUP_DB_STALL_THRESHOLD_MINUTES = 40
 # Library audit: scan_library_for_rule has time_limit=7500s (2h5m).
 # 2h15m gives a 10-min buffer.
 LIBRARY_AUDIT_STALL_THRESHOLD_MINUTES = 135
@@ -624,20 +619,16 @@ def recover_stalled_backup_runs() -> int:
     """
     SessionLocal = _sync_session_factory()
     now = datetime.now(UTC)
-    db_cutoff = now - timedelta(minutes=BACKUP_DB_STALL_THRESHOLD_MINUTES)
-    slow_cutoff = now - timedelta(minutes=BACKUP_STALL_THRESHOLD_MINUTES)
-    msg = "stranded by recovery sweep (no terminal status within the stall window)"
+    cutoff = now - timedelta(minutes=BACKUP_STALL_THRESHOLD_MINUTES)
+    msg = (
+        f"stranded by recovery sweep (no terminal status after "
+        f"{BACKUP_STALL_THRESHOLD_MINUTES // 60}h)"
+    )
     with SessionLocal() as session:
         result = session.execute(
             update(BackupRun)
             .where(BackupRun.status.in_(["running", "restoring"]))
-            # db backups/restores are fast (40-min window); images run hours (7h).
-            .where(
-                or_(
-                    and_(BackupRun.kind == "db", BackupRun.started_at < db_cutoff),
-                    and_(BackupRun.kind != "db", BackupRun.started_at < slow_cutoff),
-                )
-            )
+            .where(BackupRun.started_at < cutoff)
             .values(status="error", finished_at=now, error=msg)
         )
         session.commit()
@@ -656,30 +647,19 @@ def recover_stalled_library_audit_runs() -> int:
     guard in start_audit_run — a SIGKILL'd run would block all future
     audits until manual DB surgery. (The guard is now age-aware, but
     this sweep is what makes that work in practice.)
-
-    Measures staleness from last_progress_at (alembic 0039), NOT started_at:
-    a chunked scan stays 'running' across many re-enqueued chunks and can
-    legitimately run for hours on a big library — only flag one that hasn't
-    made progress in the threshold window (a dead chunk that never re-enqueued).
-    Falls back to started_at for pre-0039 / never-ticked rows.
     """
     SessionLocal = _sync_session_factory()
     now = datetime.now(UTC)
     cutoff = now - timedelta(minutes=LIBRARY_AUDIT_STALL_THRESHOLD_MINUTES)
     msg = (
-        f"stranded by recovery sweep (no progress for "
+        f"stranded by recovery sweep (no terminal status after "
         f"{LIBRARY_AUDIT_STALL_THRESHOLD_MINUTES} min)"
     )
     with SessionLocal() as session:
         result = session.execute(
             update(LibraryAuditRun)
             .where(LibraryAuditRun.status == "running")
-            .where(
-                func.coalesce(
-                    LibraryAuditRun.last_progress_at,
-                    LibraryAuditRun.started_at,
-                ) < cutoff
-            )
+            .where(LibraryAuditRun.started_at < cutoff)
             .values(status="error", finished_at=now, error=msg)
         )
         session.commit()

backend/app/utils/paths.py

@@ -2,26 +2,6 @@
 
 from pathlib import Path
 
-_MAX_EXT_LEN = 16
-
-
-def safe_ext(name: str | Path) -> str:
-    """Conservatively extract a short, alphanumeric file extension.
-
-    gallery-dl and Patreon CDN URLs produce basenames with URL-encoded
-    query-string artifacts, so `Path.suffix` can return 50+ chars of base64-ish
-    junk that blows bounded VARCHAR columns (e.g. PostAttachment.ext varchar(32)).
-    Accept only a suffix ≤16 chars whose post-dot characters are all alphanumeric;
-    otherwise return "" (no known extension). Operator-flagged 2026-05-25 — ONE
-    impl for the importer and the native Patreon client.
-    """
-    suffix = Path(name).suffix.lower()
-    if not suffix or len(suffix) > _MAX_EXT_LEN:
-        return ""
-    if not all(c.isalnum() for c in suffix[1:]):
-        return ""
-    return suffix
-
 
 def derive_subdir(source_path: Path, import_root: Path) -> str:
     """Returns the relative subdirectory of source_path under import_root.

backend/app/utils/safe_probe.py

@@ -149,8 +149,9 @@ def _run_probe(path_str: str) -> tuple[str, str | None]:
     call the same code path.
     """
     path = Path(path_str)
+    ext = path.suffix.lower()
     try:
-        total, test_bad = _inspect_archive(path)
+        total, test_bad = _inspect_archive(path, ext)
     except Exception as exc:  # noqa: BLE001 — clean rejection
         return ("error", f"{type(exc).__name__}: {exc}")
     if total is not None and total > MAX_ARCHIVE_UNCOMPRESSED_BYTES:
@@ -161,29 +162,24 @@ def _run_probe(path_str: str) -> tuple[str, str | None]:
     return ("ok", None)
 
 
-def _inspect_archive(path: Path):
+def _inspect_archive(path: Path, ext: str):
     """Return (total_uncompressed_bytes | None, first_bad_member | None)
-    for the archive. Format detected by extension OR magic bytes (so a
-    mis-named archive is still bomb-guarded + integrity-tested, matching the
-    extractor's gate); raises on a structurally-broken container (caught by the
-    child as a clean rejection)."""
-    from ..services.archive_extractor import detect_archive_format
-
-    fmt = detect_archive_format(path)
-    if fmt == "zip":
+    for the archive. Format-specific; raises on a structurally-broken
+    container (caught by the child as a clean rejection)."""
+    if ext in (".zip", ".cbz"):
         import zipfile
 
         with zipfile.ZipFile(path) as zf:
             total = sum(zi.file_size for zi in zf.infolist())
             return total, zf.testzip()
-    if fmt == "rar":
+    if ext == ".rar":
         import rarfile
 
         with rarfile.RarFile(path) as rf:
             total = sum(getattr(ri, "file_size", 0) for ri in rf.infolist())
             rf.testrar()
             return total, None
-    if fmt == "7z":
+    if ext == ".7z":
         import py7zr
 
         with py7zr.SevenZipFile(path, "r") as zf:
@@ -191,5 +187,5 @@ def _inspect_archive(path: Path):
             total = getattr(info, "uncompressed", None)
             ok = zf.test()  # True / None when all members pass
             return total, (None if ok in (True, None) else "7z test reported corruption")
-    # Not a recognised archive — nothing to test; treat as clean.
+    # Unknown extension — nothing to test; treat as clean.
     return None, None

docker-compose.override.yml

@@ -35,15 +35,6 @@ services:
     volumes:
       - ./backend:/app/backend
 
-  maintenance-long:
-    build:
-      context: .
-      dockerfile: Dockerfile
-    environment:
-      LOG_LEVEL: DEBUG
-    volumes:
-      - ./backend:/app/backend
-
   ml-worker:
     build:
       context: .

docker-compose.yml

@@ -27,17 +27,6 @@ services:
       POSTGRES_DB: ${DB_NAME:-fabledcurator}
     volumes:
       - postgres_data:/var/lib/postgresql/data
-      # Docker's default /dev/shm is 64MB; VACUUM (ANALYZE) + parallel queries
-      # allocate a POSIX dynamic-shared-memory segment in /dev/shm and fail with
-      # "could not resize shared memory segment ... No space left on device"
-      # (operator-flagged 2026-06-07, vacuum_analyze on import_task needed 67MB).
-      # NOTE: a `shm_size:` key is SILENTLY IGNORED under Docker Swarm
-      # (`docker stack deploy` — the prod deployment here), so size /dev/shm via
-      # a tmpfs mount instead — honored by both Swarm and plain Compose.
-      - type: tmpfs
-        target: /dev/shm
-        tmpfs:
-          size: 536870912  # 512 MB
     healthcheck:
       test: ["CMD-SHELL", "pg_isready -U ${DB_USER:-fabledcurator}"]
       interval: 10s
@@ -63,6 +52,7 @@ services:
     volumes:
       - ./images:/images
       - ./import:/import
+      - ./downloads:/downloads
       # FC-5 legacy migration: bind-mount the host's ImageRepo images dir
       # under /import (FC's existing filesystem scan picks them up). Read-only
       # is sufficient — FC copies into /images during the scan. The worker +
@@ -81,11 +71,10 @@ services:
       <<: *app_env
       CELERY_QUEUES: default,import,thumbnail,download
       CELERY_CONCURRENCY: "2"
-    # /downloads dropped — nothing in the app references it (operator-flagged
-    # 2026-06-07: it wasn't mapped in prod and everything worked).
     volumes:
       - ./images:/images
       - ./import:/import
+      - ./downloads:/downloads
     depends_on:
       postgres: { condition: service_healthy }
       redis:    { condition: service_healthy }
@@ -99,25 +88,7 @@ services:
     volumes:
       - ./images:/images
       - ./import:/import
-    depends_on:
-      postgres: { condition: service_healthy }
-      redis:    { condition: service_healthy }
-
-  # Dedicated lane for long one-shot maintenance (DB backups, library audits,
-  # admin maintenance). Kept off the scheduler's quick `maintenance` lane so a
-  # 30-min backup or a multi-chunk audit can never starve the 5-min recovery
-  # sweeps / vacuum (operator-flagged 2026-06-07). One slot — these are heavy.
-  maintenance-long:
-    image: git.fabledsword.com/bvandeusen/fabledcurator:dev
-    command: ["worker"]
-    environment:
-      <<: *app_env
-      CELERY_QUEUES: maintenance_long
-      CELERY_CONCURRENCY: "1"
-    # Only /images: backups write to /images/_backups, audits read /images, and
-    # the admin tasks (re-extract/cascade-delete/normalize) operate on /images.
-    volumes:
-      - ./images:/images
+      - ./downloads:/downloads
     depends_on:
       postgres: { condition: service_healthy }
       redis:    { condition: service_healthy }

frontend/src/components/discovery/ArtistCard.vue

@@ -49,11 +49,9 @@ function onCardClick() {
   position: relative;
   display: grid; grid-template-columns: repeat(3, 1fr);
   gap: 2px; aspect-ratio: 3 / 1;
-  /* Floor so tall source images can't escape the slot. NO max-height: an
-     aspect-ratio box capped by max-height shrinks its own WIDTH to keep the
-     ratio, leaving dead space beside the previews on a wide (single-column)
-     card. The grid below keeps columns ≲400px so the strip stays a sane height. */
-  min-height: 120px;
+  /* 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));
 }

frontend/src/components/modal/FandomPicker.vue

@@ -2,37 +2,17 @@
   <v-card>
     <v-card-title>Pick a fandom</v-card-title>
     <v-card-text>
-      <!-- Keyboard flow (operator-specified 2026-06-07):
-           1. Focus lands here (driven by the dialog's @after-enter calling the
-              exposed focusSearch — `autofocus` is unreliable inside a v-dialog
-              because the focus-trap activates after mount and steals it).
-           2. Tab moves to the "new fandom" field below.
-           3. Selecting a fandom (click, or arrow+Enter to pick from the menu)
-              leaves it in the field; a SECOND Enter (menu closed) accepts it.
-           We bind v-model:menu so the Enter handler can tell "pick from the open
-           menu" (let Vuetify handle) apart from "accept the chosen value". -->
       <v-autocomplete
-        ref="fandomRef"
         v-model="selectedId"
-        v-model:menu="menuOpen"
         :items="store.fandomCache"
         :item-title="(f) => f.name"
         :item-value="(f) => f.id"
         label="Fandom" clearable density="compact"
-        @keydown.enter.capture="onSearchEnter"
-        @keydown.tab="onSearchTab"
       />
       <v-divider class="my-3" />
       <p class="text-caption mb-2">Or create a new fandom:</p>
       <div class="d-flex" style="gap: 8px;">
-        <!-- Enter creates; on success focus returns to the dropdown (onCreate),
-             so the operator can immediately Enter-to-accept the new fandom. -->
-        <v-text-field
-          ref="newNameRef"
-          v-model="newName" placeholder="New fandom name"
-          density="compact" hide-details
-          @keydown.enter.prevent="onCreate"
-        />
+        <v-text-field v-model="newName" placeholder="New fandom name" density="compact" hide-details />
         <v-btn :disabled="!newName.trim()" rounded="pill" @click="onCreate">Create</v-btn>
       </div>
     </v-card-text>
@@ -49,7 +29,7 @@
 </template>
 
 <script setup>
-import { nextTick, onMounted, ref } from 'vue'
+import { onMounted, ref } from 'vue'
 import { useTagStore } from '../../stores/tags.js'
 
 const emit = defineEmits(['confirm', 'cancel'])
@@ -57,69 +37,21 @@ const store = useTagStore()
 
 const selectedId = ref(null)
 const newName = ref('')
-const menuOpen = ref(false)
-const fandomRef = ref(null)
-const newNameRef = ref(null)
 
-// Always refresh on open so the list reflects fandoms created elsewhere (#712).
-onMounted(() => store.loadFandoms())
+onMounted(() => { if (store.fandomCache.length === 0) store.loadFandoms() })
 
-// Exposed so the parent dialog can focus the search field on @after-enter —
-// the reliable point to grab focus (the dialog transition + focus-trap are done).
-function focusSearch () {
-  nextTick(() => {
-    fandomRef.value?.focus?.()
-    // Keep the menu closed after a programmatic focus so the very next Enter
-    // accepts the value instead of being swallowed as a menu interaction.
-    menuOpen.value = false
-  })
-}
-defineExpose({ focusSearch })
-
-async function onCreate () {
-  const name = newName.value.trim()
-  if (!name) return
-  const f = await store.createFandom(name)
+async function onCreate() {
+  const f = await store.createFandom(newName.value.trim())
   selectedId.value = f.id
   newName.value = ''
-  // The created fandom is now the selection; hand focus back to the dropdown
-  // so a single Enter accepts it (operator-specified flow 2026-06-07).
-  focusSearch()
 }
-
-function onConfirm () {
+function onConfirm() {
   const f = store.fandomCache.find(x => x.id === selectedId.value)
   if (f) emit('confirm', f)
 }
-
-// Enter on the search field — bound in the CAPTURE phase so this runs BEFORE
-// Vuetify's own input keydown handler (which opens the menu on Enter). When the
-// menu is open, let Vuetify pick the highlighted item (it sets selectedId +
-// closes the menu). When it's closed and a fandom is already chosen, accept it
-// and stop the event so Vuetify never (re)opens the menu — that re-open was the
-// bug: Enter popped the dropdown instead of submitting (operator-flagged
-// 2026-06-07).
-function onSearchEnter (e) {
-  if (menuOpen.value) return
-  if (selectedId.value != null) {
-    e.preventDefault()
-    e.stopPropagation()
-    onConfirm()
-  }
-}
-
-// Tab from the search field goes to the "new fandom" text field (operator-
-// specified). Shift+Tab keeps normal reverse traversal.
-function onSearchTab (e) {
-  if (e.shiftKey) return
-  e.preventDefault()
-  menuOpen.value = false
-  nextTick(() => newNameRef.value?.focus?.())
-}
-
 // Create the character with no fandom. Emits null so the caller knows this
 // was a deliberate "unassigned", not a cancel.
-function onNoFandom () {
+function onNoFandom() {
   emit('confirm', null)
 }
 </script>

frontend/src/components/modal/FandomSetDialog.vue

@@ -8,7 +8,6 @@
           :items="store.fandomCache"
           :item-title="(f) => f.name" :item-value="(f) => f.id"
           label="Fandom" clearable density="compact"
-          autofocus
           :hint="selectedId == null
             ? 'No fandom — the character will be unassigned.' : ''"
           persistent-hint
@@ -85,8 +84,7 @@ const busy = ref(false)
 const error = ref(null)
 const collision = ref(null)
 
-// Always refresh on open so the list reflects fandoms created elsewhere (#712).
-onMounted(() => store.loadFandoms())
+onMounted(() => { if (store.fandomCache.length === 0) store.loadFandoms() })
 
 async function onCreate() {
   const name = newName.value.trim()

frontend/src/components/modal/ImageViewer.vue

@@ -8,25 +8,6 @@
         <v-icon>mdi-close</v-icon>
       </button>
 
-      <!-- C8: keyboard cheatsheet. '?' toggles; the corner hint advertises it. -->
-      <button
-        class="fc-viewer__help-hint" aria-label="Keyboard shortcuts (press ?)"
-        @click="showHelp = !showHelp"
-      >?</button>
-      <div v-if="showHelp" class="fc-viewer__help" @click.self="showHelp = false">
-        <div class="fc-viewer__help-card" role="dialog" aria-label="Keyboard shortcuts">
-          <h3>Keyboard shortcuts</h3>
-          <dl>
-            <div><dt>← / →</dt><dd>Previous / next image</dd></div>
-            <div><dt>T or /</dt><dd>Jump to the tag input</dd></div>
-            <div><dt>↑ / ↓</dt><dd>Move through tag suggestions</dd></div>
-            <div><dt>Enter / Tab</dt><dd>Accept the highlighted tag</dd></div>
-            <div><dt>Esc</dt><dd>Close a dialog, then the viewer</dd></div>
-            <div><dt>?</dt><dd>Toggle this help</dd></div>
-          </dl>
-        </div>
-      </div>
-
       <v-chip
         v-if="integrityBadge"
         :color="integrityBadge.color"
@@ -47,17 +28,12 @@
         <v-icon size="32">mdi-chevron-right</v-icon>
       </button>
 
-      <!-- Large centered loading spinner (operator-flagged 2026-06-07: the old
-           size-36 one sat tiny in the top-left). Dual counter-rotating accent
-           rings, centered over the whole modal. pointer-events:none so the
-           close/nav controls underneath stay clickable. -->
-      <div v-if="modal.loading" class="fc-viewer__loading">
-        <div class="fc-viewer__spinner" />
-      </div>
-
       <div class="fc-viewer__body">
         <div class="fc-viewer__media">
-          <template v-if="modal.current">
+          <v-progress-circular
+            v-if="modal.loading" indeterminate color="accent" size="36"
+          />
+          <template v-else-if="modal.current">
             <ImageCanvas
               v-if="!isVideo" :src="modal.current.image_url" :alt="`Image ${modal.current.id}`"
               @close-request="$emit('close')"
@@ -87,7 +63,7 @@
 <script setup>
 import { computed, onMounted, onUnmounted, ref, watch } from 'vue'
 import { useModalStore } from '../../stores/modal.js'
-import { arrowNavAllowed, isTextEntry } from '../../utils/textEntry.js'
+import { arrowNavAllowed } from '../../utils/textEntry.js'
 import ImageCanvas from './ImageCanvas.vue'
 import VideoCanvas from './VideoCanvas.vue'
 import TagPanel from './TagPanel.vue'
@@ -98,7 +74,6 @@ const emit = defineEmits(['close'])
 
 const modal = useModalStore()
 const rootEl = ref(null)
-const showHelp = ref(false)
 
 const isVideo = computed(() =>
   modal.current?.mime && modal.current.mime.startsWith('video/')
@@ -121,27 +96,17 @@ let prevBodyOverflow = null
 // that. Filtered via isTextEntry so tag/comment inputs still get their
 // own keystrokes.
 function onKeyDown(ev) {
-  if (ev.key === 'Escape' && showHelp.value) {
-    // Close the cheatsheet first; a second Esc closes the modal.
-    ev.preventDefault()
-    showHelp.value = false
-    return
-  }
   if (ev.key === 'Escape') {
-    // Escape closes the modal even from inside a text input — the universal
-    // "get me out of here" expectation; the autofocused tag field would
-    // otherwise trap focus (operator-flagged 2026-06-01). EXCEPTION: when the
-    // keystroke originates INSIDE an open Vuetify overlay's content (a rename/
-    // fandom/alias dialog, or a kebab menu), let that overlay handle its own
-    // Esc and don't close the whole modal.
-    //
-    // #700 re-fix (2026-06-07): the prior guard queried for ANY active overlay
-    // anywhere in the DOM and suppressed the close — so a lingering overlay
-    // after accepting a suggestion (focus drops to <body>, not into any
-    // overlay) wrongly blocked Esc. Keying off the event's origin instead means
-    // a stray overlay no longer traps the modal: only an Esc pressed from
-    // within overlay content defers to that overlay.
-    if (ev.target?.closest?.('.v-overlay__content')) return
+    // Escape closes the modal even from inside a text input — that's
+    // the universal "get me out of here" expectation, and the
+    // autofocused tag-entry field would otherwise trap focus with no
+    // visible escape (operator-flagged 2026-06-01). EXCEPTION: when a
+    // nested Vuetify overlay is open (v-menu autocomplete dropdown,
+    // FandomPicker v-dialog, per-suggestion 3-dot menu), let that
+    // overlay's own Esc handling fire instead of closing the whole
+    // modal mid-interaction. Vuetify marks open overlays with
+    // `.v-overlay--active`.
+    if (document.querySelector('.v-overlay--active')) return
     ev.preventDefault()
     emit('close')
   } else if (ev.key === 'ArrowLeft') {
@@ -154,14 +119,6 @@ function onKeyDown(ev) {
     if (!arrowNavAllowed(ev.target)) return
     ev.preventDefault()
     modal.goNext()
-  } else if ((ev.key === '/' || ev.key === 't') && !isTextEntry(ev.target)) {
-    // C9 (2026-06-07): jump focus to the tag input from anywhere in the modal.
-    const input = document.querySelector('.fc-tag-autocomplete input')
-    if (input) { ev.preventDefault(); input.focus() }
-  } else if (ev.key === '?' && !isTextEntry(ev.target)) {
-    // C8: toggle the keyboard cheatsheet.
-    ev.preventDefault()
-    showHelp.value = !showHelp.value
   }
 }
 
@@ -217,43 +174,6 @@ function nextFrame() {
 }
 .fc-viewer__nav:disabled { opacity: 0.3; cursor: not-allowed; }
 .fc-viewer__close { top: 16px; right: 16px; transform: none; }
-.fc-viewer__help-hint {
-  position: absolute; top: 16px; right: 64px; z-index: 2;
-  width: 32px; height: 32px; border-radius: 50%;
-  background: rgba(255, 255, 255, 0.08);
-  color: rgb(var(--v-theme-on-surface));
-  font-weight: 700; cursor: pointer; opacity: 0.6;
-}
-.fc-viewer__help-hint:hover { opacity: 1; }
-.fc-viewer__help {
-  position: absolute; inset: 0; z-index: 4;
-  display: flex; align-items: center; justify-content: center;
-  background: rgba(0, 0, 0, 0.4);
-}
-.fc-viewer__help-card {
-  background: rgb(var(--v-theme-surface));
-  border: 1px solid rgb(var(--v-theme-surface-light));
-  border-radius: 10px; padding: 20px 24px; min-width: 320px;
-  box-shadow: 0 8px 32px rgba(0, 0, 0, 0.5);
-}
-.fc-viewer__help-card h3 {
-  font-family: 'Fraunces', Georgia, serif;
-  font-size: 18px; margin-bottom: 12px;
-  color: rgb(var(--v-theme-on-surface));
-}
-.fc-viewer__help-card dl { display: flex; flex-direction: column; gap: 8px; }
-.fc-viewer__help-card dl > div {
-  display: flex; gap: 16px; align-items: baseline;
-}
-.fc-viewer__help-card dt {
-  flex: 0 0 96px; text-align: right;
-  font-family: 'JetBrains Mono', monospace; font-size: 12px;
-  color: rgb(var(--v-theme-accent));
-}
-.fc-viewer__help-card dd {
-  flex: 1; font-size: 14px;
-  color: rgb(var(--v-theme-on-surface));
-}
 .fc-viewer__integrity {
   position: absolute; top: 72px; right: 16px; z-index: 3;
 }
@@ -264,34 +184,6 @@ function nextFrame() {
   right: calc(var(--fc-side-w) + 16px);
   transform: translateY(-50%);
 }
-/* Centered loading overlay — sits over the media band; non-interactive so the
-   close/nav controls underneath keep working. */
-.fc-viewer__loading {
-  position: absolute; inset: 0; z-index: 2;
-  display: flex; align-items: center; justify-content: center;
-  pointer-events: none;
-}
-/* Two concentric accent rings counter-rotating, each lit on a different arc —
-   reads as one orbiting object. ~108px so it's unmistakably the focal point. */
-.fc-viewer__spinner {
-  width: 108px; height: 108px; border-radius: 50%;
-  border: 5px solid rgb(var(--v-theme-accent), 0.16);
-  border-top-color: rgb(var(--v-theme-accent));
-  animation: fc-viewer-spin 0.95s cubic-bezier(0.5, 0.1, 0.5, 0.9) infinite;
-  position: relative;
-  box-shadow: 0 0 24px rgb(var(--v-theme-accent), 0.18);
-}
-.fc-viewer__spinner::after {
-  content: ''; position: absolute; inset: 14px; border-radius: 50%;
-  border: 5px solid rgb(var(--v-theme-accent), 0.10);
-  border-bottom-color: rgb(var(--v-theme-accent));
-  animation: fc-viewer-spin 1.5s cubic-bezier(0.5, 0.1, 0.5, 0.9) infinite reverse;
-}
-@keyframes fc-viewer-spin { to { transform: rotate(360deg); } }
-@media (prefers-reduced-motion: reduce) {
-  .fc-viewer__spinner, .fc-viewer__spinner::after { animation-duration: 3s; }
-}
-
 .fc-viewer__body {
   flex: 1; display: flex; min-height: 0;
 }
@@ -311,37 +203,14 @@ function nextFrame() {
 }
 
 @media (max-width: 900px) {
-  /* Stacked layout (operator-flagged 2026-06-05): pin the image pane and
-     its controls at the top while the metadata panel scrolls beneath it.
-     Previously the image + a 40vh panel shared one fixed viewport and the
-     controls could be pushed out of view; now the body scrolls and the
-     media is sticky, so the image + prev/next/close (and the integrity
-     badge) stay visible no matter how far the panel is scrolled. */
-  .fc-viewer__body {
-    flex-direction: column;
-    overflow-y: auto;
-  }
-  .fc-viewer__media {
-    flex: none;
-    height: 55vh;
-    position: sticky;
-    top: 0;
-    z-index: 1;
-    /* Opaque obsidian so the scrolling panel never bleeds through the
-       haze behind the pinned image. */
-    background: rgb(20, 23, 26);
-  }
+  .fc-viewer__body { flex-direction: column; }
+  /* Side panel drops below the image — the next arrow uses the full width. */
+  .fc-viewer__nav--next { right: 16px; }
   .fc-viewer__side {
     width: 100%;
-    max-height: none;
+    max-height: 40vh;
     border-left: none;
     border-top: 1px solid rgb(var(--v-theme-surface-light));
   }
-  /* Re-center the prev/next arrows over the 55vh image band (their base
-     top:50% would land on the scrolling panel); next uses the full width
-     now that the panel is below, not beside. Close + integrity badge keep
-     their top:16px/72px — already within the pinned band. */
-  .fc-viewer__nav { top: 27.5vh; }
-  .fc-viewer__nav--next { right: 16px; }
 }
 </style>

frontend/src/components/modal/SuggestionItem.vue

@@ -37,7 +37,6 @@
       />
       <v-menu
         v-model="menuOpen" activator="parent" :open-on-click="false"
-        :z-index="2400"
       >
         <v-list density="compact">
           <v-list-item @click="$emit('alias', suggestion)">

frontend/src/components/modal/TagAutocomplete.vue

@@ -8,87 +8,46 @@
       @keydown.down.prevent="moveHighlight(1)"
       @keydown.up.prevent="moveHighlight(-1)"
       @keydown.enter.prevent="onEnter"
-      @keydown.tab="onTab"
       @keydown.esc="$emit('cancel')"
     />
     <v-list
-      v-if="rows.length"
-      ref="listRef"
+      v-if="hits.length || allowCreate"
       density="compact" class="fc-tag-autocomplete__list"
     >
-      <template v-for="(row, idx) in rows" :key="row.key">
-        <!-- Existing tag match (server autocomplete). -->
-        <v-list-item
-          v-if="row.type === 'hit'"
-          :active="idx === highlight" @click="onPickRow(row)"
-        >
-          <template #prepend>
-            <v-icon size="small" :color="store.colorFor(row.hit.kind)">
-              {{ iconFor(row.hit.kind) }}
-            </v-icon>
-          </template>
-          <v-list-item-title>
-            {{ row.hit.name }}
-            <span v-if="row.hit.fandom_name" class="text-caption">— {{ row.hit.fandom_name }}</span>
-          </v-list-item-title>
-          <template #append>
-            <span class="text-caption">{{ row.hit.kind }}</span>
-          </template>
-        </v-list-item>
-
-        <!-- ML suggestion for THIS image that matches the typed query. Picking
-             it routes through the same accept path as the Suggestions panel, so
-             it's recorded + drops out of the panel (operator-asked 2026-06-07). -->
-        <v-list-item
-          v-else-if="row.type === 'suggestion'"
-          :active="idx === highlight" @click="onPickRow(row)"
-          class="fc-tag-autocomplete__sugg"
-        >
-          <template #prepend>
-            <v-icon size="small" :color="store.colorFor(row.sugg.category)">
-              {{ iconFor(row.sugg.category) }}
-            </v-icon>
-          </template>
-          <v-list-item-title>
-            {{ row.sugg.display_name }}
-            <span v-if="row.sugg.creates_new_tag" class="text-caption">— new</span>
-          </v-list-item-title>
-          <template #append>
-            <span class="fc-tag-autocomplete__sugg-tag">
-              <v-icon size="x-small">mdi-auto-fix</v-icon>
-              {{ scorePct(row.sugg) }}
-            </span>
-          </template>
-        </v-list-item>
-
-        <!-- Create-new row. -->
-        <v-list-item
-          v-else
-          :active="idx === highlight" @click="onPickRow(row)"
-        >
-          <template #prepend>
-            <v-icon size="small" :color="store.colorFor(parsedKind)">
-              {{ iconFor(parsedKind) }}
-            </v-icon>
-          </template>
-          <v-list-item-title>
-            Create "{{ parsedName }}" as {{ parsedKind }}
-          </v-list-item-title>
-        </v-list-item>
-      </template>
+      <v-list-item
+        v-for="(h, idx) in hits" :key="h.id"
+        :active="idx === highlight" @click="onPick(h)"
+      >
+        <template #prepend>
+          <v-icon size="small" :color="store.colorFor(h.kind)">
+            {{ iconFor(h.kind) }}
+          </v-icon>
+        </template>
+        <v-list-item-title>
+          {{ h.name }}
+          <span v-if="h.fandom_name" class="text-caption">— {{ h.fandom_name }}</span>
+        </v-list-item-title>
+        <template #append>
+          <span class="text-caption">{{ h.kind }}</span>
+        </template>
+      </v-list-item>
+      <v-list-item
+        v-if="allowCreate" :active="highlight === hits.length"
+        @click="onCreate"
+      >
+        <template #prepend>
+          <v-icon size="small" :color="store.colorFor(parsedKind)">
+            {{ iconFor(parsedKind) }}
+          </v-icon>
+        </template>
+        <v-list-item-title>
+          Create "{{ parsedName }}" as {{ parsedKind }}
+        </v-list-item-title>
+      </v-list-item>
     </v-list>
 
-    <!-- @after-enter is the reliable moment to focus the picker: the dialog's
-         open transition + focus-trap have settled, so focusSearch lands in the
-         fandom field instead of fighting the trap (operator-flagged 2026-06-07). -->
-    <v-dialog
-      v-model="fandomDialog" max-width="480"
-      @after-enter="fandomPickerRef?.focusSearch?.()"
-    >
-      <FandomPicker
-        ref="fandomPickerRef"
-        @confirm="onFandomChosen" @cancel="onFandomCancel"
-      />
+    <v-dialog v-model="fandomDialog" max-width="480">
+      <FandomPicker @confirm="onFandomChosen" @cancel="fandomDialog = false" />
     </v-dialog>
   </div>
 </template>
@@ -96,33 +55,18 @@
 <script setup>
 import { computed, nextTick, onMounted, ref, watch } from 'vue'
 import { useTagStore } from '../../stores/tags.js'
-import { useSuggestionsStore } from '../../stores/suggestions.js'
 import FandomPicker from './FandomPicker.vue'
 
-const emit = defineEmits(['pick-existing', 'pick-new', 'accept-suggestion', 'cancel'])
+const emit = defineEmits(['pick-existing', 'pick-new', 'cancel'])
 const store = useTagStore()
-// The image's ML (Camie) suggestions, surfaced inline in this dropdown so the
-// operator can pick a suggestion while typing instead of hunting for it in the
-// Suggestions panel below (operator-asked 2026-06-07).
-const suggestions = useSuggestionsStore()
 
 // Autofocus on modal open so the operator can type the moment the view
 // modal renders, no extra click required (operator-asked 2026-06-01).
 // Vuetify's v-text-field exposes .focus() on the component instance;
 // nextTick waits for the modal's mount to finish so the inner <input>
 // element exists.
-//
-// EXCEPTION — not on mobile (operator-flagged 2026-06-05): focusing the
-// field pops the soft keyboard, which shrinks the visual viewport and
-// shoves the pinned image + its nav/close controls out of view. 900px is
-// the modal's stacked-layout breakpoint (ImageViewer.vue). matchMedia is
-// safe on plain HTTP (not a secure-context API).
 const inputRef = ref(null)
-function focusInput () {
-  if (window.matchMedia?.('(max-width: 900px)')?.matches) return
-  nextTick(() => inputRef.value?.focus?.())
-}
-onMounted(focusInput)
+onMounted(() => { nextTick(() => inputRef.value?.focus?.()) })
 
 // Single text input; no kind dropdown. Client-side mirror of the
 // backend's parse_kind_prefix lives below — kept in sync with
@@ -132,9 +76,7 @@ onMounted(focusInput)
 const query = ref('')
 const hits = ref([])
 const highlight = ref(0)
-const listRef = ref(null)
 const fandomDialog = ref(false)
-const fandomPickerRef = ref(null)
 let pendingNewName = null
 
 const KNOWN_KINDS = new Set([
@@ -182,66 +124,13 @@ const allowCreate = computed(() => {
   )
 })
 
-function scorePct (s) { return `${Math.round(s.score * 100)}%` }
-
-// This image's suggestions that match the typed query, minus any the server
-// autocomplete already returned (same name+kind) so a tag never shows twice.
-const suggestionHits = computed(() => {
-  const q = parsedName.value.toLowerCase()
-  if (!q) return []
-  const seen = new Set(hits.value.map(h => `${h.kind}:${h.name.toLowerCase()}`))
-  const out = []
-  for (const list of Object.values(suggestions.byCategory)) {
-    for (const s of list || []) {
-      const key = `${s.category}:${s.display_name.toLowerCase()}`
-      if (!s.display_name.toLowerCase().includes(q)) continue
-      if (seen.has(key)) continue
-      seen.add(key)
-      out.push(s)
-    }
-  }
-  // Best matches first; cap so the dropdown stays scannable.
-  out.sort((a, b) => b.score - a.score)
-  return out.slice(0, 6)
-})
-
-// One ordered list backing both the rendered dropdown and keyboard nav, so the
-// highlight index maps 1:1 to a row regardless of which section it's in.
-const rows = computed(() => {
-  const r = hits.value.map(h => ({ type: 'hit', key: `h${h.id}`, hit: h }))
-  for (const s of suggestionHits.value) {
-    r.push({ type: 'suggestion', key: `s${s.category}:${s.display_name}`, sugg: s })
-  }
-  if (allowCreate.value) r.push({ type: 'create', key: 'create' })
-  return r
-})
-
 function moveHighlight (delta) {
-  const total = rows.value.length
+  const total = hits.value.length + (allowCreate.value ? 1 : 0)
   if (total === 0) return
   highlight.value = (highlight.value + delta + total) % total
-  // Keep the highlighted row visible — the list is capped at 240px and arrowing
-  // past the fold otherwise left the active item off-screen (operator-flagged
-  // 2026-06-07). block:'nearest' scrolls the minimum needed.
-  nextTick(() => {
-    listRef.value?.$el
-      ?.querySelector('.v-list-item--active')
-      ?.scrollIntoView({ block: 'nearest' })
-  })
 }
 
-// Dispatch a chosen dropdown row by its type.
-function onPickRow (row) {
-  if (row.type === 'hit') { emit('pick-existing', row.hit); reset() }
-  else if (row.type === 'suggestion') { onPickSuggestion(row.sugg) }
-  else { onCreate() }
-}
-
-// Picking an ML suggestion hands it to the parent, which runs the same accept
-// flow as the Suggestions panel (creates the tag if raw, records acceptance,
-// drops it from the panel). reset() clears the query; the panel-drop makes it
-// fall out of suggestionHits reactively.
-function onPickSuggestion (s) { emit('accept-suggestion', s); reset() }
+function onPick (hit) { emit('pick-existing', hit); reset() }
 
 function onCreate () {
   const name = parsedName.value
@@ -269,31 +158,13 @@ function onFandomChosen (fandom) {
   })
   pendingNewName = null
   reset()
-  // Return focus to the tag input so the keyboard flow continues straight into
-  // the next tag instead of dropping to <body> (operator-flagged 2026-06-07).
-  focusInput()
-}
-
-// Cancelling the fandom dialog drops the pending character and hands focus back
-// to the tag input, same as a successful pick.
-function onFandomCancel () {
-  fandomDialog.value = false
-  pendingNewName = null
-  focusInput()
 }
 
 function onEnter () {
-  const row = rows.value[highlight.value]
-  if (row) onPickRow(row)
-}
-
-// B5 (2026-06-07): when the dropdown is open, Tab accepts the highlighted row
-// (standard autocomplete convention) instead of leaving the field. With the
-// list closed it falls through to normal focus traversal.
-function onTab (e) {
-  if (rows.value.length) {
-    e.preventDefault()
-    onEnter()
+  if (highlight.value < hits.value.length) {
+    onPick(hits.value[highlight.value])
+  } else if (allowCreate.value) {
+    onCreate()
   }
 }
 
@@ -309,14 +180,4 @@ function reset () { query.value = ''; hits.value = []; highlight.value = 0 }
   border: 1px solid rgb(var(--v-theme-surface-light));
   border-radius: 6px;
 }
-/* Mark ML-suggestion rows so they read as distinct from typed/known matches. */
-.fc-tag-autocomplete__sugg {
-  border-left: 2px solid rgb(var(--v-theme-accent), 0.5);
-}
-.fc-tag-autocomplete__sugg-tag {
-  display: inline-flex; align-items: center; gap: 2px;
-  font-size: 11px;
-  font-family: 'JetBrains Mono', monospace;
-  color: rgb(var(--v-theme-accent));
-}
 </style>

frontend/src/components/modal/TagChip.vue

@@ -1,65 +0,0 @@
-<template>
-  <!-- One tag chip + its kebab. The kebab uses the SAME explicit-menu pattern
-       as SuggestionItem (activator="parent" + :open-on-click="false" + a manual
-       v-model), because the #activator / v-bind="props" pattern never toggles a
-       v-menu inside the teleported ImageViewer modal (#711, re-fixed 2026-06-07
-       after the first attempt used the broken pattern). -->
-  <span class="fc-tag-chip">
-    <v-chip
-      size="small" closable
-      :color="store.colorFor(tag.kind)" variant="tonal"
-      @click:close="$emit('remove', tag.id)"
-    >
-      <v-icon start size="x-small">{{ iconFor(tag.kind) }}</v-icon>
-      {{ tag.name }}<span v-if="tag.fandom_id">→</span>
-    </v-chip>
-    <span class="fc-tag-chip__menu-wrap">
-      <v-btn
-        class="fc-tag-chip__kebab"
-        icon="mdi-dots-vertical" size="x-small"
-        variant="text" density="comfortable"
-        :aria-label="`More actions for ${tag.name}`"
-        @click.stop="menuOpen = !menuOpen"
-      />
-      <!-- :z-index above the modal's 2000 — the teleported menu otherwise lands
-           BEHIND .fc-viewer and gets blurred by its backdrop-filter (operator
-           saw it ghosted behind the sidebar, 2026-06-07). THIS is what made the
-           kebab look dead the whole time: it opened, just underneath. -->
-      <v-menu v-model="menuOpen" activator="parent" :open-on-click="false" :z-index="2400">
-        <v-list density="compact">
-          <v-list-item @click="$emit('rename', tag)">
-            <v-list-item-title>Rename…</v-list-item-title>
-          </v-list-item>
-          <v-list-item v-if="tag.kind === 'character'" @click="$emit('set-fandom', tag)">
-            <v-list-item-title>Set fandom…</v-list-item-title>
-          </v-list-item>
-        </v-list>
-      </v-menu>
-    </span>
-  </span>
-</template>
-
-<script setup>
-import { ref } from 'vue'
-import { useTagStore } from '../../stores/tags.js'
-
-defineProps({ tag: { type: Object, required: true } })
-defineEmits(['remove', 'rename', 'set-fandom'])
-
-const store = useTagStore()
-const menuOpen = ref(false)
-
-const KIND_ICONS = {
-  general: 'mdi-tag', character: 'mdi-account-circle',
-  fandom: 'mdi-book-open-page-variant', series: 'mdi-bookshelf',
-  meta: 'mdi-cog-outline', rating: 'mdi-shield-check-outline',
-}
-function iconFor (k) { return KIND_ICONS[k] || 'mdi-tag' }
-</script>
-
-<style scoped>
-.fc-tag-chip { display: inline-flex; align-items: center; gap: 1px; }
-.fc-tag-chip__menu-wrap { display: inline-flex; align-items: center; }
-.fc-tag-chip__kebab { opacity: 0.7; }
-.fc-tag-chip:hover .fc-tag-chip__kebab { opacity: 1; }
-</style>

frontend/src/components/modal/TagPanel.vue

@@ -2,11 +2,45 @@
   <aside class="fc-tag-panel" aria-label="Tags for this image">
     <h3 class="fc-tag-panel__title">Tags</h3>
     <div class="fc-tag-panel__chips">
-      <TagChip
+      <v-chip
         v-for="tag in modal.current?.tags || []"
-        :key="tag.id" :tag="tag"
-        @remove="onRemove" @rename="openRename" @set-fandom="openSetFandom"
-      />
+        :key="tag.id" size="small" closable
+        :color="store.colorFor(tag.kind)" variant="tonal"
+        @click:close="onRemove(tag.id)"
+      >
+        <v-icon start size="x-small">{{ iconFor(tag.kind) }}</v-icon>
+        {{ tag.name }}<span v-if="tag.fandom_id">→</span>
+        <!-- Operator-flagged 2026-06-04: the `#activator` + `v-bind` menu
+             never opened inside this teleported modal. Drive it explicitly
+             instead (same mechanism as the dialogs below, which work): the
+             icon toggles `openTagId` with @click.stop (shielding the chip's
+             close button), and `activator="parent"` + `:open-on-click=false`
+             anchors the menu for positioning only. One tag's menu open at a
+             time, so a single id is enough. -->
+        <span class="kebab-wrap">
+          <v-icon
+            size="x-small" class="ml-1 kebab-icon"
+            icon="mdi-dots-vertical"
+            @click.stop="openTagId = openTagId === tag.id ? null : tag.id"
+          />
+          <v-menu
+            :model-value="openTagId === tag.id"
+            activator="parent" :open-on-click="false"
+            @update:model-value="v => { if (!v) openTagId = null }"
+          >
+            <v-list density="compact">
+              <v-list-item @click="openRename(tag)">
+                <v-list-item-title>Rename…</v-list-item-title>
+              </v-list-item>
+              <v-list-item
+                v-if="tag.kind === 'character'" @click="openSetFandom(tag)"
+              >
+                <v-list-item-title>Set fandom…</v-list-item-title>
+              </v-list-item>
+            </v-list>
+          </v-menu>
+        </span>
+      </v-chip>
       <span v-if="!modal.current?.tags?.length" class="text-caption">No tags yet.</span>
     </div>
 
@@ -14,7 +48,6 @@
 
     <TagAutocomplete
       @pick-existing="onPickExisting" @pick-new="onPickNew"
-      @accept-suggestion="onAcceptSuggestion"
     />
 
     <v-alert v-if="errorMsg" type="error" variant="tonal" class="mt-2" closable>
@@ -46,16 +79,25 @@
 <script setup>
 import { ref } from 'vue'
 import { useModalStore } from '../../stores/modal.js'
-import { useSuggestionsStore } from '../../stores/suggestions.js'
-import TagChip from './TagChip.vue'
+import { useTagStore } from '../../stores/tags.js'
 import TagAutocomplete from './TagAutocomplete.vue'
 import SuggestionsPanel from './SuggestionsPanel.vue'
 import TagRenameDialog from './TagRenameDialog.vue'
 import FandomSetDialog from './FandomSetDialog.vue'
 
 const modal = useModalStore()
-const suggestions = useSuggestionsStore()
+const store = useTagStore()
 const errorMsg = ref(null)
+// Which tag chip's kebab menu is open (only one at a time). Drives each
+// chip menu's v-model so opening never depends on Vuetify's activator click.
+const openTagId = ref(null)
+
+const KIND_ICONS = {
+  general: 'mdi-tag', character: 'mdi-account-circle',
+  fandom: 'mdi-book-open-page-variant', series: 'mdi-bookshelf',
+  meta: 'mdi-cog-outline', rating: 'mdi-shield-check-outline'
+}
+function iconFor(k) { return KIND_ICONS[k] || 'mdi-tag' }
 
 async function onRemove(tagId) {
   errorMsg.value = null
@@ -72,16 +114,6 @@ async function onPickNew(payload) {
   try { await modal.createAndAdd(payload) }
   catch (e) { errorMsg.value = e.message }
 }
-// A suggestion picked from the autocomplete dropdown runs the SAME path as the
-// Suggestions panel's Accept: the store creates the tag if it's raw, records the
-// acceptance, and drops it from the panel; then we refresh the chip rail.
-async function onAcceptSuggestion(s) {
-  errorMsg.value = null
-  try {
-    await suggestions.accept(s)
-    await modal.reloadTags()
-  } catch (e) { errorMsg.value = e.message }
-}
 
 const renameDialog = ref(false)
 const renameTarget = ref(null)
@@ -121,4 +153,6 @@ async function onFandomUpdated() {
   margin-bottom: 12px;
 }
 .fc-tag-panel__chips { display: flex; flex-wrap: wrap; gap: 6px; }
+.kebab-wrap { display: inline-flex; align-items: center; }
+.kebab-icon { cursor: pointer; }
 </style>

frontend/src/components/posts/PostCard.vue

@@ -15,7 +15,6 @@
         · {{ totalImages }} image{{ totalImages === 1 ? '' : 's' }}
       </span>
       <v-spacer />
-      <PostSeriesMenu :post="post" />
       <v-btn
         v-if="post.post_url"
         :href="post.post_url" target="_blank" rel="noopener"
@@ -35,10 +34,7 @@
           >
             <img :src="hero.thumbnail_url" alt="hero thumbnail" loading="lazy" />
           </button>
-          <div
-            v-if="rail.length || moreCount" class="fc-post-card__rail"
-            :style="{ '--fc-rail-cols': railCols }"
-          >
+          <div v-if="rail.length || moreCount" class="fc-post-card__rail">
             <button
               v-for="t in rail" :key="t.image_id" type="button"
               class="fc-post-card__rail-cell"
@@ -102,7 +98,6 @@ import { useModalStore } from '../../stores/modal.js'
 import { usePostsStore } from '../../stores/posts.js'
 import { toPlainText } from '../../utils/htmlSanitize.js'
 import PostEmptyThumbs from './PostEmptyThumbs.vue'
-import PostSeriesMenu from './PostSeriesMenu.vue'
 
 const props = defineProps({
   post: { type: Object, required: true },
@@ -122,29 +117,13 @@ const totalImages = computed(() => images.value.length + (props.post.thumbnails_
 const plainTitle = computed(() => toPlainText(props.post.post_title))
 
 const hero = computed(() => images.value[0])
-// The thumbnail strip spans the hero's full width (CSS grid, equal columns),
-// rather than a fixed 3-cell cap. Show up to RAIL_MAX cells; when there are
-// more images than fit, the last cell becomes a "+N" overflow tile so the
-// count stays accurate.
-const RAIL_MAX = 5
-const serverMore = computed(() => props.post.thumbnails_more || 0)
-const afterHero = computed(() => images.value.slice(1))
-const hasOverflow = computed(
-  () => serverMore.value > 0 || afterHero.value.length > RAIL_MAX,
-)
-const rail = computed(() =>
-  hasOverflow.value
-    ? afterHero.value.slice(0, RAIL_MAX - 1)
-    : afterHero.value.slice(0, RAIL_MAX),
-)
+const rail = computed(() => images.value.slice(1, 4))
 const visibleCount = computed(() => (images.value.length ? 1 + rail.value.length : 0))
 const moreCount = computed(() => {
-  if (!hasOverflow.value) return 0
-  return serverMore.value + Math.max(0, images.value.length - visibleCount.value)
+  const more = props.post.thumbnails_more || 0
+  const extraShown = Math.max(0, images.value.length - visibleCount.value)
+  return more + extraShown
 })
-// Columns = thumbnails shown plus the overflow tile, so the strip fills the
-// hero's full width with equal cells.
-const railCols = computed(() => rail.value.length + (moreCount.value > 0 ? 1 : 0))
 
 const sortDateIso = computed(() => props.post.post_date || props.post.downloaded_at)
 const absoluteDate = computed(() => new Date(sortDateIso.value).toLocaleString())
@@ -300,16 +279,11 @@ function formatBytes (n) {
 .fc-post-card__hero:hover img,
 .fc-post-card__rail-cell:hover img { transform: scale(1.03); filter: brightness(1.08); }
 
-/* Equal columns spanning the hero's full width — cells stretch horizontally
-   (1fr) at a fixed height, so the strip always reaches the edge of the hero
-   regardless of how many thumbnails the post has. */
 .fc-post-card__rail {
-  display: grid;
-  grid-template-columns: repeat(var(--fc-rail-cols, 3), 1fr);
-  gap: 6px; margin-top: 6px;
+  display: flex; gap: 6px; margin-top: 6px; flex-wrap: wrap;
 }
 .fc-post-card__rail-cell {
-  width: 100%; height: 84px;
+  width: 80px; height: 80px;
   overflow: hidden; border-radius: 4px;
 }
 .fc-post-card__rail-cell img {
@@ -317,7 +291,7 @@ function formatBytes (n) {
   object-fit: cover; display: block;
 }
 .fc-post-card__rail-more {
-  width: 100%; height: 84px;
+  width: 80px; height: 80px;
   display: flex; align-items: center; justify-content: center;
   border: 1px dashed rgb(var(--v-theme-on-surface-variant));
   border-radius: 4px;

frontend/src/components/posts/PostSeriesMenu.vue

@@ -1,123 +0,0 @@
-<template>
-  <span class="fc-post-series">
-    <v-btn
-      size="x-small" variant="text" prepend-icon="mdi-book-plus-outline"
-      append-icon="mdi-menu-down"
-      :aria-label="`Add post ${post.id} to a series`"
-      @click.stop="menuOpen = !menuOpen"
-    >Series</v-btn>
-    <v-menu v-model="menuOpen" activator="parent" :open-on-click="false">
-      <v-list density="compact">
-        <v-list-item :disabled="busy" @click="onPromote">
-          <template #prepend>
-            <v-icon size="small">mdi-book-plus</v-icon>
-          </template>
-          <v-list-item-title>New series from this post</v-list-item-title>
-        </v-list-item>
-        <v-list-item @click="openPicker">
-          <template #prepend>
-            <v-icon size="small">mdi-playlist-plus</v-icon>
-          </template>
-          <v-list-item-title>Add to existing series…</v-list-item-title>
-        </v-list-item>
-      </v-list>
-    </v-menu>
-
-    <v-dialog v-model="pickerOpen" max-width="460">
-      <v-card>
-        <v-card-title>Add to series</v-card-title>
-        <v-card-text>
-          <p class="text-caption mb-2">
-            Appends this post as the next chapter of the chosen series.
-          </p>
-          <v-autocomplete
-            v-model="picked"
-            :items="hits"
-            :item-title="(s) => s.name"
-            :item-value="(s) => s.id"
-            :loading="searching"
-            label="Series" density="compact" autofocus clearable
-          />
-        </v-card-text>
-        <v-card-actions>
-          <v-spacer />
-          <v-btn @click="pickerOpen = false">Cancel</v-btn>
-          <v-btn
-            color="primary" rounded="pill"
-            :disabled="picked == null || busy" @click="onAddExisting"
-          >Add</v-btn>
-        </v-card-actions>
-      </v-card>
-    </v-dialog>
-  </span>
-</template>
-
-<script setup>
-import { ref } from 'vue'
-import { useRouter } from 'vue-router'
-import { useApi } from '../../composables/useApi.js'
-import { toast } from '../../utils/toast.js'
-
-const props = defineProps({ post: { type: Object, required: true } })
-
-const api = useApi()
-const router = useRouter()
-
-const menuOpen = ref(false)
-const busy = ref(false)
-const pickerOpen = ref(false)
-const picked = ref(null)
-const hits = ref([])
-const searching = ref(false)
-
-async function onPromote() {
-  menuOpen.value = false
-  busy.value = true
-  try {
-    const out = await api.post('/api/series/from-post', {
-      body: { post_id: props.post.id }
-    })
-    toast({ text: `Series created: ${out.name}`, type: 'success' })
-    router.push({ name: 'series-manage', params: { tagId: out.series_tag_id } })
-  } catch (e) {
-    toast({ text: `Could not create series: ${e.message}`, type: 'error' })
-  } finally {
-    busy.value = false
-  }
-}
-
-async function openPicker() {
-  menuOpen.value = false
-  picked.value = null
-  hits.value = []
-  pickerOpen.value = true
-  // Load the full series list so the picker is browsable (the tag autocomplete
-  // returns nothing for an empty query — the #712 trap). v-autocomplete filters
-  // client-side by name as the operator types.
-  searching.value = true
-  try {
-    const body = await api.get('/api/series', { params: { sort: 'name' } })
-    hits.value = body.series || []
-  } catch (e) {
-    toast({ text: `Could not load series: ${e.message}`, type: 'error' })
-  } finally {
-    searching.value = false
-  }
-}
-
-async function onAddExisting() {
-  if (picked.value == null) return
-  busy.value = true
-  try {
-    await api.post(`/api/series/${picked.value}/add-post`, {
-      body: { post_id: props.post.id }
-    })
-    pickerOpen.value = false
-    toast({ text: 'Added to series', type: 'success' })
-  } catch (e) {
-    toast({ text: `Could not add to series: ${e.message}`, type: 'error' })
-  } finally {
-    busy.value = false
-  }
-}
-</script>

frontend/src/components/settings/ArchiveReextractCard.vue

@@ -1,46 +0,0 @@
-<template>
-  <!-- #713: re-extract PostAttachments that are really archives but were filed
-       opaquely before the magic-byte gate, and attach their images to the post. -->
-  <v-card>
-    <v-card-title>Re-extract archive attachments</v-card-title>
-    <v-card-text>
-      <p class="text-body-2 mb-3">
-        Some posts attached an archive (zip) whose images weren't extracted
-        because the downloaded file had no usable extension. This scans existing
-        attachments, extracts any that are really archives, and attaches their
-        images to the post. Idempotent — safe to run more than once.
-      </p>
-      <v-btn color="primary" rounded="pill" :loading="busy" @click="run">
-        <v-icon start>mdi-folder-zip-outline</v-icon> Re-extract archives now
-      </v-btn>
-      <span v-if="queued" class="ml-3 text-caption text-success">Queued ✓</span>
-      <QueueStatusBar queue="maintenance" queue-label="Maintenance" />
-    </v-card-text>
-  </v-card>
-</template>
-
-<script setup>
-import { ref } from 'vue'
-
-import { useApi } from '../../composables/useApi.js'
-import { toast } from '../../utils/toast.js'
-import QueueStatusBar from './QueueStatusBar.vue'
-
-const api = useApi()
-const busy = ref(false)
-const queued = ref(false)
-
-async function run () {
-  busy.value = true
-  queued.value = false
-  try {
-    await api.post('/api/admin/maintenance/reextract-archives')
-    queued.value = true
-    toast({ text: 'Archive re-extract queued', type: 'success' })
-  } catch (e) {
-    toast({ text: e?.body?.detail || e?.message || 'Failed to queue', type: 'error' })
-  } finally {
-    busy.value = false
-  }
-}
-</script>

frontend/src/components/settings/MaintenancePanel.vue

@@ -15,7 +15,6 @@
     <AllowlistTable class="mt-4" />
     <AliasTable class="mt-4" />
     <DbMaintenanceCard class="mt-6" />
-    <ArchiveReextractCard class="mt-6" />
     <BackupCard class="mt-6" />
     <!-- TagMaintenanceCard moved to Cleanup tab (v26.05.25.7) — it
          operates on the existing library which fits the Cleanup-tab
@@ -34,7 +33,6 @@ import MLThresholdSliders from './MLThresholdSliders.vue'
 import AllowlistTable from './AllowlistTable.vue'
 import AliasTable from './AliasTable.vue'
 import DbMaintenanceCard from './DbMaintenanceCard.vue'
-import ArchiveReextractCard from './ArchiveReextractCard.vue'
 import BackupCard from './BackupCard.vue'
 import { useSystemActivityStore } from '../../stores/systemActivity.js'
 

frontend/src/components/settings/TagMaintenanceCard.vue

@@ -136,60 +136,6 @@
         >Delete {{ resetPreview.count }} content tag(s) +
           {{ resetPreview.applications }} application(s)</v-btn>
       </div>
-
-      <v-divider class="my-5" />
-
-      <!-- #714: standardize existing tags to the canonical Title-Case form
-           and merge case/whitespace-variant duplicates. -->
-      <p class="text-body-2 mb-2">
-        <strong>Standardize tag casing.</strong>
-        New tags are saved Title Case, but older tags keep whatever casing they
-        were created with. This renames every existing tag to
-        <code>Title Case</code> (collapsing extra spaces) and
-        <strong>merges</strong> tags that differ only by case or spacing
-        (e.g. <code>hatsune miku</code> + <code>Hatsune&nbsp;&nbsp;Miku</code>)
-        into one — repointing their images, allowlist entries and series pages.
-      </p>
-      <v-alert type="warning" variant="tonal" density="compact" class="mb-3">
-        The merges are irreversible — back up first
-        (Settings → Maintenance → Backup). Safe to run more than once.
-      </v-alert>
-
-      <v-btn
-        color="accent" variant="flat" rounded="pill"
-        prepend-icon="mdi-magnify"
-        :loading="loadingNormPreview"
-        class="mb-3"
-        @click="onNormPreview"
-      >Preview tag standardization</v-btn>
-
-      <div v-if="normPreview">
-        <p class="text-body-2 mb-2">
-          <strong>{{ normPreview.total_changes }}</strong> tag group(s) to
-          change — <strong>{{ normPreview.tags_to_rename }}</strong> rename(s),
-          <strong>{{ normPreview.collisions }}</strong> collision(s) merging
-          <strong>{{ normPreview.tags_to_merge }}</strong> tag(s) away.
-        </p>
-        <div v-if="normPreview.sample?.length" class="fc-name-grid mb-3">
-          <span
-            v-for="s in normPreview.sample" :key="s.to + s.kind"
-            class="fc-name"
-          >
-            <template v-if="s.merge">{{ s.from.join(' + ') }} → </template>{{ s.to }}
-          </span>
-        </div>
-        <v-btn
-          color="error" variant="flat" rounded="pill"
-          prepend-icon="mdi-format-letter-case"
-          :disabled="!normPreview.total_changes || normResult === 'queued'"
-          :loading="normCommitting"
-          @click="onNormCommit"
-        >Standardize {{ normPreview.total_changes }} tag group(s)</v-btn>
-        <span v-if="normResult === 'queued'" class="ml-3 text-caption text-success">
-          Queued ✓ — runs in the background (you can leave this page). It
-          processes in chunks, so re-run “Preview” later to confirm it's all done.
-        </span>
-      </div>
     </v-card-text>
   </v-card>
 </template>
@@ -209,10 +155,6 @@ const kindCommitting = ref(false)
 const resetPreview = ref(null)
 const loadingResetPreview = ref(false)
 const resetCommitting = ref(false)
-const normPreview = ref(null)
-const loadingNormPreview = ref(false)
-const normCommitting = ref(false)
-const normResult = ref(null)
 
 async function onPreview() {
   loadingPreview.value = true
@@ -270,36 +212,6 @@ async function onResetCommit() {
     resetCommitting.value = false
   }
 }
-
-async function onNormPreview() {
-  loadingNormPreview.value = true
-  normResult.value = null
-  try {
-    normPreview.value = await store.normalizeTags({ dryRun: true })
-  } finally {
-    loadingNormPreview.value = false
-  }
-}
-
-async function onNormCommit() {
-  normCommitting.value = true
-  normResult.value = null
-  try {
-    // Fire-and-forget: the task is time-boxed and self-resuming across chunks
-    // (a large back-catalog can't finish in one run), so we DON'T poll-until-
-    // done — that would falsely report "complete" after the first chunk. Just
-    // confirm it's queued; the operator can re-run Preview later to verify.
-    await store.normalizeTags({ dryRun: false })
-    normResult.value = 'queued'
-    // Keep the preview showing what was QUEUED — do NOT zero it out. Overwriting
-    // it with a zeroed object made the card read "0 tag groups to change" the
-    // instant Standardize was clicked, which looked like nothing happened
-    // (operator-flagged 2026-06-07). The button disables on `queued`; re-running
-    // Preview later reflects the real remaining count as the task applies.
-  } finally {
-    normCommitting.value = false
-  }
-}
 </script>
 
 <style scoped>

frontend/src/components/subscriptions/ActiveDownloadsPanel.vue

@@ -23,18 +23,6 @@
         <span class="fc-active__dot" />
         <PlatformChip :platform="e.platform" size="x-small" />
         <span class="fc-active__artist">{{ e.artist_name || '—' }}</span>
-        <!-- #709: live per-file progress for native (Patreon) downloads. -->
-        <span v-if="e.live" class="fc-active__counts">
-          <span class="fc-active__count" title="downloaded">↓ {{ e.live.downloaded }}</span>
-          <span v-if="e.live.skipped" class="fc-active__count" title="skipped">
-            ⤼ {{ e.live.skipped }}</span>
-          <span
-            v-if="e.live.errors" class="fc-active__count fc-active__count--err"
-            title="errors"
-          >✕ {{ e.live.errors }}</span>
-          <span class="fc-active__count fc-active__count--posts" title="posts scanned">
-            {{ e.live.posts }} posts</span>
-        </span>
         <v-spacer />
         <span class="fc-active__timer" :title="`started ${e.started_at}`">
           {{ elapsed(e.started_at) }}
@@ -135,13 +123,6 @@ function elapsed (startedIso) {
   font-size: 0.78rem; text-transform: uppercase; letter-spacing: 0.05em;
   color: rgb(var(--v-theme-on-surface-variant));
 }
-.fc-active__counts {
-  display: inline-flex; align-items: center; gap: 8px;
-  font-variant-numeric: tabular-nums; font-size: 0.78rem;
-}
-.fc-active__count { color: rgb(var(--v-theme-on-surface-variant)); }
-.fc-active__count--err { color: rgb(var(--v-theme-error)); }
-.fc-active__count--posts { opacity: 0.7; }
 @keyframes fc-active-pulse {
   0%, 100% { opacity: 1; transform: scale(1); }
   50% { opacity: 0.35; transform: scale(0.7); }

frontend/src/components/subscriptions/FailingSourcesCard.vue

@@ -93,7 +93,6 @@ const ERROR_TYPE_COLOR = {
   unsupported_url: 'error',
   http_error: 'error',
   unknown_error: 'error',
-  api_drift: 'error',
   partial: 'info',
   tier_limited: 'info',
   no_new_content: 'info',
@@ -109,7 +108,6 @@ const ERROR_TYPE_HINT = {
   http_error: 'Generic HTTP error — see Logs.',
   unsupported_url: 'gallery-dl does not support this URL pattern.',
   unknown_error: 'Could not classify — see Logs.',
-  api_drift: 'Patreon changed its API shape — the native ingester needs a code update.',
 }
 function errorTypeColor(t) { return ERROR_TYPE_COLOR[t] || 'error' }
 function errorTypeHint(t) { return ERROR_TYPE_HINT[t] || '' }

frontend/src/components/subscriptions/PreviewDialog.vue

@@ -1,113 +0,0 @@
-<template>
-  <!-- Plan #708 B4: dry-run preview — what a backfill WOULD download, without
-       downloading. Self-fetches on open; loading / error / empty / result. -->
-  <v-dialog
-    :model-value="modelValue" max-width="520"
-    @update:model-value="$emit('update:modelValue', $event)"
-  >
-    <v-card v-if="source">
-      <v-card-title class="d-flex align-center">
-        <v-icon class="mr-2">mdi-eye-outline</v-icon>
-        Preview · {{ source.artist_name }}
-      </v-card-title>
-      <v-card-subtitle class="fc-preview__url">{{ source.url }}</v-card-subtitle>
-
-      <v-card-text>
-        <div v-if="loading" class="text-center py-8">
-          <v-progress-circular indeterminate color="accent" />
-          <div class="text-caption mt-3 text-medium-emphasis">Walking the feed…</div>
-        </div>
-
-        <v-alert
-          v-else-if="error" type="error" variant="tonal" density="comfortable"
-        >{{ error }}</v-alert>
-
-        <div v-else-if="result">
-          <div class="text-h5">
-            {{ result.total_new }} new item{{ result.total_new === 1 ? '' : 's' }}
-          </div>
-          <div class="text-body-2 text-medium-emphasis mb-3">
-            in the {{ result.posts_scanned }} most recent
-            post{{ result.posts_scanned === 1 ? '' : 's' }}<span
-              v-if="result.has_more"> · more pages not scanned</span>
-          </div>
-
-          <div
-            v-if="result.total_new === 0"
-            class="text-body-2 text-medium-emphasis"
-          >You’re caught up — nothing new to download.</div>
-
-          <v-list v-else density="compact" class="fc-preview__list" bg-color="transparent">
-            <v-list-item v-for="(row, i) in result.sample" :key="i" class="px-0">
-              <template #prepend>
-                <v-chip
-                  size="x-small" color="info" variant="tonal" label class="mr-3"
-                >+{{ row.new }}</v-chip>
-              </template>
-              <v-list-item-title class="text-body-2">{{ row.title }}</v-list-item-title>
-              <v-list-item-subtitle v-if="row.date" class="text-caption">
-                {{ formatRelative(row.date) }}
-              </v-list-item-subtitle>
-            </v-list-item>
-          </v-list>
-        </div>
-      </v-card-text>
-
-      <v-card-actions>
-        <v-spacer />
-        <v-btn variant="text" @click="$emit('update:modelValue', false)">Close</v-btn>
-        <v-btn
-          v-if="result && result.total_new > 0"
-          color="accent" variant="flat"
-          @click="$emit('backfill', source)"
-        >Start backfill</v-btn>
-      </v-card-actions>
-    </v-card>
-  </v-dialog>
-</template>
-
-<script setup>
-import { ref, watch } from 'vue'
-import { useSourcesStore } from '../../stores/sources.js'
-import { formatRelative } from '../../utils/date.js'
-
-const props = defineProps({
-  modelValue: { type: Boolean, default: false },
-  source: { type: Object, default: null },
-})
-defineEmits(['update:modelValue', 'backfill'])
-
-const store = useSourcesStore()
-const loading = ref(false)
-const error = ref('')
-const result = ref(null)
-
-watch(
-  () => props.modelValue,
-  async (open) => {
-    if (!open || !props.source) return
-    loading.value = true
-    error.value = ''
-    result.value = null
-    try {
-      result.value = await store.previewSource(props.source.id)
-    } catch (e) {
-      error.value = e?.body?.detail || e?.message || 'Preview failed'
-    } finally {
-      loading.value = false
-    }
-  },
-)
-</script>
-
-<style scoped>
-.fc-preview__url {
-  overflow: hidden;
-  text-overflow: ellipsis;
-  white-space: nowrap;
-}
-.fc-preview__list {
-  max-height: 320px;
-  overflow-y: auto;
-}
-</style>

frontend/src/components/subscriptions/SourceActions.vue

@@ -1,81 +0,0 @@
-<template>
-  <!-- Shared per-source action cluster (desktop SourceRow + mobile SourceCard).
-       Frequent actions stay as buttons; low-frequency / destructive ones move
-       into a labelled overflow menu so they can't be mis-clicked, and the
-       labels spell out what each one does (recover vs backfill). -->
-  <div class="fc-src-actions">
-    <v-btn
-      icon size="small" variant="text" :loading="checking"
-      @click.stop="emit('check', source)"
-    >
-      <v-icon>mdi-play</v-icon>
-      <v-tooltip activator="parent" location="top">
-        Check now — fetch posts published since the last check
-      </v-tooltip>
-    </v-btn>
-    <v-btn
-      icon size="small" variant="text"
-      :color="running ? 'warning' : undefined"
-      @click.stop="emit('backfill', source)"
-    >
-      <v-icon>{{ running ? 'mdi-stop' : 'mdi-magnify-scan' }}</v-icon>
-      <v-tooltip activator="parent" location="top">
-        {{ running
-          ? (recovering ? 'Stop recovery' : 'Stop backfill')
-          : 'Backfill — one-time walk of the FULL post history until complete' }}
-      </v-tooltip>
-    </v-btn>
-
-    <v-menu location="bottom end">
-      <template #activator="{ props: menuProps }">
-        <v-btn icon size="small" variant="text" v-bind="menuProps" @click.stop>
-          <v-icon>mdi-dots-vertical</v-icon>
-          <v-tooltip activator="parent" location="top">More actions</v-tooltip>
-        </v-btn>
-      </template>
-      <v-list density="compact" min-width="260">
-        <v-list-item
-          v-if="isPatreon && !running"
-          prepend-icon="mdi-backup-restore"
-          @click="emit('recover', source)"
-        >
-          <v-list-item-title>Recover dropped near-duplicates</v-list-item-title>
-          <v-list-item-subtitle>
-            Re-fetch images previously dropped as near-dups and re-judge them
-            under the current similarity threshold
-          </v-list-item-subtitle>
-        </v-list-item>
-        <v-divider v-if="isPatreon && !running" />
-        <v-list-item
-          base-color="error"
-          prepend-icon="mdi-delete-outline"
-          @click="emit('remove', source)"
-        >
-          <v-list-item-title>Remove source</v-list-item-title>
-        </v-list-item>
-      </v-list>
-    </v-menu>
-  </div>
-</template>
-
-<script setup>
-import { computed } from 'vue'
-
-const props = defineProps({
-  source: { type: Object, required: true },
-  checking: { type: Boolean, default: false },
-})
-const emit = defineEmits(['check', 'backfill', 'recover', 'remove'])
-
-const running = computed(() => props.source.backfill_state === 'running')
-const recovering = computed(() => !!props.source.backfill_bypass_seen)
-const isPatreon = computed(() => props.source.platform === 'patreon')
-</script>
-
-<style scoped>
-.fc-src-actions {
-  display: inline-flex;
-  align-items: center;
-  gap: 2px;
-}
-</style>

frontend/src/components/subscriptions/SourceCard.vue

@@ -12,19 +12,10 @@
       />
     </div>
 
-    <div class="fc-source-card__url-row">
-      <a
-        :href="source.url" target="_blank" rel="noopener"
-        class="fc-source-card__url" @click.stop
-      >{{ source.url }}</a>
-      <v-btn
-        icon="mdi-pencil" size="x-small" variant="text"
-        @click.stop="$emit('edit', source)"
-      >
-        <v-icon>mdi-pencil</v-icon>
-        <v-tooltip activator="parent" location="top">Edit source</v-tooltip>
-      </v-btn>
-    </div>
+    <a
+      :href="source.url" target="_blank" rel="noopener"
+      class="fc-source-card__url" @click.stop
+    >{{ source.url }}</a>
 
     <div class="fc-source-card__meta">
       <span>Last {{ formatRelative(source.last_checked_at) }}</span>
@@ -32,42 +23,45 @@
       <v-chip
         v-if="(source.consecutive_failures || 0) > 0"
         size="x-small" color="error" variant="tonal" label
-      >{{ source.consecutive_failures }} err
-        <v-tooltip v-if="source.last_error" activator="parent" location="top" max-width="480">
-          <span style="white-space: pre-wrap; word-break: break-word;">{{ source.last_error }}</span>
-        </v-tooltip>
-      </v-chip>
+      >{{ source.consecutive_failures }} err</v-chip>
       <v-chip
-        v-else-if="source.backfill_state === 'running'"
+        v-else-if="(source.backfill_runs_remaining || 0) > 0"
         size="x-small" color="info" variant="tonal" label
-      >{{ source.backfill_bypass_seen ? 'Recovering' : 'Backfilling'
-        }}{{ source.backfill_posts
-              ? ` · ${source.backfill_posts} posts`
-              : (source.backfill_chunks ? ` (${source.backfill_chunks})` : '') }}</v-chip>
-      <v-chip
-        v-else-if="source.backfill_state === 'complete'"
-        size="x-small" color="success" variant="tonal" label
-      >Backfilled</v-chip>
-      <v-chip
-        v-else-if="source.backfill_state === 'stalled'"
-        size="x-small" color="warning" variant="tonal" label
-      >Stalled</v-chip>
+      >backfill ({{ source.backfill_runs_remaining }}×)</v-chip>
     </div>
 
     <div class="fc-source-card__actions">
-      <SourceActions
-        :source="source" :checking="checking"
-        @check="$emit('check', $event)"
-        @backfill="$emit('backfill', $event)"
-        @recover="$emit('recover', $event)"
-        @remove="$emit('remove', $event)"
-      />
+      <v-btn
+        size="x-small" variant="text" :loading="checking"
+        @click.stop="$emit('check', source)"
+      >
+        <v-icon>mdi-play</v-icon>
+        <v-tooltip activator="parent" location="top">Check now</v-tooltip>
+      </v-btn>
+      <v-btn
+        size="x-small" variant="text"
+        :disabled="(source.backfill_runs_remaining || 0) > 0"
+        @click.stop="$emit('backfill', source)"
+      >
+        <v-icon>mdi-magnify-scan</v-icon>
+        <v-tooltip activator="parent" location="top">Deep scan</v-tooltip>
+      </v-btn>
+      <v-btn size="x-small" variant="text" @click.stop="$emit('edit', source)">
+        <v-icon>mdi-pencil</v-icon>
+        <v-tooltip activator="parent" location="top">Edit</v-tooltip>
+      </v-btn>
+      <v-btn
+        size="x-small" variant="text" color="error"
+        @click.stop="$emit('remove', source)"
+      >
+        <v-icon>mdi-close</v-icon>
+        <v-tooltip activator="parent" location="top">Remove</v-tooltip>
+      </v-btn>
     </div>
   </div>
 </template>
 
 <script setup>
-import SourceActions from './SourceActions.vue'
 import SourceHealthDot from './SourceHealthDot.vue'
 import { formatRelative } from '../../utils/date.js'
 
@@ -79,7 +73,7 @@ const props = defineProps({
   checking: { type: Boolean, default: false },
   warningThreshold: { type: Number, default: 5 },
 })
-const emit = defineEmits(['edit', 'remove', 'toggle', 'check', 'backfill', 'recover'])
+const emit = defineEmits(['edit', 'remove', 'toggle', 'check', 'backfill'])
 
 function onToggleEnabled(value) {
   emit('toggle', { source: props.source, enabled: value })
@@ -95,15 +89,11 @@ function onToggleEnabled(value) {
   display: flex; flex-direction: column; gap: 6px;
 }
 .fc-source-card__top { display: flex; align-items: center; gap: 8px; }
-.fc-source-card__url-row {
-  display: flex; align-items: center; gap: 4px;
-}
 .fc-source-card__url {
   color: rgb(var(--v-theme-on-surface-variant));
   text-decoration: none;
   font-size: 0.8rem;
   word-break: break-all;
-  flex: 1 1 auto; min-width: 0;
 }
 .fc-source-card__url:hover { color: rgb(var(--v-theme-accent)); }
 .fc-source-card__meta {

frontend/src/components/subscriptions/SourceRow.vue

@@ -7,22 +7,10 @@
       <v-chip size="x-small" variant="tonal" label>{{ source.platform }}</v-chip>
     </td>
     <td class="fc-source-row__url-cell">
-      <div class="fc-source-row__url-wrap">
-        <a :href="source.url" target="_blank" rel="noopener" class="fc-source-row__url"
-           @click.stop>
-          {{ source.url }}
-        </a>
-        <!-- Edit sits next to the source identity (operator-requested), not in
-             the action cluster where it was easy to fat-finger Remove. -->
-        <v-btn
-          icon="mdi-pencil" size="x-small" variant="text"
-          class="fc-source-row__edit"
-          @click.stop="$emit('edit', source)"
-        >
-          <v-icon>mdi-pencil</v-icon>
-          <v-tooltip activator="parent" location="top">Edit source</v-tooltip>
-        </v-btn>
-      </div>
+      <a :href="source.url" target="_blank" rel="noopener" class="fc-source-row__url"
+         @click.stop>
+        {{ source.url }}
+      </a>
     </td>
     <td>
       <v-switch
@@ -42,43 +30,53 @@
       <v-chip
         v-if="(source.consecutive_failures || 0) > 0"
         size="x-small" color="error" variant="tonal" label
-      >{{ source.consecutive_failures }}
-        <!-- #1: show the actual failure reason on hover instead of a bare count. -->
-        <v-tooltip v-if="source.last_error" activator="parent" location="top" max-width="480">
-          <span class="fc-source-row__err-text">{{ source.last_error }}</span>
-        </v-tooltip>
-      </v-chip>
+      >{{ source.consecutive_failures }}</v-chip>
       <v-chip
-        v-else-if="source.backfill_state === 'running'"
+        v-else-if="(source.backfill_runs_remaining || 0) > 0"
         size="x-small" color="info" variant="tonal" label
-      >{{ source.backfill_bypass_seen ? 'Recovering' : 'Backfilling'
-        }}{{ source.backfill_posts
-              ? ` · ${source.backfill_posts} posts`
-              : (source.backfill_chunks ? ` (${source.backfill_chunks})` : '') }}</v-chip>
-      <v-chip
-        v-else-if="source.backfill_state === 'complete'"
-        size="x-small" color="success" variant="tonal" label
-      >Backfilled</v-chip>
-      <v-chip
-        v-else-if="source.backfill_state === 'stalled'"
-        size="x-small" color="warning" variant="tonal" label
-      >Stalled</v-chip>
+      >
+        backfill ({{ source.backfill_runs_remaining }}×)
+      </v-chip>
       <span v-else class="fc-source-row__zero">0</span>
     </td>
     <td class="fc-source-row__actions">
-      <SourceActions
-        :source="source" :checking="checking"
-        @check="$emit('check', $event)"
-        @backfill="$emit('backfill', $event)"
-        @recover="$emit('recover', $event)"
-        @remove="$emit('remove', $event)"
-      />
+      <v-btn
+        icon="mdi-play" size="x-small" variant="text"
+        :loading="checking"
+        @click.stop="$emit('check', source)"
+      >
+        <v-icon>mdi-play</v-icon>
+        <v-tooltip activator="parent" location="top">Check now</v-tooltip>
+      </v-btn>
+      <v-btn
+        icon="mdi-magnify-scan" size="x-small" variant="text"
+        :disabled="(source.backfill_runs_remaining || 0) > 0"
+        @click.stop="$emit('backfill', source)"
+      >
+        <v-icon>mdi-magnify-scan</v-icon>
+        <v-tooltip activator="parent" location="top">
+          Deep scan — walk full history for next few runs
+        </v-tooltip>
+      </v-btn>
+      <v-btn
+        icon="mdi-pencil" size="x-small" variant="text"
+        @click.stop="$emit('edit', source)"
+      >
+        <v-icon>mdi-pencil</v-icon>
+        <v-tooltip activator="parent" location="top">Edit</v-tooltip>
+      </v-btn>
+      <v-btn
+        icon="mdi-close" size="x-small" variant="text" color="error"
+        @click.stop="$emit('remove', source)"
+      >
+        <v-icon>mdi-close</v-icon>
+        <v-tooltip activator="parent" location="top">Remove</v-tooltip>
+      </v-btn>
     </td>
   </tr>
 </template>
 
 <script setup>
-import SourceActions from './SourceActions.vue'
 import SourceHealthDot from './SourceHealthDot.vue'
 import { formatRelative } from '../../utils/date.js'
 
@@ -87,7 +85,7 @@ const props = defineProps({
   checking: { type: Boolean, default: false },
   warningThreshold: { type: Number, default: 5 },
 })
-const emit = defineEmits(['edit', 'remove', 'toggle', 'check', 'backfill', 'recover'])
+const emit = defineEmits(['edit', 'remove', 'toggle', 'check', 'backfill'])
 
 function onToggleEnabled(value) {
   emit('toggle', { source: props.source, enabled: value })
@@ -100,27 +98,18 @@ function onToggleEnabled(value) {
   padding-right: 0 !important;
 }
 .fc-source-row__url-cell {
-  max-width: 420px;
+  max-width: 400px;
   overflow: hidden;
 }
-.fc-source-row__url-wrap {
-  display: flex;
-  align-items: center;
-  gap: 2px;
-  min-width: 0;
-}
 .fc-source-row__url {
   color: rgb(var(--v-theme-on-surface-variant));
   text-decoration: none;
-  flex: 0 1 auto;
-  min-width: 0;
+  display: inline-block;
+  max-width: 100%;
   overflow: hidden;
   text-overflow: ellipsis;
   white-space: nowrap;
 }
-/* Keep the edit affordance subtle until the row is hovered. */
-.fc-source-row__edit { opacity: 0.5; flex: 0 0 auto; }
-.fc-source-row:hover .fc-source-row__edit { opacity: 1; }
 .fc-source-row__url:hover { color: rgb(var(--v-theme-accent)); }
 .fc-source-row__when {
   color: rgb(var(--v-theme-on-surface-variant));
@@ -131,10 +120,6 @@ function onToggleEnabled(value) {
   color: rgb(var(--v-theme-on-surface-variant));
   opacity: 0.6;
 }
-.fc-source-row__err-text {
-  white-space: pre-wrap;
-  word-break: break-word;
-}
 .fc-source-row__actions {
   white-space: nowrap;
   text-align: right;

frontend/src/components/subscriptions/SubscriptionsTab.vue

@@ -54,9 +54,6 @@
           <v-btn size="small" variant="text" prepend-icon="mdi-toggle-switch-off-outline" @click="bulkSetEnabled(false)">
             Disable all
           </v-btn>
-          <v-btn size="small" variant="text" prepend-icon="mdi-magnify-scan" @click="bulkBackfill">
-            Backfill
-          </v-btn>
           <v-btn size="small" variant="text" color="error" prepend-icon="mdi-delete" @click="bulkDelete">
             Delete
           </v-btn>
@@ -85,7 +82,6 @@
         item-value="key"
         v-model="selected"
         v-model:expanded="expanded"
-        v-model:sort-by="sortBy"
         :items-per-page="50"
         :items-per-page-options="ITEMS_PER_PAGE_OPTIONS"
         density="comfortable"
@@ -93,13 +89,7 @@
         @click:row="onRowClick"
       >
         <template #item.name="{ item }">
-          <div class="fc-subs__name">{{ item.artist.name }}</div>
-          <!-- #2: single-source subscriptions show their URL inline (no expand). -->
-          <a
-            v-if="item.singleSource"
-            :href="item.singleSource.url" target="_blank" rel="noopener"
-            class="fc-subs__sub-url" @click.stop
-          >{{ item.singleSource.url }}</a>
+          <span class="fc-subs__name">{{ item.artist.name }}</span>
         </template>
 
         <template #item.platforms="{ item }">
@@ -131,60 +121,32 @@
         </template>
 
         <template #item.actions="{ item }">
-          <!-- #2: single source → its own actions inline (Check / Backfill / ⋮),
-               plus Edit, so the row is actionable without expanding. -->
-          <template v-if="item.singleSource">
-            <SourceActions
-              :source="item.singleSource"
-              :checking="store.checkingIds.has(item.singleSource.id)"
-              @check="onCheck" @backfill="onBackfill"
-              @recover="onRecover" @remove="removeSource"
-            />
-            <v-btn icon size="small" variant="text" @click.stop="openEditSource(item.singleSource)">
-              <v-icon>mdi-pencil</v-icon>
-              <v-tooltip activator="parent" location="top">Edit source</v-tooltip>
-            </v-btn>
-            <v-btn icon size="small" variant="text" @click.stop="openAddSource(item.artist)">
-              <v-icon>mdi-plus</v-icon>
-              <v-tooltip activator="parent" location="top">Add another source</v-tooltip>
-            </v-btn>
-            <v-btn
-              icon size="small" variant="text"
-              :to="`/artist/${item.artist.slug}`" @click.stop
-            >
-              <v-icon>mdi-account</v-icon>
-              <v-tooltip activator="parent" location="top">Open artist page</v-tooltip>
-            </v-btn>
-          </template>
-          <!-- Multi-source → artist-level actions; expand for per-source rows. -->
-          <template v-else>
-            <v-btn
-              icon size="small" variant="text"
-              :loading="anyChecking(item.sources)"
-              @click.stop="checkAll(item)"
-            >
-              <v-icon>mdi-refresh</v-icon>
-              <v-tooltip activator="parent" location="top">Check all sources</v-tooltip>
-            </v-btn>
-            <v-btn icon size="small" variant="text" @click.stop="openAddSource(item.artist)">
-              <v-icon>mdi-plus</v-icon>
-              <v-tooltip activator="parent" location="top">Add source</v-tooltip>
-            </v-btn>
-            <v-btn
-              icon size="small" variant="text"
-              :to="`/posts?artist_id=${item.artist.id}`" @click.stop
-            >
-              <v-icon>mdi-rss</v-icon>
-              <v-tooltip activator="parent" location="top">View posts</v-tooltip>
-            </v-btn>
-            <v-btn
-              icon size="small" variant="text"
-              :to="`/artist/${item.artist.slug}`" @click.stop
-            >
-              <v-icon>mdi-account</v-icon>
-              <v-tooltip activator="parent" location="top">Open artist page</v-tooltip>
-            </v-btn>
-          </template>
+          <v-btn
+            icon size="small" variant="text"
+            :loading="anyChecking(item.sources)"
+            @click.stop="checkAll(item)"
+          >
+            <v-icon>mdi-refresh</v-icon>
+            <v-tooltip activator="parent" location="top">Check all sources</v-tooltip>
+          </v-btn>
+          <v-btn icon size="small" variant="text" @click.stop="openAddSource(item.artist)">
+            <v-icon>mdi-plus</v-icon>
+            <v-tooltip activator="parent" location="top">Add source</v-tooltip>
+          </v-btn>
+          <v-btn
+            icon size="small" variant="text"
+            :to="`/posts?artist_id=${item.artist.id}`" @click.stop
+          >
+            <v-icon>mdi-rss</v-icon>
+            <v-tooltip activator="parent" location="top">View posts</v-tooltip>
+          </v-btn>
+          <v-btn
+            icon size="small" variant="text"
+            :to="`/artist/${item.artist.slug}`" @click.stop
+          >
+            <v-icon>mdi-account</v-icon>
+            <v-tooltip activator="parent" location="top">Open artist page</v-tooltip>
+          </v-btn>
         </template>
 
         <template #expanded-row="{ columns, item }">
@@ -213,7 +175,6 @@
                     @toggle="toggleSourceEnabled"
                     @check="onCheck"
                     @backfill="onBackfill"
-                    @recover="onRecover"
                   />
                   <tr v-if="item.sources.length === 0">
                     <td colspan="8" class="fc-subs__sources-empty">
@@ -290,7 +251,6 @@
             @toggle="toggleSourceEnabled"
             @check="onCheck"
             @backfill="onBackfill"
-            @recover="onRecover"
           />
           <div v-if="item.sources.length === 0" class="fc-subs__sources-empty">
             No sources yet. Tap + to add one.
@@ -322,7 +282,6 @@ import SourceCard from './SourceCard.vue'
 import SourceHealthDot from './SourceHealthDot.vue'
 import SourceFormDialog from './SourceFormDialog.vue'
 import ArtistCreateDialog from './ArtistCreateDialog.vue'
-import SourceActions from './SourceActions.vue'
 import PlatformChip from './PlatformChip.vue'
 import SchedulerStatusBar from './SchedulerStatusBar.vue'
 import { formatRelative } from '../../utils/date.js'
@@ -356,12 +315,6 @@ const statusFilter = ref('all')
 const needsAttention = ref(false)
 const expanded = ref([])
 const selected = ref([])
-// #3: default to worst-health first, name as the tiebreak (the table header
-// stays clickable to re-sort by any column).
-const sortBy = ref([
-  { key: 'health', order: 'desc' },
-  { key: 'name', order: 'asc' },
-])
 
 // Mobile card list drives the same `selected`/`expanded` key arrays the
 // desktop v-data-table binds, so selection + bulk actions work identically.
@@ -417,7 +370,7 @@ const headers = [
   { title: 'Subscription', key: 'name',          sortable: true,  align: 'start' },
   { title: 'Platforms',    key: 'platforms',     sortable: false, align: 'start', width: 240 },
   { title: 'Sources',      key: 'sources_count', sortable: true,  align: 'start', width: 90 },
-  { title: 'Health',       key: 'health',        sortable: true,  align: 'start', width: 80 },
+  { title: 'Health',       key: 'health',        sortable: false, align: 'start', width: 80 },
   { title: 'Last activity',key: 'last_activity', sortable: true,  align: 'start', width: 140 },
   { title: 'Actions',      key: 'actions',       sortable: false, align: 'end',   width: 200 },
 ]
@@ -438,26 +391,10 @@ const groups = computed(() => {
       lastActivity,
       name: g.artist.name,
       last_activity: lastActivity ?? '',
-      // #3: numeric health rank so the Health column is sortable worst-first.
-      health: healthRank(worstSource, failureThreshold.value),
-      // #2: when a subscription has exactly one source, surface it on the
-      // artist row directly (URL + actions) so the common case needs no expand.
-      singleSource: g.sources.length === 1 ? g.sources[0] : null,
     }
   })
 })
 
-// 0 never-checked · 1 ok · 2 warning · 3 failing — higher = worse, so a
-// descending sort floats the broken subscriptions to the top.
-function healthRank(source, threshold) {
-  if (!source) return 0
-  if (!source.last_checked_at) return 0
-  const f = source.consecutive_failures || 0
-  if (f === 0) return 1
-  if (f < threshold) return 2
-  return 3
-}
-
 // A group "needs attention" if any of its sources has accumulated
 // failures OR has never been checked — the ones worth acting on.
 function groupNeedsAttention(g) {
@@ -542,26 +479,13 @@ function openEditSource(source) {
 }
 
 async function removeSource(source) {
-  // Confirm — Remove moved into the overflow menu, but a destructive action
-  // still deserves a guard (single-source removal had none before).
-  if (!globalThis.window?.confirm(
-    `Remove this ${source.platform} source?\n${source.url}\n\nThe artist and already-downloaded images stay.`,
-  )) return
-  try {
-    // store._dropSource patches the cache in place — no full reload, so the
-    // table + expansion don't reset (operator-flagged lock/reload, 2026-06-07).
-    await store.remove(source.id, source.artist_id)
-  } catch (e) {
-    toast({ text: `Remove failed: ${e?.body?.detail || e?.message || e}`, type: 'error' })
-  }
+  await store.remove(source.id, source.artist_id)
+  await refresh()
 }
 
 async function toggleSourceEnabled({ source, enabled }) {
-  try {
-    await store.update(source.id, { enabled }, source.artist_id)
-  } catch (e) {
-    toast({ text: `Update failed: ${e?.body?.detail || e?.message || e}`, type: 'error' })
-  }
+  await store.update(source.id, { enabled }, source.artist_id)
+  await refresh()
 }
 
 async function onSourceSaved() {
@@ -608,39 +532,31 @@ async function onCheck(source) {
   }
 }
 
-// Plan #693: toggle a run-until-done backfill. Starting walks the full post
-// history in time-boxed chunks across ticks until it reaches the bottom (the
-// row badge tracks progress / completion); stopping cancels back to tick mode.
+// Plan #544: arm a source for backfill mode (gallery-dl walks the full
+// post history) for the next N download runs. Default 3 — enough budget
+// to finish a deep creator without re-prompting the operator across
+// timeout boundaries. The chip on the row reflects the remaining count.
 async function onBackfill(source) {
-  const running = source.backfill_state === 'running'
-  try {
-    if (running) {
-      await store.stopBackfill(source.id, source.artist_id)
-      toast({ text: `Backfill stopped for ${source.artist_name}`, type: 'success' })
-    } else {
-      await store.startBackfill(source.id, source.artist_id)
-      toast({ text: `Backfill started for ${source.artist_name}`, type: 'success' })
-    }
-    // startBackfill/stopBackfill patch the source in place — no full reload.
-  } catch (e) {
-    toast({
-      text: `Backfill ${running ? 'stop' : 'start'} failed: ${e?.body?.detail || e?.message || e}`,
-      type: 'error',
-    })
+  const raw = globalThis.window?.prompt(
+    `Deep scan "${source.artist_name} (${source.platform})" — walk full history for the next how many download runs? (1–10, default 3)`,
+    '3',
+  )
+  if (raw == null) return
+  const runs = parseInt(raw, 10)
+  if (!Number.isFinite(runs) || runs < 1 || runs > 10) {
+    toast({ text: 'Deep scan: runs must be 1–10', type: 'error' })
+    return
   }
-}
-
-// Plan #697: arm a recovery walk (Patreon-only) — re-fetches dropped-and-deleted
-// near-dups and re-evaluates them under the current pHash threshold. Reuses the
-// backfill lifecycle/badge; stop via the same Stop control (onBackfill).
-async function onRecover(source) {
   try {
-    await store.recoverSource(source.id, source.artist_id)
-    toast({ text: `Recovery started for ${source.artist_name}`, type: 'success' })
-    // recoverSource patches the source in place — no full reload.
+    await store.setBackfill(source.id, runs, source.artist_id)
+    toast({
+      text: `Deep scan armed for ${runs} run${runs === 1 ? '' : 's'}`,
+      type: 'success',
+    })
+    await store.loadAll()
   } catch (e) {
     toast({
-      text: `Recovery start failed: ${e?.body?.detail || e?.message || e}`,
+      text: `Deep scan failed: ${e?.detail || e?.message || e}`,
       type: 'error',
     })
   }
@@ -701,28 +617,6 @@ async function bulkSetEnabled(enabled) {
   selected.value = []
 }
 
-// #5: arm a run-until-done backfill on every enabled source in the selected
-// subscriptions (skips ones already backfilling). Patches each row in place.
-async function bulkBackfill() {
-  const sel = resolveSelectedGroups()
-  let armed = 0
-  for (const g of sel) {
-    for (const s of g.sources) {
-      if (!s.enabled || s.backfill_state === 'running') continue
-      try {
-        await store.startBackfill(s.id, s.artist_id)
-        armed += 1
-      } catch { /* keep going */ }
-    }
-  }
-  toast({
-    text: armed ? `Backfill started on ${armed} source${armed === 1 ? '' : 's'}`
-      : 'Nothing to backfill (already running or disabled)',
-    type: armed ? 'success' : 'info',
-  })
-  selected.value = []
-}
-
 async function bulkDelete() {
   const groups = resolveSelectedGroups()
   const total = groups.reduce((n, g) => n + g.sources.length, 0)
@@ -807,17 +701,6 @@ async function bulkDelete() {
 }
 .fc-subs__mactions { display: flex; gap: 2px; }
 .fc-subs__name { font-weight: 600; }
-.fc-subs__sub-url {
-  display: block;
-  max-width: 460px;
-  overflow: hidden;
-  text-overflow: ellipsis;
-  white-space: nowrap;
-  font-size: 0.76rem;
-  color: rgb(var(--v-theme-on-surface-variant));
-  text-decoration: none;
-}
-.fc-subs__sub-url:hover { color: rgb(var(--v-theme-accent)); }
 .fc-subs__chips {
   display: flex; flex-wrap: wrap; gap: 4px;
 }

frontend/src/router.js

@@ -4,7 +4,6 @@ import GalleryView from './views/GalleryView.vue'
 import ShowcaseView from './views/ShowcaseView.vue'
 import TagsView from './views/TagsView.vue'
 import ArtistView from './views/ArtistView.vue'
-import SeriesView from './views/SeriesView.vue'
 import SeriesManageView from './views/SeriesManageView.vue'
 import SeriesReaderView from './views/SeriesReaderView.vue'
 import SubscriptionsView from './views/SubscriptionsView.vue'
@@ -26,9 +25,7 @@ const routes = [
   { path: '/tags',     name: 'tags',     component: TagsView,     meta: { title: 'Tags'     } },
   // Artist detail — no meta.title (reached by clicking an artist, not nav).
   { path: '/artist/:slug', name: 'artist', component: ArtistView },
-  // Series browse — a nav entry (meta.title). Peer of Posts.
-  { path: '/series', name: 'series', component: SeriesView, meta: { title: 'Series' } },
-  // Series management — no meta.title (reached from a series card/tag).
+  // Series management — no meta.title (reached from a series tag card).
   { path: '/series/:tagId', name: 'series-manage', component: SeriesManageView },
   // Series reader — immersive (no top nav, no meta.title).
   { path: '/series/:tagId/read', name: 'series-read', component: SeriesReaderView, meta: { immersive: true } },
@@ -57,19 +54,4 @@ const router = createRouter({
   routes
 })
 
-const DEFAULT_TITLE = 'FabledCurator'
-
-// Keep the tab title in sync with the route on EVERY navigation. List routes set
-// it from meta.title; detail routes (artist, series) have no meta.title, so they
-// reset to the default here and then overwrite it with their own dynamic title
-// (e.g. ArtistView sets "<artist> — FabledCurator"). Without this reset the
-// previous view's dynamic title stuck around — an artist name showing on the
-// Showcase tab, operator-flagged 2026-06-07.
-router.afterEach((to) => {
-  if (typeof document === 'undefined') return
-  document.title = to.meta?.title
-    ? `${to.meta.title} — ${DEFAULT_TITLE}`
-    : DEFAULT_TITLE
-})
-
 export default router

frontend/src/stores/admin.js

@@ -142,22 +142,6 @@ export const useAdminStore = defineStore('admin', () => {
     }
   }
 
-  // #714: Title-Case the back-catalog + merge case/whitespace-variant tags.
-  // dry-run returns a projection inline; live returns {task_id} (long op —
-  // FK repoints) the caller polls via pollTaskUntilDone.
-  async function normalizeTags({ dryRun = true } = {}) {
-    lastError.value = null
-    try {
-      return await api.post(
-        '/api/admin/tags/normalize',
-        { body: { dry_run: dryRun } },
-      )
-    } catch (e) {
-      lastError.value = e.message
-      throw e
-    }
-  }
-
   // --- Task progress polling (taps FC-3i activity dashboard) --------
 
   /**
@@ -194,7 +178,6 @@ export const useAdminStore = defineStore('admin', () => {
     pruneUnusedTags,
     purgeLegacyTags,
     resetContentTagging,
-    normalizeTags,
     pollTaskUntilDone,
   }
 })

frontend/src/stores/artistDirectory.js

@@ -14,10 +14,6 @@ export const useArtistDirectoryStore = defineStore('artistDirectory', () => {
   const q = ref('')
   const platform = ref(null)
   let started = false
-  // Has a load ATTEMPT completed yet? Gates the "No artists match" empty state
-  // so it can't flash on the very first render (loading is still false before
-  // onMounted fires the first loadMore) or between a query change and its fetch.
-  const loaded = ref(false)
   // Typed "alice" then "alice bob" used to drop the second fetch
   // entirely (loading flag still true from the first), so the UI
   // showed alice results while the input said "alice bob". Inflight
@@ -39,7 +35,6 @@ export const useArtistDirectoryStore = defineStore('artistDirectory', () => {
       nextCursor.value = body.next_cursor
       started = true
     })
-    if (t.isCurrent()) loaded.value = true
   }
 
   async function reset() {
@@ -47,7 +42,6 @@ export const useArtistDirectoryStore = defineStore('artistDirectory', () => {
     cards.value = []
     nextCursor.value = null
     started = false
-    loaded.value = false
     await loadMore()
   }
 
@@ -56,8 +50,7 @@ export const useArtistDirectoryStore = defineStore('artistDirectory', () => {
 
   const hasMore = computed(() => !started || nextCursor.value !== null)
   const isEmpty = computed(
-    () => loaded.value && !loading.value
-      && cards.value.length === 0 && error.value === null
+    () => !loading.value && cards.value.length === 0 && error.value === null
   )
 
   return {

frontend/src/stores/seriesBrowse.js

@@ -1,31 +0,0 @@
-import { defineStore } from 'pinia'
-import { ref } from 'vue'
-import { useApi } from '../composables/useApi.js'
-import { useAsyncAction } from '../composables/useAsyncAction.js'
-
-// Backs the Series browse view (FC-6.2). list_series returns every series as a
-// card; homelab scale is small enough to load the whole list and sort/filter
-// server-side.
-export const useSeriesBrowseStore = defineStore('seriesBrowse', () => {
-  const api = useApi()
-  const series = ref([])
-  const sort = ref('recent')
-  const artistId = ref(null)
-  const { loading, error, run } = useAsyncAction({ errorAs: 'message' })
-
-  async function load() {
-    await run(async () => {
-      const params = { sort: sort.value }
-      if (artistId.value != null) params.artist_id = artistId.value
-      const body = await api.get('/api/series', { params })
-      series.value = body.series || []
-    })
-  }
-
-  function setSort(s) {
-    sort.value = s
-    return load()
-  }
-
-  return { series, sort, artistId, loading, error, load, setSort }
-})

frontend/src/stores/seriesManage.js

@@ -16,11 +16,7 @@ export const useSeriesManageStore = defineStore('seriesManage', () => {
 
   const tagId = ref(null)
   const series = ref(null)
-  const chapters = ref([])           // [{id, chapter_number, stated_part, title, is_placeholder, stated_page_start/end, source_post, pages:[...]}]
-  const gaps = ref([])               // missing-page gaps: [{after_chapter_id, start, end}]
-  const partGaps = ref([])           // missing-Part gaps: [{after_chapter_id, start, end}]
-  const pageCount = ref(0)
-  const targetChapterId = ref(null)  // which chapter the picker adds into
+  const pages = ref([])              // [{image_id, page_number, thumbnail_url}]
   const picker = ref([])             // gallery scroll results
   const pickerCursor = ref(null)
   const pickerSelection = ref([])    // image ids
@@ -32,15 +28,7 @@ export const useSeriesManageStore = defineStore('seriesManage', () => {
     try {
       const body = await api.get(`/api/series/${id}/pages`)
       series.value = body.series
-      chapters.value = body.chapters || []
-      gaps.value = body.gaps || []
-      partGaps.value = body.part_gaps || []
-      pageCount.value = (body.pages || []).length
-      // Keep a valid add-target: the selected chapter, else the first one.
-      const ids = chapters.value.map(c => c.id)
-      if (!ids.includes(targetChapterId.value)) {
-        targetChapterId.value = ids.length ? ids[0] : null
-      }
+      pages.value = body.pages
     } finally {
       loading.value = false
     }
@@ -50,79 +38,6 @@ export const useSeriesManageStore = defineStore('seriesManage', () => {
     if (tagId.value != null) await load(tagId.value)
   }
 
-  function gapAfter(chapterId) {
-    return gaps.value.find(g => g.after_chapter_id === chapterId) || null
-  }
-
-  function partGapAfter(chapterId) {
-    return partGaps.value.find(g => g.after_chapter_id === chapterId) || null
-  }
-
-  // ---- chapters ----
-  async function createChapter({ title = null, isPlaceholder = false } = {}) {
-    await api.post(`/api/series/${tagId.value}/chapters`, {
-      body: { title, is_placeholder: isPlaceholder }
-    })
-    await refresh()
-  }
-
-  async function renameChapter(chapterId, title) {
-    await api.patch(`/api/series/${tagId.value}/chapters/${chapterId}`, {
-      body: { title }
-    })
-    await refresh()
-  }
-
-  async function setChapterPart(chapterId, part) {
-    await api.patch(`/api/series/${tagId.value}/chapters/${chapterId}`, {
-      body: { stated_part: part }
-    })
-    await refresh()
-  }
-
-  async function setChapterStated(chapterId, start, end) {
-    await api.patch(`/api/series/${tagId.value}/chapters/${chapterId}`, {
-      body: { stated_page_start: start, stated_page_end: end }
-    })
-    await refresh()
-  }
-
-  async function reorderChapters(orderedChapterIds) {
-    await api.post(`/api/series/${tagId.value}/chapters/reorder`, {
-      body: { chapter_ids: orderedChapterIds }
-    })
-    await refresh()
-  }
-
-  async function moveChapter(chapterId, dir) {
-    const ids = chapters.value.map(c => c.id)
-    const from = ids.indexOf(chapterId)
-    const to = from + dir
-    if (from === -1 || to < 0 || to >= ids.length) return
-    await reorderChapters(moveItem(ids, from, to))
-  }
-
-  async function deleteChapter(chapterId) {
-    await api.delete(`/api/series/${tagId.value}/chapters/${chapterId}`)
-    await refresh()
-  }
-
-  async function mergeChapter(sourceId, targetId) {
-    await api.post(`/api/series/${tagId.value}/chapters/${sourceId}/merge`, {
-      body: { target_chapter_id: targetId }
-    })
-    await refresh()
-    toast({ text: 'Chapters merged', type: 'success' })
-  }
-
-  async function reorderPages(chapterId, orderedImageIds) {
-    await api.post(`/api/series/${tagId.value}/chapters/${chapterId}/reorder`, {
-      body: { image_ids: orderedImageIds }
-    })
-    await refresh()
-  }
-
-  // ---- picker / pages ----
   async function loadPicker(reset = false) {
     if (reset) { picker.value = []; pickerCursor.value = null }
     const params = { limit: 50 }
@@ -140,13 +55,12 @@ export const useSeriesManageStore = defineStore('seriesManage', () => {
 
   async function addSelected() {
     if (pickerSelection.value.length === 0) return
-    if (targetChapterId.value == null) await createChapter()
     await api.post(`/api/series/${tagId.value}/pages`, {
-      body: { image_ids: pickerSelection.value, chapter_id: targetChapterId.value }
+      body: { image_ids: pickerSelection.value }
     })
     pickerSelection.value = []
     await refresh()
-    toast({ text: 'Added to chapter', type: 'success' })
+    toast({ text: 'Added to series', type: 'success' })
   }
 
   async function remove(imageId) {
@@ -156,6 +70,13 @@ export const useSeriesManageStore = defineStore('seriesManage', () => {
     await refresh()
   }
 
+  async function reorder(orderedImageIds) {
+    await api.post(`/api/series/${tagId.value}/reorder`, {
+      body: { image_ids: orderedImageIds }
+    })
+    await refresh()
+  }
+
   async function setCover(imageId) {
     await api.post(`/api/series/${tagId.value}/cover`, {
       body: { image_id: imageId }
@@ -164,11 +85,8 @@ export const useSeriesManageStore = defineStore('seriesManage', () => {
   }
 
   return {
-    tagId, series, chapters, gaps, partGaps, pageCount, targetChapterId,
-    picker, pickerCursor, pickerSelection, loading,
-    load, refresh, gapAfter, partGapAfter,
-    createChapter, renameChapter, setChapterPart, setChapterStated,
-    reorderChapters, moveChapter, deleteChapter, mergeChapter, reorderPages,
-    loadPicker, togglePick, addSelected, remove, setCover
+    tagId, series, pages, picker, pickerCursor, pickerSelection, loading,
+    load, refresh, loadPicker, togglePick, addSelected, remove,
+    reorder, setCover
   }
 })

frontend/src/stores/seriesReader.js

@@ -42,27 +42,7 @@ export const useSeriesReaderStore = defineStore('seriesReader', () => {
     await run(async () => {
       const body = await api.get(`/api/series/${tagId}/pages`)
       series.value = body.series
-      // page_number is now WITHIN a chapter, so it can't anchor scroll/jump
-      // (two chapters both have a page 1). Decorate each page with a global
-      // `seq` (reading-order position) for anchors, plus chapter-divider info
-      // so the reader can mark where each chapter begins.
-      const chapters = body.chapters || []
-      const labelById = {}
-      for (const c of chapters) {
-        labelById[c.id] = c.title || `Chapter ${c.chapter_number}`
-      }
-      let prevChapter = null
-      pages.value = (body.pages || []).map((p, i) => {
-        const isChapterStart =
-          chapters.length > 1 && p.chapter_id !== prevChapter
-        prevChapter = p.chapter_id
-        return {
-          ...p,
-          seq: i + 1,
-          isChapterStart,
-          chapterLabel: labelById[p.chapter_id] || null
-        }
-      })
+      pages.value = body.pages
     })
   }
 

frontend/src/stores/seriesSuggestions.js

@@ -1,75 +0,0 @@
-import { defineStore } from 'pinia'
-import { ref } from 'vue'
-import { useApi } from '../composables/useApi.js'
-import { useAsyncAction } from '../composables/useAsyncAction.js'
-import { toast } from '../utils/toast.js'
-
-// Backs the Series → Suggestions tab (FC-6.3). The matcher is confirm-only:
-// accept turns a suggestion into a chapter; skip dismisses it. enabled +
-// threshold are the operator-tunable knobs (DB-backed via /settings/import).
-export const useSeriesSuggestionsStore = defineStore('seriesSuggestions', () => {
-  const api = useApi()
-  const suggestions = ref([])
-  const enabled = ref(true)
-  const threshold = ref(0.5)
-  const { loading, error, run } = useAsyncAction({ errorAs: 'message' })
-
-  async function load() {
-    await run(async () => {
-      const body = await api.get('/api/series/suggestions')
-      suggestions.value = body.suggestions || []
-    })
-  }
-
-  async function loadSettings() {
-    const s = await api.get('/api/settings/import')
-    enabled.value = s.series_suggest_enabled
-    threshold.value = s.series_suggest_threshold
-  }
-
-  async function saveSettings(patch) {
-    await api.patch('/api/settings/import', { body: patch })
-  }
-
-  async function setEnabled(v) {
-    enabled.value = v
-    await saveSettings({ series_suggest_enabled: v })
-  }
-
-  async function setThreshold(v) {
-    threshold.value = v
-    await saveSettings({ series_suggest_threshold: v })
-  }
-
-  async function accept(id) {
-    try {
-      await api.post(`/api/series/suggestions/${id}/accept`, {})
-      suggestions.value = suggestions.value.filter(s => s.id !== id)
-      toast({ text: 'Added to series', type: 'success' })
-    } catch (e) {
-      toast({ text: `Accept failed: ${e.message}`, type: 'error' })
-    }
-  }
-
-  async function dismiss(id) {
-    try {
-      await api.post(`/api/series/suggestions/${id}/dismiss`, {})
-      suggestions.value = suggestions.value.filter(s => s.id !== id)
-    } catch (e) {
-      toast({ text: `Skip failed: ${e.message}`, type: 'error' })
-    }
-  }
-
-  async function rescan() {
-    await api.post('/api/series/suggestions/rescan', {})
-    toast({
-      text: 'Rescan queued — runs in the background; reload to see new matches.',
-      type: 'info'
-    })
-  }
-
-  return {
-    suggestions, enabled, threshold, loading, error,
-    load, loadSettings, setEnabled, setThreshold, accept, dismiss, rescan
-  }
-})

frontend/src/stores/sources.js

@@ -32,31 +32,6 @@ export const useSourcesStore = defineStore('sources', () => {
     if (artistId != null) byArtist.value.delete(artistId)
   }
 
-  // Patch a single updated source into every cached array IN PLACE, rather than
-  // invalidating + refetching the whole list. The full refetch was what locked
-  // the Subscriptions UI and collapsed the expanded row after every inline
-  // action (check / backfill / recover / toggle) — operator-flagged 2026-06-07.
-  // Map.set on a Vue-reactive ref(Map) triggers re-render; we hand each cache a
-  // NEW array so the table diffs just the one changed row.
-  function _patchSource(updated) {
-    if (!updated || updated.id == null) return
-    for (const [key, arr] of byArtist.value) {
-      const i = arr.findIndex((s) => s.id === updated.id)
-      if (i === -1) continue
-      const next = arr.slice()
-      next[i] = { ...arr[i], ...updated }
-      byArtist.value.set(key, next)
-    }
-  }
-
-  function _dropSource(id) {
-    for (const [key, arr] of byArtist.value) {
-      if (arr.some((s) => s.id === id)) {
-        byArtist.value.set(key, arr.filter((s) => s.id !== id))
-      }
-    }
-  }
-
   async function create(payload) {
     const body = await api.post('/api/sources', { body: payload })
     _invalidate(payload.artist_id)
@@ -65,13 +40,13 @@ export const useSourcesStore = defineStore('sources', () => {
 
   async function update(id, patch, artistIdHint = null) {
     const body = await api.patch(`/api/sources/${id}`, { body: patch })
-    _patchSource(body)
+    _invalidate(artistIdHint ?? body.artist_id)
     return body
   }
 
   async function remove(id, artistIdHint = null) {
     await api.delete(`/api/sources/${id}`)
-    _dropSource(id)
+    _invalidate(artistIdHint)
   }
 
   async function findOrCreateArtist(name) {
@@ -110,33 +85,14 @@ export const useSourcesStore = defineStore('sources', () => {
     }
   }
 
-  // Plan #693: start/stop a run-until-done backfill. 'start' walks the full
-  // post history in time-boxed chunks until it reaches the bottom (then the
-  // source shows backfill_state 'complete'); 'stop' cancels back to tick mode.
-  async function startBackfill(id, artistIdHint = null) {
-    const body = await api.post(`/api/sources/${id}/backfill`, { body: { action: 'start' } })
-    _patchSource(body)
+  // Plan #544: arm a source for backfill mode. The next `runs` download
+  // runs (default 3) walk gallery-dl's full post history instead of
+  // exiting early at the first contiguous archived block.
+  async function setBackfill(id, runs = 3, artistIdHint = null) {
+    const body = await api.post(`/api/sources/${id}/backfill`, { body: { runs } })
+    _invalidate(artistIdHint ?? body.artist_id)
     return body
   }
-  async function stopBackfill(id, artistIdHint = null) {
-    const body = await api.post(`/api/sources/${id}/backfill`, { body: { action: 'stop' } })
-    _patchSource(body)
-    return body
-  }
-  // Plan #697: arm a recovery walk — a backfill that bypasses the Patreon
-  // seen-ledger to re-fetch dropped-and-deleted near-dups and re-evaluate them
-  // under the current pHash threshold. Stop with stopBackfill (shared lifecycle).
-  async function recoverSource(id, artistIdHint = null) {
-    const body = await api.post(`/api/sources/${id}/backfill`, { body: { action: 'recover' } })
-    _patchSource(body)
-    return body
-  }
-
-  // Plan #708 B4: dry-run preview — count what a backfill WOULD download
-  // (native platforms), without downloading. Read-only, so no cache invalidate.
-  async function previewSource(id) {
-    return await api.post(`/api/sources/${id}/preview`)
-  }
 
   function sourcesByArtistGrouped() {
     // returns [{artist: {id,name,slug}, sources: [...]}, ...]
@@ -164,10 +120,7 @@ export const useSourcesStore = defineStore('sources', () => {
     loadAll, loadForArtist,
     create, update, remove,
     checkNow,
-    startBackfill,
-    stopBackfill,
-    recoverSource,
-    previewSource,
+    setBackfill,
     findOrCreateArtist, autocompleteArtist,
     loadScheduleStatus,
     sourcesByArtistGrouped,

frontend/src/stores/tags.js

@@ -34,20 +34,10 @@ export const useTagStore = defineStore('tags', () => {
   }
 
   async function loadFandoms() {
-    // List ALL fandoms via the cursor-paged tag directory. (#712: the old impl
-    // abused /tags/autocomplete with q=' ', which the backend strips to empty
-    // and returns [] — so the picker showed no existing fandoms.)
-    const all = []
-    let cursor = null
-    do {
-      const params = { kind: 'fandom', limit: 200 }
-      if (cursor) params.cursor = cursor
-      const body = await api.get('/api/tags/directory', { params })
-      all.push(...(body.cards || []))
-      cursor = body.next_cursor
-    } while (cursor)
-    fandomCache.value = all
-    return all
+    fandomCache.value = await api.get('/api/tags/autocomplete', {
+      params: { q: ' ', kind: 'fandom', limit: 200 }
+    })
+    return fandomCache.value
   }
 
   async function createFandom(name) {

frontend/src/views/ArtistsView.vue

@@ -76,11 +76,9 @@ onMounted(async () => {
 }
 .fc-artists__grid {
   display: grid;
-  /* min(360px, 100%) so a card never exceeds the viewport (phones collapse to
-     one column). 360 keeps a LONE column under ~732px wide, so the 3/1 preview
-     strip stays ≲244px tall and always fills the card width (no dead space),
-     while giving 2+ columns on a normal desktop. */
-  grid-template-columns: repeat(auto-fill, minmax(min(360px, 100%), 1fr));
+  /* min(440px, 100%) so a card never exceeds the viewport — on phones the
+     min-track collapses to 100% (single column) instead of overflowing. */
+  grid-template-columns: repeat(auto-fill, minmax(min(440px, 100%), 1fr));
   gap: 12px;
 }
 .fc-artists__sentinel {

frontend/src/views/SeriesManageView.vue

@@ -1,311 +1,83 @@
 <template>
-  <v-container fluid class="pt-2 pb-10 fc-series">
+  <v-container fluid class="pt-2 pb-6">
     <div class="fc-series__head">
       <span class="fc-series__name">{{ store.series?.name || 'Series' }}</span>
-      <span class="fc-series__count">
-        {{ store.chapters.length }} part(s) · {{ store.pageCount }} page(s)
-      </span>
-      <v-spacer />
+      <span class="fc-series__count">{{ store.pages.length }} page(s)</span>
       <v-btn
-        v-if="store.pageCount > 0"
+        v-if="store.pages.length > 0"
         size="small" variant="tonal" color="accent"
         prepend-icon="mdi-book-open"
+        class="fc-series__read"
         @click="$router.push({ name: 'series-read', params: { tagId: store.tagId } })"
       >Read</v-btn>
     </div>
 
-    <p class="fc-series__hint">
-      Each part is one installment — drag pages to set their order, and set the
-      <strong>Part #</strong> to mark where it falls in the story. Use
-      <strong>Add pages</strong> to pull images from the gallery into the part
-      you're working on.
-    </p>
-
-    <!-- Parts, full width -->
-    <div class="fc-parts">
-      <template v-for="(ch, ci) in store.chapters" :key="ch.id">
-        <section class="fc-part">
-          <header class="fc-part__head">
-            <label class="fc-part__partfield" :title="'Part number for this installment'">
-              <span class="fc-part__partlabel">Part</span>
-              <input
-                class="fc-part__partnum" type="number" min="1"
-                :value="partDraft[ch.id] ?? ''"
-                :placeholder="String(ch.chapter_number)"
-                @input="partDraft[ch.id] = $event.target.value"
-                @keydown.enter.prevent="commitPart(ch)"
-                @blur="commitPart(ch)"
-              >
-            </label>
-
-            <v-text-field
-              v-model="titleDraft[ch.id]"
-              :placeholder="`Untitled — Part ${ch.stated_part ?? ch.chapter_number}`"
-              density="compact" variant="plain" hide-details
-              class="fc-part__title"
-              @keydown.enter="commitTitle(ch)"
-              @blur="commitTitle(ch)"
-            />
-
-            <span
-              v-if="ch.source_post?.title" class="fc-part__src"
-              :title="ch.source_post.title"
-            >
-              <v-icon size="x-small">mdi-link-variant</v-icon>
-              {{ ch.source_post.title }}
-            </span>
-
-            <span v-if="ch.is_placeholder" class="fc-part__badge">placeholder</span>
-            <span v-else class="fc-part__pc">{{ ch.pages.length }} pg</span>
-
-            <v-spacer />
-
-            <v-btn
-              v-if="!ch.is_placeholder"
-              size="small" variant="tonal" color="accent"
-              prepend-icon="mdi-image-plus"
-              @click="openPicker(ch.id)"
-            >Add pages</v-btn>
-
-            <v-menu location="bottom end">
-              <template #activator="{ props }">
-                <v-btn
-                  v-bind="props" size="small" variant="text"
-                  icon="mdi-dots-vertical" title="Part actions"
-                />
-              </template>
-              <v-list density="compact">
-                <v-list-item
-                  prepend-icon="mdi-chevron-up" title="Move up"
-                  :disabled="ci === 0"
-                  @click="store.moveChapter(ch.id, -1)"
-                />
-                <v-list-item
-                  prepend-icon="mdi-chevron-down" title="Move down"
-                  :disabled="ci === store.chapters.length - 1"
-                  @click="store.moveChapter(ch.id, 1)"
-                />
-                <v-list-item
-                  prepend-icon="mdi-arrow-collapse-up" title="Merge into previous"
-                  :disabled="ci === 0"
-                  @click="store.mergeChapter(ch.id, store.chapters[ci - 1].id)"
-                />
-                <v-divider />
-                <v-list-item
-                  prepend-icon="mdi-delete-outline" title="Delete part"
-                  base-color="error"
-                  @click="confirmDelete(ch)"
-                />
-              </v-list>
-            </v-menu>
-          </header>
-
-          <div class="fc-part__statedrow">
-            <span class="fc-part__statedlabel">Printed pages</span>
-            <input
-              class="fc-part__statedin" type="number" min="0"
-              :value="ch.stated_page_start ?? ''" placeholder="start"
-              @change="onStated(ch, 'start', $event)"
-            >
-            <span class="fc-part__dash">–</span>
-            <input
-              class="fc-part__statedin" type="number" min="0"
-              :value="ch.stated_page_end ?? ''" placeholder="end"
-              @change="onStated(ch, 'end', $event)"
-            >
-            <span class="fc-part__statedhelp">
-              optional — the page numbers printed in this installment
-            </span>
+    <div class="fc-series__body">
+      <div class="fc-series__pages">
+        <div
+          v-for="(p, idx) in store.pages" :key="p.image_id"
+          class="fc-series__page" draggable="true"
+          @dragstart="dragFrom = idx"
+          @dragover.prevent
+          @drop="onDrop(idx)"
+        >
+          <span class="fc-series__pn">{{ p.page_number }}</span>
+          <img :src="p.thumbnail_url" alt="" loading="lazy" />
+          <div class="fc-series__pageactions">
+            <v-btn size="x-small" variant="text" icon="mdi-image-frame"
+                   title="Make cover" @click="store.setCover(p.image_id)" />
+            <v-btn size="x-small" variant="text" icon="mdi-close"
+                   title="Remove" @click="store.remove(p.image_id)" />
           </div>
-
-          <div v-if="ch.is_placeholder" class="fc-part__reserved">
-            Reserved slot — a part you don't have yet.
-          </div>
-          <div v-else-if="ch.pages.length === 0" class="fc-part__empty">
-            No pages yet.
-            <v-btn
-              size="small" variant="text" color="accent"
-              prepend-icon="mdi-image-plus" @click="openPicker(ch.id)"
-            >Add pages</v-btn>
-          </div>
-          <div v-else class="fc-part__pages">
-            <div
-              v-for="(p, pi) in ch.pages" :key="p.image_id"
-              class="fc-page" draggable="true"
-              :class="{ 'fc-page--dragging': drag && drag.chapterId === ch.id && drag.idx === pi }"
-              @dragstart="drag = { chapterId: ch.id, idx: pi }"
-              @dragend="drag = null"
-              @dragover.prevent
-              @drop="onPageDrop(ch, pi)"
-            >
-              <span class="fc-page__pn">{{ p.page_number }}</span>
-              <img :src="p.thumbnail_url" alt="" loading="lazy" />
-              <div class="fc-page__actions">
-                <v-btn size="x-small" variant="flat" icon="mdi-image-frame"
-                       title="Make series cover" @click="store.setCover(p.image_id)" />
-                <v-btn size="x-small" variant="flat" icon="mdi-close"
-                       title="Remove from series" @click="store.remove(p.image_id)" />
-              </div>
-            </div>
-          </div>
-        </section>
-
-        <div v-if="store.partGapAfter(ch.id)" class="fc-gap">
-          <v-icon size="x-small">mdi-alert-outline</v-icon>
-          Missing
-          <template v-if="store.partGapAfter(ch.id).start === store.partGapAfter(ch.id).end">
-            Part {{ store.partGapAfter(ch.id).start }}
-          </template>
-          <template v-else>
-            Parts {{ store.partGapAfter(ch.id).start }}–{{ store.partGapAfter(ch.id).end }}
-          </template>
         </div>
-        <div v-if="store.gapAfter(ch.id)" class="fc-gap">
-          <v-icon size="x-small">mdi-alert-outline</v-icon>
-          Gap: printed pages {{ store.gapAfter(ch.id).start }}–{{ store.gapAfter(ch.id).end }} missing
+        <div v-if="store.pages.length === 0" class="fc-series__empty">
+          No pages yet — add images from the right.
         </div>
-      </template>
-
-      <div v-if="store.chapters.length === 0" class="fc-series__empty">
-        No parts yet — add one, then add pages from the gallery.
       </div>
 
-      <div class="fc-parts__add">
-        <v-btn size="small" variant="tonal" prepend-icon="mdi-plus"
-               @click="store.createChapter()">Add part</v-btn>
-        <v-btn size="small" variant="text" prepend-icon="mdi-bookmark-outline"
-               @click="store.createChapter({ isPlaceholder: true })">
-          Add placeholder
-        </v-btn>
-      </div>
-    </div>
-
-    <!-- Picker slide-over -->
-    <v-navigation-drawer
-      v-model="pickerOpen" location="right" temporary width="420"
-      class="fc-picker"
-    >
-      <div class="fc-picker__head">
-        <div class="fc-picker__title">
-          <span>Add pages</span>
-          <v-btn size="x-small" variant="text" icon="mdi-close"
-                 @click="pickerOpen = false" />
-        </div>
-        <v-select
-          v-model="store.targetChapterId"
-          :items="chapterItems" item-title="label" item-value="id"
-          density="compact" variant="outlined" hide-details
-          label="Add to part" class="mt-2"
-        />
-        <div class="fc-picker__bar">
+      <div class="fc-series__picker">
+        <div class="fc-series__pickerhead">
           <span>{{ store.pickerSelection.length }} selected</span>
           <v-btn
             size="small" color="accent" variant="flat"
             :disabled="store.pickerSelection.length === 0"
             @click="store.addSelected()"
-          >Add to part</v-btn>
+          >Add to series</v-btn>
         </div>
-      </div>
-      <div class="fc-picker__grid">
-        <div
-          v-for="img in store.picker" :key="img.id"
-          class="fc-picker__pick"
-          :class="{ on: store.pickerSelection.includes(img.id) }"
-          @click="store.togglePick(img.id)"
-        >
-          <img :src="img.thumbnail_url" alt="" loading="lazy" />
-          <v-icon
-            v-if="store.pickerSelection.includes(img.id)"
-            class="fc-picker__check" color="accent" size="small"
-          >mdi-check-circle</v-icon>
+        <div class="fc-series__pickergrid">
+          <div
+            v-for="img in store.picker" :key="img.id"
+            class="fc-series__pick"
+            :class="{ on: store.pickerSelection.includes(img.id) }"
+            @click="store.togglePick(img.id)"
+          >
+            <img :src="img.thumbnail_url" alt="" loading="lazy" />
+          </div>
         </div>
-        <div ref="sentinel" class="fc-picker__sentinel" />
+        <div ref="sentinel" class="fc-series__sentinel" />
       </div>
-    </v-navigation-drawer>
+    </div>
   </v-container>
 </template>
 
 <script setup>
-import { computed, onMounted, reactive, ref, watch } from 'vue'
+import { onMounted, ref } from 'vue'
 import { useRoute } from 'vue-router'
 import { useSeriesManageStore, moveItem } from '../stores/seriesManage.js'
 import { useInfiniteScroll } from '../composables/useInfiniteScroll.js'
 
 const route = useRoute()
 const store = useSeriesManageStore()
+const dragFrom = ref(null)
 const sentinel = ref(null)
-const drag = ref(null)            // { chapterId, idx }
-const titleDraft = reactive({})   // chapterId -> draft title
-const partDraft = reactive({})    // chapterId -> draft stated_part (string)
-const pickerOpen = ref(false)
 
-// Keep local drafts in sync with loaded chapters. Edits commit on blur/Enter,
-// which refreshes and resets the draft to the saved value.
-watch(() => store.chapters, (chs) => {
-  for (const k of Object.keys(titleDraft)) delete titleDraft[k]
-  for (const k of Object.keys(partDraft)) delete partDraft[k]
-  for (const c of chs) {
-    titleDraft[c.id] = c.title || ''
-    partDraft[c.id] = c.stated_part == null ? '' : String(c.stated_part)
-  }
-}, { immediate: true })
-
-const chapterItems = computed(() =>
-  store.chapters.map(c => ({
-    id: c.id,
-    label: `Part ${c.stated_part ?? c.chapter_number}` +
-      (c.title ? ` — ${c.title}` : '') +
-      (c.is_placeholder ? ' (placeholder)' : ` · ${c.pages.length} pg`),
-  }))
-)
-
-function commitTitle(ch) {
-  const v = (titleDraft[ch.id] || '').trim()
-  if (v === (ch.title || '')) return
-  store.renameChapter(ch.id, v || null)
-}
-
-function commitPart(ch) {
-  const raw = (partDraft[ch.id] ?? '').trim()
-  const next = raw === '' ? null : parseInt(raw, 10)
-  if (raw !== '' && (Number.isNaN(next) || next < 1)) {
-    partDraft[ch.id] = ch.stated_part == null ? '' : String(ch.stated_part)
-    return
-  }
-  if (next === (ch.stated_part ?? null)) return
-  store.setChapterPart(ch.id, next)
-}
-
-function onStated(ch, which, ev) {
-  const raw = ev.target.value
-  const n = raw === '' ? null : parseInt(raw, 10)
-  const start = which === 'start' ? n : ch.stated_page_start
-  const end = which === 'end' ? n : ch.stated_page_end
-  store.setChapterStated(ch.id, start, end)
-}
-
-function onPageDrop(chapter, toIdx) {
-  if (!drag.value || drag.value.chapterId !== chapter.id) { drag.value = null; return }
-  if (drag.value.idx === toIdx) { drag.value = null; return }
+function onDrop(toIdx) {
+  if (dragFrom.value === null || dragFrom.value === toIdx) return
   const ordered = moveItem(
-    chapter.pages.map(p => p.image_id), drag.value.idx, toIdx
+    store.pages.map(p => p.image_id), dragFrom.value, toIdx
   )
-  drag.value = null
-  store.reorderPages(chapter.id, ordered)
-}
-
-function confirmDelete(ch) {
-  const label = `Part ${ch.stated_part ?? ch.chapter_number}`
-  const n = ch.pages.length
-  const msg = n
-    ? `Delete ${label} and remove its ${n} page(s) from the series?`
-    : `Delete ${label}?`
-  if (window.confirm(msg)) store.deleteChapter(ch.id)
-}
-
-function openPicker(chapterId) {
-  store.targetChapterId = chapterId
-  pickerOpen.value = true
+  dragFrom.value = null
+  store.reorder(ordered)
 }
 
 useInfiniteScroll(sentinel, () => store.loadPicker())
@@ -318,146 +90,51 @@ onMounted(async () => {
 
 <style scoped>
 .fc-series__head {
-  display: flex; align-items: baseline; gap: 12px; margin-bottom: 4px;
+  display: flex; align-items: baseline; gap: 12px; margin-bottom: 16px;
+}
+.fc-series__name {
+  font-family: 'Fraunces', Georgia, serif; font-size: 22px;
 }
-.fc-series__name { font-family: 'Fraunces', Georgia, serif; font-size: 24px; }
 .fc-series__count {
   font-size: 13px; color: rgb(var(--v-theme-on-surface-variant));
 }
-.fc-series__hint {
-  font-size: 13px; color: rgb(var(--v-theme-on-surface-variant));
-  margin-bottom: 18px; max-width: 760px;
+.fc-series__body { display: flex; gap: 16px; align-items: flex-start; }
+.fc-series__pages { flex: 1; min-width: 0; display: flex;
+  flex-direction: column; gap: 6px; }
+.fc-series__page {
+  display: flex; align-items: center; gap: 10px; padding: 6px;
+  background: rgb(var(--v-theme-surface)); border-radius: 6px;
+  cursor: grab;
 }
-
-/* Parts — full width, stacked */
-.fc-parts { display: flex; flex-direction: column; gap: 14px; max-width: 1100px; }
-.fc-part {
-  border: 1px solid rgb(var(--v-theme-surface-light));
-  border-radius: 10px; padding: 12px 14px;
-  background: rgb(var(--v-theme-surface));
+.fc-series__page img {
+  width: 64px; height: 64px; object-fit: cover; border-radius: 4px;
 }
-.fc-part__head { display: flex; align-items: center; gap: 12px; flex-wrap: wrap; }
-
-.fc-part__partfield {
-  display: inline-flex; align-items: center; gap: 6px;
-  background: rgb(var(--v-theme-accent), 0.12);
-  border: 1px solid rgb(var(--v-theme-accent), 0.35);
-  border-radius: 8px; padding: 4px 8px;
-}
-.fc-part__partlabel {
-  font-size: 12px; text-transform: uppercase; letter-spacing: 0.06em;
-  color: rgb(var(--v-theme-accent));
-}
-.fc-part__partnum {
-  width: 46px; text-align: center; font-size: 18px; font-weight: 600;
-  font-variant-numeric: tabular-nums;
-  background: transparent; border: none; color: rgb(var(--v-theme-on-surface));
-}
-.fc-part__partnum:focus { outline: none; }
-.fc-part__partnum::placeholder { color: rgb(var(--v-theme-on-surface-variant)); opacity: 0.6; }
-
-.fc-part__title { flex: 1 1 180px; min-width: 120px; font-size: 15px; }
-.fc-part__src {
-  display: inline-flex; align-items: center; gap: 4px; max-width: 240px;
-  overflow: hidden; text-overflow: ellipsis; white-space: nowrap;
-  font-size: 12px; color: rgb(var(--v-theme-on-surface-variant));
-}
-.fc-part__badge {
-  font-size: 11px; text-transform: uppercase; letter-spacing: 0.04em;
+.fc-series__pn {
+  width: 28px; text-align: center;
   color: rgb(var(--v-theme-on-surface-variant));
-}
-.fc-part__pc {
-  font-size: 12px; color: rgb(var(--v-theme-on-surface-variant));
   font-variant-numeric: tabular-nums;
 }
-
-.fc-part__statedrow {
-  display: flex; align-items: center; gap: 6px; margin: 8px 0 2px;
-  font-size: 12px; color: rgb(var(--v-theme-on-surface-variant));
-}
-.fc-part__statedlabel { text-transform: uppercase; letter-spacing: 0.04em; font-size: 11px; }
-.fc-part__statedin {
-  width: 56px; text-align: center; font-size: 12px;
-  background: rgb(var(--v-theme-surface-light));
-  border: 1px solid transparent; border-radius: 4px; padding: 2px 4px;
-  color: rgb(var(--v-theme-on-surface));
-}
-.fc-part__statedin:focus { outline: none; border-color: rgb(var(--v-theme-accent)); }
-.fc-part__dash { opacity: 0.6; }
-.fc-part__statedhelp { font-size: 11px; opacity: 0.7; }
-
-.fc-part__reserved, .fc-part__empty {
-  padding: 18px; text-align: center; font-size: 13px;
-  color: rgb(var(--v-theme-on-surface-variant));
-}
-
-/* Big page grid */
-.fc-part__pages {
-  display: grid; gap: 10px; margin-top: 10px;
-  grid-template-columns: repeat(auto-fill, minmax(150px, 1fr));
-}
-.fc-page {
-  position: relative; border-radius: 8px; overflow: hidden; cursor: grab;
-  background: rgb(var(--v-theme-surface-light));
-  border: 1px solid rgb(var(--v-theme-surface-light));
-}
-.fc-page--dragging { opacity: 0.4; }
-.fc-page img {
-  width: 100%; aspect-ratio: 3 / 4; object-fit: contain;
-  display: block; background: rgb(var(--v-theme-background));
-}
-.fc-page__pn {
-  position: absolute; top: 6px; left: 6px; z-index: 2;
-  min-width: 24px; height: 24px; padding: 0 6px; border-radius: 12px;
-  display: inline-flex; align-items: center; justify-content: center;
-  font-size: 13px; font-weight: 600; font-variant-numeric: tabular-nums;
-  background: rgb(var(--v-theme-accent)); color: rgb(var(--v-theme-on-accent, 0 0 0));
-}
-.fc-page__actions {
-  position: absolute; top: 4px; right: 4px; z-index: 2;
-  display: flex; gap: 2px; opacity: 0; transition: opacity 0.12s;
-}
-.fc-page:hover .fc-page__actions, .fc-page:focus-within .fc-page__actions { opacity: 1; }
-.fc-page__actions :deep(.v-btn) {
-  background: rgba(0, 0, 0, 0.55); color: #fff;
-}
-
-.fc-gap {
-  display: flex; align-items: center; gap: 6px; padding: 2px 10px;
-  font-size: 12px; color: rgb(var(--v-theme-warning, var(--v-theme-accent)));
-}
-.fc-parts__add { display: flex; gap: 8px; margin-top: 4px; }
+.fc-series__pageactions { margin-left: auto; }
 .fc-series__empty {
-  padding: 40px; text-align: center; color: rgb(var(--v-theme-on-surface-variant));
+  padding: 32px; text-align: center;
+  color: rgb(var(--v-theme-on-surface-variant));
 }
+.fc-series__picker { flex: 1; min-width: 0; }
+.fc-series__pickerhead {
+  display: flex; align-items: center; justify-content: space-between;
+  margin-bottom: 8px;
+}
+.fc-series__pickergrid {
+  display: grid; grid-template-columns: repeat(auto-fill, minmax(120px, 1fr));
+  gap: 6px;
+}
+.fc-series__pick { cursor: pointer; aspect-ratio: 1; overflow: hidden;
+  border-radius: 4px; outline: 2px solid transparent; }
+.fc-series__pick img { width: 100%; height: 100%; object-fit: cover; }
+.fc-series__pick.on { outline-color: rgb(var(--v-theme-accent)); }
+.fc-series__sentinel { height: 40px; }
 
-/* Picker slide-over */
-.fc-picker__head {
-  position: sticky; top: 0; z-index: 1; padding: 12px 14px;
-  background: rgb(var(--v-theme-surface));
-  border-bottom: 1px solid rgb(var(--v-theme-surface-light));
+@media (max-width: 900px) {
+  .fc-series__body { flex-direction: column; }
 }
-.fc-picker__title {
-  display: flex; align-items: center; justify-content: space-between;
-  font-family: 'Fraunces', Georgia, serif; font-size: 18px;
-}
-.fc-picker__bar {
-  display: flex; align-items: center; justify-content: space-between;
-  margin-top: 10px; font-size: 13px;
-}
-.fc-picker__grid {
-  display: grid; gap: 6px; padding: 12px;
-  grid-template-columns: repeat(auto-fill, minmax(110px, 1fr));
-}
-.fc-picker__pick {
-  position: relative; cursor: pointer; aspect-ratio: 1; overflow: hidden;
-  border-radius: 6px; outline: 2px solid transparent;
-}
-.fc-picker__pick img { width: 100%; height: 100%; object-fit: cover; }
-.fc-picker__pick.on { outline-color: rgb(var(--v-theme-accent)); }
-.fc-picker__check {
-  position: absolute; top: 4px; right: 4px;
-  background: rgba(0, 0, 0, 0.5); border-radius: 50%;
-}
-.fc-picker__sentinel { grid-column: 1 / -1; height: 40px; }
 </style>

frontend/src/views/SeriesReaderView.vue

@@ -20,9 +20,9 @@
           :value="currentPage" @change="jumpTo($event.target.value)"
         >
           <option
-            v-for="p in store.pages" :key="p.seq"
-            :value="p.seq"
-          >Page {{ p.seq }}</option>
+            v-for="p in store.pages" :key="p.page_number"
+            :value="p.page_number"
+          >Page {{ p.page_number }}</option>
         </select>
         <v-btn
           icon="mdi-menu" variant="text" size="small"
@@ -43,27 +43,23 @@
         <div
           v-for="p in store.pages" :key="p.image_id"
           class="fc-reader__thumb"
-          :class="{ active: p.seq === currentPage }"
-          @click="jumpTo(p.seq, true)"
+          :class="{ active: p.page_number === currentPage }"
+          @click="jumpTo(p.page_number, true)"
         >
           <img :src="p.thumbnail_url" alt="" loading="lazy" />
-          <span class="fc-reader__thumbnum">{{ p.seq }}</span>
+          <span class="fc-reader__thumbnum">{{ p.page_number }}</span>
         </div>
       </div>
     </aside>
 
     <div ref="scrollEl" class="fc-reader__content" @scroll="onScroll">
-      <template v-for="p in store.pages" :key="p.image_id">
-        <div v-if="p.isChapterStart" class="fc-reader__chdiv">
-          {{ p.chapterLabel }}
-        </div>
-        <div
-          class="fc-reader__page" :id="'fc-page-' + p.seq"
-          :data-page="p.seq"
-        >
-          <img :src="p.image_url" :alt="'Page ' + p.seq" loading="lazy" />
-        </div>
-      </template>
+      <div
+        v-for="p in store.pages" :key="p.image_id"
+        class="fc-reader__page" :id="'fc-page-' + p.page_number"
+        :data-page="p.page_number"
+      >
+        <img :src="p.image_url" :alt="'Page ' + p.page_number" loading="lazy" />
+      </div>
       <div v-if="store.error" class="fc-reader__empty">
         {{ store.error }}
         <v-btn variant="text" color="accent" @click="goBack">Back</v-btn>
@@ -119,9 +115,9 @@ function pageMetrics() {
   const root = scrollEl.value
   if (!root) return []
   return store.pages.map(p => {
-    const el = root.querySelector('#fc-page-' + p.seq)
+    const el = root.querySelector('#fc-page-' + p.page_number)
     return {
-      page_number: p.seq,
+      page_number: p.page_number,
       top: el ? el.offsetTop : 0,
       height: el ? el.offsetHeight : 0
     }
@@ -284,13 +280,6 @@ onUnmounted(() => {
   display: flex; flex-direction: column; align-items: center;
   gap: 2px; padding: 0.25rem;
 }
-.fc-reader__chdiv {
-  width: 100%; max-width: 1200px; margin: 0.5rem auto 0;
-  padding: 0.4rem 0.75rem;
-  font-family: 'Fraunces', Georgia, serif; font-size: 0.95rem;
-  color: rgb(var(--v-theme-on-surface-variant));
-  border-top: 1px solid rgb(var(--v-theme-surface-light));
-}
 .fc-reader__page { width: 100%; display: flex; justify-content: center; }
 .fc-reader__page img {
   max-width: min(100%, 1200px); height: auto; display: block;

frontend/src/views/SeriesView.vue

@@ -1,264 +0,0 @@
-<template>
-  <v-container fluid class="pt-2 pb-6">
-    <v-tabs v-model="tab" density="compact" class="mb-3">
-      <v-tab value="browse">Browse</v-tab>
-      <v-tab value="suggestions">
-        Suggestions
-        <v-badge
-          v-if="sug.suggestions.length" :content="sug.suggestions.length"
-          inline color="accent"
-        />
-      </v-tab>
-    </v-tabs>
-
-    <v-window v-model="tab">
-      <!-- ---- Browse ---- -->
-      <v-window-item value="browse">
-        <div class="fc-series-browse__controls">
-          <v-select
-            :model-value="store.sort"
-            :items="sortItems"
-            density="compact" variant="outlined" hide-details
-            label="Sort" style="max-width: 220px;"
-            @update:model-value="store.setSort($event)"
-          />
-        </div>
-
-        <v-alert v-if="store.error" type="error" variant="tonal" closable class="my-4">
-          Failed to load: {{ store.error }}
-        </v-alert>
-        <div
-          v-else-if="!store.loading && store.series.length === 0"
-          class="fc-series-browse__empty"
-        >
-          No series yet. Make one from a post's “Series ▾ → New series from this
-          post”, or create a <code>series:</code> tag and add images.
-        </div>
-
-        <div class="fc-series-browse__grid">
-          <article
-            v-for="s in store.series" :key="s.id" class="fc-sbcard"
-            @click="manage(s.id)"
-          >
-            <div class="fc-sbcard__cover">
-              <img v-if="s.cover_thumbnail_url" :src="s.cover_thumbnail_url" alt="" loading="lazy" />
-              <div v-else class="fc-sbcard__cover--empty">
-                <v-icon size="large">mdi-book-open-page-variant</v-icon>
-              </div>
-              <span v-if="s.has_gap" class="fc-sbcard__gap" title="Has a missing-page gap">
-                <v-icon size="x-small">mdi-alert-outline</v-icon> gap
-              </span>
-            </div>
-            <div class="fc-sbcard__body">
-              <h3 class="fc-sbcard__name">{{ s.name }}</h3>
-              <div class="fc-sbcard__meta">
-                <span v-if="s.artist_name" class="fc-sbcard__artist">{{ s.artist_name }}</span>
-                <span class="fc-sbcard__counts">
-                  {{ s.chapter_count }} ch · {{ s.page_count }} pg
-                </span>
-              </div>
-              <div class="fc-sbcard__actions">
-                <v-btn
-                  size="x-small" variant="tonal" color="accent"
-                  prepend-icon="mdi-book-open"
-                  :disabled="s.page_count === 0"
-                  @click.stop="read(s.id)"
-                >Read</v-btn>
-                <v-btn
-                  size="x-small" variant="text"
-                  prepend-icon="mdi-pencil" @click.stop="manage(s.id)"
-                >Manage</v-btn>
-              </div>
-            </div>
-          </article>
-        </div>
-
-        <div v-if="store.loading" class="fc-series-browse__sentinel">
-          <v-progress-circular indeterminate color="accent" size="28" />
-        </div>
-      </v-window-item>
-
-      <!-- ---- Suggestions ---- -->
-      <v-window-item value="suggestions">
-        <div class="fc-sug__controls">
-          <v-switch
-            :model-value="sug.enabled" color="accent" density="compact"
-            hide-details label="Matching on"
-            @update:model-value="sug.setEnabled($event)"
-          />
-          <v-text-field
-            :model-value="sug.threshold"
-            type="number" min="0" max="1" step="0.05"
-            density="compact" variant="outlined" hide-details
-            label="Threshold" style="max-width: 130px;"
-            @change="sug.setThreshold(Number($event.target.value))"
-          />
-          <v-spacer />
-          <v-btn
-            variant="tonal" prepend-icon="mdi-radar" @click="sug.rescan()"
-          >Rescan</v-btn>
-        </div>
-
-        <v-alert v-if="sug.error" type="error" variant="tonal" closable class="my-4">
-          {{ sug.error }}
-        </v-alert>
-        <div
-          v-else-if="!sug.loading && sug.suggestions.length === 0"
-          class="fc-series-browse__empty"
-        >
-          No pending suggestions. Click <strong>Rescan</strong> to look for posts
-          that continue an existing series.
-        </div>
-
-        <div class="fc-sug__list">
-          <div v-for="s in sug.suggestions" :key="s.id" class="fc-sug__row">
-            <div class="fc-sug__main">
-              <div class="fc-sug__title">
-                {{ s.post_title }}
-                <v-icon size="x-small">mdi-arrow-right</v-icon>
-                <span class="fc-sug__series">{{ s.series_name }}</span>
-              </div>
-              <div class="fc-sug__signals">
-                <span class="fc-sug__score">{{ pct(s.score) }}</span>
-                <span
-                  v-for="k in signalKeys(s)" :key="k" class="fc-sug__chip"
-                >{{ k }} {{ pct(s.signals[k]) }}</span>
-              </div>
-            </div>
-            <div class="fc-sug__actions">
-              <v-btn
-                size="small" color="accent" variant="flat"
-                @click="sug.accept(s.id)"
-              >Add</v-btn>
-              <v-btn size="small" variant="text" @click="sug.dismiss(s.id)">Skip</v-btn>
-            </div>
-          </div>
-        </div>
-      </v-window-item>
-    </v-window>
-  </v-container>
-</template>
-
-<script setup>
-import { onMounted, ref } from 'vue'
-import { useRouter } from 'vue-router'
-import { useSeriesBrowseStore } from '../stores/seriesBrowse.js'
-import { useSeriesSuggestionsStore } from '../stores/seriesSuggestions.js'
-
-const router = useRouter()
-const store = useSeriesBrowseStore()
-const sug = useSeriesSuggestionsStore()
-const tab = ref('browse')
-
-const sortItems = [
-  { title: 'Recently updated', value: 'recent' },
-  { title: 'Name', value: 'name' },
-  { title: 'Size', value: 'size' }
-]
-
-function manage(id) {
-  router.push({ name: 'series-manage', params: { tagId: id } })
-}
-function read(id) {
-  router.push({ name: 'series-read', params: { tagId: id } })
-}
-
-function pct(v) { return `${Math.round((v || 0) * 100)}%` }
-// Only the signals that actually fired (strength > 0), in a stable order.
-function signalKeys(s) {
-  return ['title', 'artist', 'pages', 'tags'].filter(k => (s.signals?.[k] || 0) > 0)
-}
-
-onMounted(() => {
-  store.load()
-  sug.loadSettings()
-  sug.load()
-})
-</script>
-
-<style scoped>
-.fc-series-browse__controls {
-  display: flex; gap: 12px; margin-bottom: 16px;
-}
-.fc-series-browse__empty {
-  padding: 48px; text-align: center;
-  color: rgb(var(--v-theme-on-surface-variant));
-}
-.fc-series-browse__grid {
-  display: grid;
-  grid-template-columns: repeat(auto-fill, minmax(220px, 1fr));
-  gap: 14px;
-}
-.fc-sbcard {
-  border: 1px solid rgb(var(--v-theme-surface-light));
-  border-radius: 8px; overflow: hidden; cursor: pointer;
-  background: rgb(var(--v-theme-surface));
-  transition: border-color 120ms ease, transform 120ms ease;
-}
-.fc-sbcard:hover {
-  border-color: rgb(var(--v-theme-accent), 0.6);
-  transform: translateY(-2px);
-}
-.fc-sbcard__cover {
-  position: relative; aspect-ratio: 3 / 2; background: rgb(var(--v-theme-surface-light));
-}
-.fc-sbcard__cover img { width: 100%; height: 100%; object-fit: cover; display: block; }
-.fc-sbcard__cover--empty {
-  width: 100%; height: 100%;
-  display: flex; align-items: center; justify-content: center;
-  color: rgb(var(--v-theme-on-surface-variant));
-}
-.fc-sbcard__gap {
-  position: absolute; top: 6px; right: 6px;
-  display: inline-flex; align-items: center; gap: 2px;
-  padding: 1px 6px; border-radius: 999px; font-size: 11px;
-  background: rgba(20, 23, 26, 0.75);
-  color: rgb(var(--v-theme-warning, var(--v-theme-accent)));
-}
-.fc-sbcard__body { padding: 8px 10px; }
-.fc-sbcard__name {
-  font-family: 'Fraunces', Georgia, serif; font-size: 15px; font-weight: 600;
-  margin: 0 0 4px; color: rgb(var(--v-theme-on-surface));
-  overflow: hidden; text-overflow: ellipsis; white-space: nowrap;
-}
-.fc-sbcard__meta {
-  display: flex; justify-content: space-between; gap: 8px;
-  font-size: 12px; color: rgb(var(--v-theme-on-surface-variant));
-  margin-bottom: 8px;
-}
-.fc-sbcard__artist { overflow: hidden; text-overflow: ellipsis; white-space: nowrap; }
-.fc-sbcard__counts { flex: 0 0 auto; font-variant-numeric: tabular-nums; }
-.fc-sbcard__actions { display: flex; gap: 6px; }
-
-/* Suggestions tab */
-.fc-sug__controls {
-  display: flex; align-items: center; gap: 16px; margin-bottom: 12px;
-}
-.fc-sug__list { display: flex; flex-direction: column; gap: 8px; }
-.fc-sug__row {
-  display: flex; align-items: center; gap: 12px;
-  padding: 10px 12px; border-radius: 8px;
-  border: 1px solid rgb(var(--v-theme-surface-light));
-  background: rgb(var(--v-theme-surface));
-}
-.fc-sug__main { flex: 1; min-width: 0; }
-.fc-sug__title {
-  display: flex; align-items: center; gap: 6px;
-  font-size: 14px; color: rgb(var(--v-theme-on-surface));
-  overflow: hidden;
-}
-.fc-sug__series { color: rgb(var(--v-theme-accent)); font-weight: 600; }
-.fc-sug__signals {
-  display: flex; flex-wrap: wrap; align-items: center; gap: 6px; margin-top: 4px;
-}
-.fc-sug__score {
-  font-family: 'JetBrains Mono', monospace; font-size: 12px; font-weight: 700;
-  color: rgb(var(--v-theme-accent));
-}
-.fc-sug__chip {
-  font-size: 11px; padding: 1px 7px; border-radius: 999px;
-  background: rgb(var(--v-theme-surface-light));
-  color: rgb(var(--v-theme-on-surface-variant));
-}
-.fc-sug__actions { flex: 0 0 auto; display: flex; gap: 6px; }
-</style>

frontend/test/components/postCard.spec.js

@@ -40,30 +40,6 @@ describe('PostCard', () => {
     expect(full.text()).not.toContain('Show more')
   })
 
-  const thumbs = (n) =>
-    Array.from({ length: n }, (_, i) => ({ image_id: 100 + i, thumbnail_url: `/t${i}` }))
-
-  it('fills the rail to full width (5 cells, no overflow) for a 6-image post', () => {
-    const post = { ...BASE, description_plain: 'd', thumbnails: thumbs(6), thumbnails_more: 0 }
-    const w = mountComponent(PostCard, { props: { post }, pinia: freshPinia() })
-    // hero + 5 rail thumbnails, no "+N" tile.
-    expect(w.findAll('.fc-post-card__rail-cell')).toHaveLength(5)
-    expect(w.find('.fc-post-card__rail-more').exists()).toBe(false)
-    // The grid spans the hero width via a column count == cells shown.
-    expect(w.find('.fc-post-card__rail').attributes('style')).toContain('--fc-rail-cols: 5')
-  })
-
-  it('reserves the last cell for a "+N" overflow tile when there are more images', () => {
-    const post = { ...BASE, description_plain: 'd', thumbnails: thumbs(6), thumbnails_more: 4 }
-    const w = mountComponent(PostCard, { props: { post }, pinia: freshPinia() })
-    // 4 thumbnails + 1 overflow tile == 5 columns; tile counts every hidden image.
-    expect(w.findAll('.fc-post-card__rail-cell')).toHaveLength(4)
-    const more = w.find('.fc-post-card__rail-more')
-    expect(more.exists()).toBe(true)
-    expect(more.text()).toBe('+5')
-    expect(w.find('.fc-post-card__rail').attributes('style')).toContain('--fc-rail-cols: 5')
-  })
-
   it('clicking an image opens the post-scoped modal (never expands the card)', async () => {
     const pinia = freshPinia()
     const modal = useModalStore()

frontend/test/seriesManage.spec.js

@@ -13,19 +13,6 @@ function stubFetch(handler) {
   })
 }
 
-const SERIES_BODY = {
-  series: { id: 7, name: 'V' },
-  chapters: [
-    {
-      id: 1, chapter_number: 1, title: null, is_placeholder: false,
-      stated_page_start: null, stated_page_end: null,
-      pages: [{ image_id: 1, page_number: 1, thumbnail_url: 't' }]
-    }
-  ],
-  gaps: [],
-  pages: [{ image_id: 1, page_number: 1, thumbnail_url: 't' }]
-}
-
 describe('seriesManage', () => {
   beforeEach(() => setActivePinia(createPinia()))
   afterEach(() => vi.restoreAllMocks())
@@ -35,88 +22,48 @@ describe('seriesManage', () => {
     expect(moveItem([1, 2, 3], 0, 2)).toEqual([2, 3, 1])
   })
 
-  it('load fetches chapters + series and picks a default target', async () => {
+  it('load fetches pages + series', async () => {
     const s = useSeriesManageStore()
-    stubFetch(() => ({ status: 200, body: SERIES_BODY }))
+    stubFetch(() => ({
+      status: 200,
+      body: { series: { id: 7, name: 'V' },
+              pages: [{ image_id: 1, page_number: 1, thumbnail_url: 't' }] }
+    }))
     await s.load(7)
     expect(s.series).toEqual({ id: 7, name: 'V' })
-    expect(s.chapters.map(c => c.id)).toEqual([1])
-    expect(s.pageCount).toBe(1)
-    expect(s.targetChapterId).toBe(1)
+    expect(s.pages.map(p => p.image_id)).toEqual([1])
   })
 
-  it('reorderPages posts ordered ids to the chapter reorder route', async () => {
+  it('reorder posts the full ordered id list', async () => {
     const s = useSeriesManageStore()
     s.tagId = 7
+    s.pages = [{ image_id: 1 }, { image_id: 2 }, { image_id: 3 }]
     const calls = []
     stubFetch((url, init) => {
       calls.push({ url, body: init.body ? JSON.parse(init.body) : null })
       if (url.includes('/reorder')) return { status: 200, body: { ok: true } }
-      return { status: 200, body: SERIES_BODY }
+      return { status: 200, body: { series: { id: 7, name: 'V' }, pages: [] } }
     })
-    await s.reorderPages(3, [30, 10, 20])
-    const r = calls.find(c => c.url.includes('/chapters/3/reorder'))
-    expect(r.url).toContain('/api/series/7/chapters/3/reorder')
-    expect(r.body).toEqual({ image_ids: [30, 10, 20] })
+    await s.reorder([3, 1, 2])
+    const r = calls.find(c => c.url.includes('/reorder'))
+    expect(r.url).toContain('/api/series/7/reorder')
+    expect(r.body).toEqual({ image_ids: [3, 1, 2] })
   })
 
-  it('moveChapter reorders via the chapter id list', async () => {
+  it('addSelected posts picker selection then reloads', async () => {
     const s = useSeriesManageStore()
     s.tagId = 7
-    s.chapters = [{ id: 1 }, { id: 2 }, { id: 3 }]
-    const calls = []
-    stubFetch((url, init) => {
-      calls.push({ url, body: init.body ? JSON.parse(init.body) : null })
-      if (url.includes('/chapters/reorder')) return { status: 200, body: { ok: true } }
-      return { status: 200, body: SERIES_BODY }
-    })
-    await s.moveChapter(2, -1)
-    const r = calls.find(c => c.url.includes('/chapters/reorder'))
-    expect(r.body).toEqual({ chapter_ids: [2, 1, 3] })
-  })
-
-  it('setChapterPart patches stated_part on the chapter', async () => {
-    const s = useSeriesManageStore()
-    s.tagId = 7
-    const calls = []
-    stubFetch((url, init) => {
-      calls.push({ url, method: init.method, body: init.body ? JSON.parse(init.body) : null })
-      if (init.method === 'PATCH') return { status: 200, body: { ok: true } }
-      return { status: 200, body: SERIES_BODY }
-    })
-    await s.setChapterPart(1, 2)
-    const p = calls.find(c => c.method === 'PATCH')
-    expect(p.url).toContain('/api/series/7/chapters/1')
-    expect(p.body).toEqual({ stated_part: 2 })
-  })
-
-  it('load surfaces part_gaps and partGapAfter looks them up', async () => {
-    const s = useSeriesManageStore()
-    stubFetch(() => ({
-      status: 200,
-      body: { ...SERIES_BODY, part_gaps: [{ after_chapter_id: 1, start: 2, end: 2 }] }
-    }))
-    await s.load(7)
-    expect(s.partGaps).toHaveLength(1)
-    expect(s.partGapAfter(1)).toEqual({ after_chapter_id: 1, start: 2, end: 2 })
-    expect(s.partGapAfter(99)).toBeNull()
-  })
-
-  it('addSelected posts selection + target chapter then clears', async () => {
-    const s = useSeriesManageStore()
-    s.tagId = 7
-    s.targetChapterId = 5
     s.pickerSelection = [9, 10]
     const calls = []
     stubFetch((url, init) => {
       calls.push({ url, body: init.body ? JSON.parse(init.body) : null })
-      if (url.endsWith('/pages') && init.method === 'POST')
+      if (url.includes('/pages') && init.method === 'POST')
         return { status: 200, body: { added_count: 2 } }
-      return { status: 200, body: SERIES_BODY }
+      return { status: 200, body: { series: { id: 7, name: 'V' }, pages: [] } }
     })
     await s.addSelected()
     const add = calls.find(c => c.url.endsWith('/api/series/7/pages'))
-    expect(add.body).toEqual({ image_ids: [9, 10], chapter_id: 5 })
+    expect(add.body).toEqual({ image_ids: [9, 10] })
     expect(s.pickerSelection).toEqual([])
   })
 })

frontend/test/sources.spec.js

@@ -41,41 +41,6 @@ describe('sources store', () => {
     expect(calls[0]).toContain('artist_id=7')
   })
 
-  it('update patches the source in place without dropping the cache', async () => {
-    const s = useSourcesStore()
-    s.byArtist.set(null, [
-      { id: 5, artist_id: 5, enabled: true, backfill_state: null },
-      { id: 6, artist_id: 5, enabled: true },
-    ])
-    stubFetch(() => ({
-      status: 200,
-      body: { id: 5, artist_id: 5, enabled: false, backfill_state: null },
-    }))
-    await s.update(5, { enabled: false }, 5)
-    // Cache survives (no full reload) and only source 5 changed.
-    expect(s.byArtist.has(null)).toBe(true)
-    const arr = s.byArtist.get(null)
-    expect(arr.find(r => r.id === 5).enabled).toBe(false)
-    expect(arr.find(r => r.id === 6).enabled).toBe(true)
-  })
-
-  it('startBackfill patches backfill_state in place', async () => {
-    const s = useSourcesStore()
-    s.byArtist.set(null, [{ id: 5, artist_id: 5, backfill_state: null }])
-    stubFetch(() => ({ status: 200, body: { id: 5, backfill_state: 'running' } }))
-    await s.startBackfill(5, 5)
-    expect(s.byArtist.get(null).find(r => r.id === 5).backfill_state).toBe('running')
-  })
-
-  it('remove drops the source from the cache in place', async () => {
-    const s = useSourcesStore()
-    s.byArtist.set(null, [{ id: 5, artist_id: 5 }, { id: 6, artist_id: 5 }])
-    stubFetch(() => ({ status: 204, body: null }))
-    await s.remove(5, 5)
-    expect(s.byArtist.has(null)).toBe(true)
-    expect(s.byArtist.get(null).map(r => r.id)).toEqual([6])
-  })
-
   it('create invalidates the all-cache and the artist-cache', async () => {
     const s = useSourcesStore()
     s.byArtist.set(null, [])

tests/fixtures/patreon_posts_page1.json

@@ -1,121 +0,0 @@
-{
-  "data": [
-    {
-      "id": "1001",
-      "type": "post",
-      "attributes": {
-        "title": "Image gallery post",
-        "published_at": "2026-05-01T12:00:00.000+00:00",
-        "post_type": "image_file",
-        "content": "<p>just a gallery</p>",
-        "url": "https://www.patreon.com/posts/1001",
-        "image": {
-          "large_url": "https://c10.patreonusercontent.com/4/large/aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa/gallery-one.jpg",
-          "url": "https://c10.patreonusercontent.com/4/orig/aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa/gallery-one.jpg"
-        }
-      },
-      "relationships": {
-        "images": {
-          "data": [
-            {"id": "9001", "type": "media"},
-            {"id": "9002", "type": "media"}
-          ]
-        }
-      }
-    },
-    {
-      "id": "1002",
-      "type": "post",
-      "attributes": {
-        "title": "Attachment post",
-        "published_at": "2026-04-15T09:30:00.000+00:00",
-        "post_type": "text_only",
-        "content": "<p>here is a zip</p>",
-        "url": "https://www.patreon.com/posts/1002",
-        "post_file": {
-          "name": "bonus-pack.zip",
-          "url": "https://www.patreon.com/file?h=cccccccccccccccccccccccccccccccc&i=1002"
-        }
-      },
-      "relationships": {
-        "attachments_media": {
-          "data": [
-            {"id": "9003", "type": "media"}
-          ]
-        }
-      }
-    },
-    {
-      "id": "1003",
-      "type": "post",
-      "attributes": {
-        "title": "Inline content post",
-        "published_at": "2026-04-01T08:00:00.000+00:00",
-        "post_type": "text_only",
-        "content": "<p>look</p><figure><img src=\"https://c10.patreonusercontent.com/4/orig/dddddddddddddddddddddddddddddddd/inline.png?token=x&amp;v=2\" alt=\"inline\"></figure>",
-        "url": "https://www.patreon.com/posts/1003"
-      },
-      "relationships": {}
-    },
-    {
-      "id": "1004",
-      "type": "post",
-      "attributes": {
-        "title": "Video post",
-        "published_at": "2026-03-20T18:45:00.000+00:00",
-        "post_type": "video_external_file",
-        "content": "<p>watch</p>",
-        "url": "https://www.patreon.com/posts/1004",
-        "post_file": {
-          "name": "clip.m3u8",
-          "url": "https://stream.mux.com/eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee.m3u8?token=jwt"
-        }
-      },
-      "relationships": {}
-    }
-  ],
-  "included": [
-    {
-      "id": "9001",
-      "type": "media",
-      "attributes": {
-        "file_name": "gallery-one.jpg",
-        "download_url": "https://c10.patreonusercontent.com/4/orig/aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa/gallery-one.jpg",
-        "image_urls": {
-          "original": "https://c10.patreonusercontent.com/4/orig/aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa/gallery-one.jpg",
-          "default": "https://c10.patreonusercontent.com/4/default/aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa/gallery-one.jpg"
-        }
-      }
-    },
-    {
-      "id": "9002",
-      "type": "media",
-      "attributes": {
-        "file_name": "gallery-two.jpg",
-        "download_url": "https://c10.patreonusercontent.com/4/orig/bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb/gallery-two.jpg",
-        "image_urls": {
-          "original": "https://c10.patreonusercontent.com/4/orig/bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb/gallery-two.jpg"
-        }
-      }
-    },
-    {
-      "id": "9003",
-      "type": "media",
-      "attributes": {
-        "file_name": "bonus-pack.zip",
-        "download_url": "https://www.patreon.com/file/download?h=cccccccccccccccccccccccccccccccc"
-      }
-    },
-    {
-      "id": "5555",
-      "type": "campaign",
-      "attributes": {
-        "name": "Example Creator",
-        "url": "https://www.patreon.com/example"
-      }
-    }
-  ],
-  "links": {
-    "next": "https://www.patreon.com/api/posts?filter%5Bcampaign_id%5D=5555&page%5Bcursor%5D=NEXTCURSOR123&sort=-published_at"
-  }
-}

tests/test_api_credentials.py

@@ -155,8 +155,6 @@ async def test_verify_no_enabled_source_is_untestable(client):
 
 @pytest.mark.asyncio
 async def test_verify_runs_gallery_dl_and_stamps_on_success(client, db, monkeypatch):
-    """A gallery-dl platform routes through GalleryDLService.verify; success
-    stamps last_verified (platform-agnostic endpoint behavior)."""
     from backend.app.models import Artist, Source
     from backend.app.services import gallery_dl as gdl_mod
 
@@ -165,45 +163,6 @@ async def test_verify_runs_gallery_dl_and_stamps_on_success(client, db, monkeypa
         return (True, "Credentials valid — the feed authenticated.")
     monkeypatch.setattr(gdl_mod.GalleryDLService, "verify", _fake_verify)
 
-    await client.post("/api/credentials", json={
-        "platform": "subscribestar", "credential_type": "cookies", "data": _NETSCAPE,
-    })
-    artist = Artist(name="Maewix", slug="maewix")
-    db.add(artist)
-    await db.flush()
-    db.add(Source(
-        artist_id=artist.id, platform="subscribestar",
-        url="https://www.subscribestar.com/maewix", enabled=True, config_overrides={},
-    ))
-    await db.commit()
-
-    resp = await client.post("/api/credentials/subscribestar/verify")
-    body = await resp.get_json()
-    assert body["valid"] is True
-    assert body["last_verified"] is not None
-
-    # The stamp is persisted on the credential record.
-    rec = await (await client.get("/api/credentials/subscribestar")).get_json()
-    assert rec["last_verified"] is not None
-
-
-@pytest.mark.asyncio
-async def test_verify_patreon_uses_native_ingester_not_gallery_dl(client, db, monkeypatch):
-    """plan #697 cutover: Patreon credential verify routes through the native
-    ingester (verify_patreon_credential), NOT gallery-dl --simulate."""
-    from backend.app.models import Artist, Source
-    from backend.app.services import gallery_dl as gdl_mod
-    from backend.app.services import patreon_ingester as pi_mod
-
-    async def _native_verify(url, cookies_path, overrides):
-        return (True, "Credentials valid — the Patreon feed authenticated.")
-    monkeypatch.setattr(pi_mod, "verify_patreon_credential", _native_verify)
-
-    # gallery-dl must NOT be consulted for Patreon.
-    async def _boom(self, *args, **kwargs):
-        raise AssertionError("gallery-dl verify must not run for patreon")
-    monkeypatch.setattr(gdl_mod.GalleryDLService, "verify", _boom)
-
     await client.post("/api/credentials", json={
         "platform": "patreon", "credential_type": "cookies", "data": _NETSCAPE,
     })
@@ -221,6 +180,10 @@ async def test_verify_patreon_uses_native_ingester_not_gallery_dl(client, db, mo
     assert body["valid"] is True
     assert body["last_verified"] is not None
 
+    # The stamp is persisted on the credential record.
+    rec = await (await client.get("/api/credentials/patreon")).get_json()
+    assert rec["last_verified"] is not None
+
 
 @pytest.mark.asyncio
 async def test_verify_reports_auth_failure(client, db, monkeypatch):

tests/test_api_sources.py

@@ -5,20 +5,6 @@ from backend.app.models import Artist, Source
 pytestmark = pytest.mark.integration
 
 
-@pytest.fixture(autouse=True)
-def _stub_preflight_verify(monkeypatch):
-    """Backfill/recover arms run a pre-flight credential verify (plan #703 #2)
-    which for Patreon would hit the network. Default it to 'valid' so the
-    endpoint tests stay network-free; the dedicated pre-flight tests override
-    this with a rejection/inconclusive verdict."""
-    from backend.app.services import download_backends as db_mod
-
-    async def _ok(**kwargs):
-        return (True, "Credentials valid")
-
-    monkeypatch.setattr(db_mod, "verify_source_credential", _ok)
-
-
 @pytest.fixture
 async def artist(db):
     a = Artist(name="Alice", slug="alice")
@@ -89,39 +75,6 @@ async def test_create_list_get_delete(client, artist):
     assert (await client.get(f"/api/sources/{sid}")).status_code == 404
 
 
-@pytest.mark.asyncio
-async def test_create_kicks_off_backfill_for_enabled_source(client, artist, monkeypatch):
-    """A new enabled source dispatches its first walk immediately (no waiting
-    for the next scheduler tick), unless its platform is in cooldown."""
-    delays: list = []
-    monkeypatch.setattr(
-        "backend.app.tasks.download.download_source.delay",
-        lambda *a, **k: delays.append(a),
-    )
-    resp = await client.post("/api/sources", json={
-        "artist_id": artist.id, "platform": "patreon",
-        "url": "https://patreon.com/kickoff",
-    })
-    assert resp.status_code == 201
-    sid = (await resp.get_json())["id"]
-    assert delays == [(sid,)]
-
-
-@pytest.mark.asyncio
-async def test_create_disabled_source_does_not_kick_off(client, artist, monkeypatch):
-    delays: list = []
-    monkeypatch.setattr(
-        "backend.app.tasks.download.download_source.delay",
-        lambda *a, **k: delays.append(a),
-    )
-    resp = await client.post("/api/sources", json={
-        "artist_id": artist.id, "platform": "patreon",
-        "url": "https://patreon.com/disabled", "enabled": False,
-    })
-    assert resp.status_code == 201
-    assert delays == []
-
-
 @pytest.mark.asyncio
 async def test_create_rejects_unknown_platform(client, artist):
     resp = await client.post("/api/sources", json={
@@ -246,7 +199,7 @@ async def test_list_derives_next_check_at_when_last_checked_set(
 
 
 @pytest.mark.asyncio
-async def test_backfill_endpoint_start_and_stop(client, artist, db):
+async def test_backfill_endpoint_arms_source(client, artist, db):
     src = Source(
         artist_id=artist.id, platform="patreon",
         url="https://patreon.com/alice-backfill", enabled=True,
@@ -255,21 +208,18 @@ async def test_backfill_endpoint_start_and_stop(client, artist, db):
     await db.commit()
     sid = src.id
 
-    resp = await client.post(f"/api/sources/{sid}/backfill", json={"action": "start"})
+    resp = await client.post(f"/api/sources/{sid}/backfill", json={"runs": 5})
     assert resp.status_code == 200
     body = await resp.get_json()
-    assert body["backfill_state"] == "running"
+    assert body["backfill_runs_remaining"] == 5
 
-    # GET reflects the running state.
+    # GET reflects the new state.
     one = await client.get(f"/api/sources/{sid}")
-    assert (await one.get_json())["backfill_state"] == "running"
-
-    stopped = await client.post(f"/api/sources/{sid}/backfill", json={"action": "stop"})
-    assert (await stopped.get_json())["backfill_state"] is None
+    assert (await one.get_json())["backfill_runs_remaining"] == 5
 
 
 @pytest.mark.asyncio
-async def test_backfill_endpoint_defaults_to_start(client, artist, db):
+async def test_backfill_endpoint_defaults_to_three(client, artist, db):
     src = Source(
         artist_id=artist.id, platform="patreon",
         url="https://patreon.com/alice-backfill-default", enabled=True,
@@ -278,136 +228,26 @@ async def test_backfill_endpoint_defaults_to_start(client, artist, db):
     await db.commit()
     resp = await client.post(f"/api/sources/{src.id}/backfill", json={})
     body = await resp.get_json()
-    assert body["backfill_state"] == "running"
+    assert body["backfill_runs_remaining"] == 3
 
 
 @pytest.mark.asyncio
-async def test_backfill_endpoint_rejects_bad_action(client, artist, db):
+async def test_backfill_endpoint_rejects_out_of_range(client, artist, db):
     src = Source(
         artist_id=artist.id, platform="patreon",
         url="https://patreon.com/alice-backfill-bad", enabled=True,
     )
     db.add(src)
     await db.commit()
-    bad = await client.post(f"/api/sources/{src.id}/backfill", json={"action": "nope"})
+    bad = await client.post(f"/api/sources/{src.id}/backfill", json={"runs": 0})
     assert bad.status_code == 400
+    too_big = await client.post(
+        f"/api/sources/{src.id}/backfill", json={"runs": 99}
+    )
+    assert too_big.status_code == 400
 
 
 @pytest.mark.asyncio
 async def test_backfill_endpoint_404_when_source_missing(client):
-    resp = await client.post("/api/sources/999999/backfill", json={"action": "start"})
+    resp = await client.post("/api/sources/999999/backfill", json={"runs": 3})
     assert resp.status_code == 404
-
-
-@pytest.mark.asyncio
-async def test_backfill_endpoint_recover_arms_bypass(client, artist, db):
-    """Plan #697: action='recover' arms a backfill that bypasses the Patreon
-    seen-ledger; the response exposes backfill_bypass_seen=True so the UI badge
-    can label it 'Recovering'. Stop clears it."""
-    src = Source(
-        artist_id=artist.id, platform="patreon",
-        url="https://patreon.com/alice-recover", enabled=True,
-    )
-    db.add(src)
-    await db.commit()
-    sid = src.id
-
-    resp = await client.post(f"/api/sources/{sid}/backfill", json={"action": "recover"})
-    assert resp.status_code == 200
-    body = await resp.get_json()
-    assert body["backfill_state"] == "running"
-    assert body["backfill_bypass_seen"] is True
-
-    stopped = await client.post(f"/api/sources/{sid}/backfill", json={"action": "stop"})
-    stopped_body = await stopped.get_json()
-    assert stopped_body["backfill_state"] is None
-    assert stopped_body["backfill_bypass_seen"] is False
-
-
-# --- Plan #703 #2: pre-flight credential verify on backfill/recover arm -----
-
-
-@pytest.mark.asyncio
-async def test_arm_blocked_when_credential_rejected(client, artist, db, monkeypatch):
-    """A definitively-rejected credential refuses the arm (409 + reason) instead
-    of starting a doomed walk; the source is NOT armed."""
-    from backend.app.services import download_backends as db_mod
-
-    async def _reject(**kwargs):
-        return (False, "Patreon rejected the credential — cookies expired")
-    monkeypatch.setattr(db_mod, "verify_source_credential", _reject)
-
-    src = Source(
-        artist_id=artist.id, platform="patreon",
-        url="https://patreon.com/alice-preflight", enabled=True,
-    )
-    db.add(src)
-    await db.commit()
-    sid = src.id
-
-    resp = await client.post(f"/api/sources/{sid}/backfill", json={"action": "recover"})
-    assert resp.status_code == 409
-    assert "expired" in ((await resp.get_json()).get("detail") or "")
-
-    # Not armed.
-    one = await (await client.get(f"/api/sources/{sid}")).get_json()
-    assert one["backfill_state"] is None
-
-
-@pytest.mark.asyncio
-async def test_arm_proceeds_when_credential_inconclusive(client, artist, db, monkeypatch):
-    """An inconclusive verify (network blip / drift) must NOT block the arm."""
-    from backend.app.services import download_backends as db_mod
-
-    async def _inconclusive(**kwargs):
-        return (None, "couldn't verify (network)")
-    monkeypatch.setattr(db_mod, "verify_source_credential", _inconclusive)
-
-    src = Source(
-        artist_id=artist.id, platform="patreon",
-        url="https://patreon.com/alice-incon", enabled=True,
-    )
-    db.add(src)
-    await db.commit()
-
-    resp = await client.post(f"/api/sources/{src.id}/backfill", json={"action": "start"})
-    assert resp.status_code == 200
-    assert (await resp.get_json())["backfill_state"] == "running"
-
-
-@pytest.mark.asyncio
-async def test_stop_never_pre_flights(client, artist, db, monkeypatch):
-    from backend.app.services import download_backends as db_mod
-
-    async def _boom(**kwargs):
-        raise AssertionError("stop must not pre-flight verify")
-    monkeypatch.setattr(db_mod, "verify_source_credential", _boom)
-
-    src = Source(
-        artist_id=artist.id, platform="patreon",
-        url="https://patreon.com/alice-stop", enabled=True,
-    )
-    db.add(src)
-    await db.commit()
-    resp = await client.post(f"/api/sources/{src.id}/backfill", json={"action": "stop"})
-    assert resp.status_code == 200
-
-
-@pytest.mark.asyncio
-async def test_gallery_dl_platform_arm_skips_pre_flight(client, artist, db, monkeypatch):
-    """Pre-flight is gated to native-ingester platforms (cheap verify); a
-    gallery-dl source arms without the slow --simulate probe."""
-    from backend.app.services import download_backends as db_mod
-
-    async def _boom(**kwargs):
-        raise AssertionError("gallery-dl arm must not pre-flight verify")
-    monkeypatch.setattr(db_mod, "verify_source_credential", _boom)
-
-    src = Source(
-        artist_id=artist.id, platform="subscribestar",
-        url="https://www.subscribestar.com/x", enabled=True,
-    )
-    db.add(src)
-    await db.commit()
-    resp = await client.post(f"/api/sources/{src.id}/backfill", json={"action": "start"})
-    assert resp.status_code == 200

tests/test_api_tags.py

@@ -52,12 +52,11 @@ async def test_create_tag_missing_name_400(client):
 
 @pytest.mark.asyncio
 async def test_create_tag_name_only_defaults_to_general(client):
-    """IR-style: name without kind and without `kind:` prefix → general.
-    #701: operator-entered names are Title-Cased at the create endpoint."""
+    """IR-style: name without kind and without `kind:` prefix → general."""
     resp = await client.post("/api/tags", json={"name": "sunset"})
     assert resp.status_code == 201
     body = await resp.get_json()
-    assert body["name"] == "Sunset"
+    assert body["name"] == "sunset"
     assert body["kind"] == "general"
 
 
@@ -74,23 +73,21 @@ async def test_create_tag_with_kind_prefix(client):
 @pytest.mark.asyncio
 async def test_explicit_kind_overrides_prefix_parsing(client):
     """If caller passes explicit kind, don't re-parse the name —
-    colon and prefix stay literal. #701: first letter capitalized, tail
-    preserved (acronym-safe), so the 'S' in Saber survives."""
+    colon and prefix stay literal."""
     resp = await client.post(
         "/api/tags", json={"name": "character:Saber", "kind": "general"}
     )
     assert resp.status_code == 201
     body = await resp.get_json()
-    assert body["name"] == "Character:Saber"
+    assert body["name"] == "character:Saber"
     assert body["kind"] == "general"
 
 
 @pytest.mark.asyncio
 async def test_unknown_prefix_kept_literal(client):
-    """`http:example` — `http` not in KNOWN_KINDS → kind=general, literal name.
-    #701: Title-Cased as one word (no internal whitespace to split on)."""
+    """`http:example` — `http` not in KNOWN_KINDS → kind=general, literal name."""
     resp = await client.post("/api/tags", json={"name": "http://example.com"})
     assert resp.status_code == 201
     body = await resp.get_json()
-    assert body["name"] == "Http://example.com"
+    assert body["name"] == "http://example.com"
     assert body["kind"] == "general"

tests/test_archive_extractor.py

@@ -39,24 +39,6 @@ def test_cbz_alias(tmp_path):
         assert [n for n, _ in members] == ["p1.jpg"]
 
 
-def test_misnamed_zip_detected_and_extracted(tmp_path):
-    """A real zip whose filename has no usable extension (Patreon attachment
-    URL-blob name) is detected by magic bytes and its members extracted —
-    previously it was filed as an opaque attachment and never opened."""
-    z = tmp_path / "01_https___www.patreon.com_media-u_v3_131083093"
-    _zip(z, {"a.jpg": b"img", "b.png": b"img2"})
-    assert is_archive(z)
-    with extract_archive(z) as members:
-        assert sorted(n for n, _ in members) == ["a.jpg", "b.png"]
-
-
-def test_non_archive_with_dotted_name_is_not_archive(tmp_path):
-    """A non-archive file with a dotted/extension-less name isn't misdetected."""
-    f = tmp_path / "01_https___www.patreon.com_media-u_v3_999"
-    f.write_bytes(b"\x89PNG\r\n\x1a\n not really but not a zip")
-    assert not is_archive(f)
-
-
 def test_corrupt_archive_yields_nothing(tmp_path):
     bad = tmp_path / "broken.zip"
     bad.write_bytes(b"not a zip at all")

tests/test_backup_service.py

@@ -5,7 +5,6 @@ without external binaries. The real subprocess behavior is exercised
 implicitly via the Celery task tests in test_tasks_backup.py.
 """
 import json
-import subprocess
 from pathlib import Path
 
 import pytest
@@ -17,9 +16,9 @@ pytestmark = pytest.mark.integration
 
 @pytest.fixture
 def fake_subprocess(monkeypatch):
-    """Fake both subprocess.run (restore path) AND subprocess.Popen (the
-    bounded-kill backup path) so tests run without external binaries. Each
-    writes the target sentinel and records the cmd in a shared list."""
+    """Replace subprocess.run with a fake that writes a sentinel to
+    the target path (for pg_dump's -f, for tar's -cf). Captures all
+    calls in a list."""
     calls = []
 
     class _FakeProc:
@@ -27,33 +26,17 @@ def fake_subprocess(monkeypatch):
         stdout = b""
         stderr = b""
 
-    def _write_sentinel(cmd):
+    def _fake_run(cmd, **kwargs):
+        calls.append(list(cmd))
         if cmd[0] == "pg_dump":
             i = cmd.index("-f")
             Path(cmd[i + 1]).write_bytes(b"-- fake pg_dump\n")
         elif cmd[0] == "tar" and "-cf" in cmd:
             i = cmd.index("-cf")
             Path(cmd[i + 1]).write_bytes(b"fake tar payload")
-
-    def _fake_run(cmd, **kwargs):
-        calls.append(list(cmd))
-        _write_sentinel(cmd)
         return _FakeProc()
 
-    class _FakePopen:
-        def __init__(self, cmd, **kwargs):
-            calls.append(list(cmd))
-            self.returncode = 0
-            _write_sentinel(cmd)
-
-        def communicate(self, timeout=None):
-            return (b"", b"")
-
-        def kill(self):
-            self.returncode = -9
-
     monkeypatch.setattr("subprocess.run", _fake_run)
-    monkeypatch.setattr("subprocess.Popen", _FakePopen)
     return calls
 
 
@@ -215,41 +198,3 @@ def test_backups_dir_created_on_first_use(tmp_path):
     d = backup_service._backups_dir(tmp_path)
     assert d.is_dir()
     assert d.name == "_backups"
-
-
-# --- bounded-kill + local-temp (FC #739) -----------------------------
-
-
-def test_backup_db_dumps_to_local_temp_not_nfs_backups_dir(tmp_path, fake_subprocess):
-    """pg_dump must target a LOCAL temp path, not the (NFS) _backups dir —
-    so its long phase can't hang uninterruptibly on an NFS write."""
-    backup_service.backup_db(db_url="postgresql://u@h/d", images_root=tmp_path)
-    cmd = fake_subprocess[0]
-    dump_target = cmd[cmd.index("-f") + 1]
-    assert "_backups" not in dump_target
-    # The finished file still ends up in _backups (moved there).
-    result = backup_service.backup_db(
-        db_url="postgresql://u@h/d", images_root=tmp_path,
-    )
-    assert "_backups" in result["sql_path"]
-
-
-def test_run_bounded_fails_fast_when_unkillable(monkeypatch):
-    """A child stuck in D-state (communicate keeps timing out even after kill)
-    must NOT block the reaper — _run_bounded kills then re-raises promptly."""
-    killed = {"n": 0}
-
-    class _Hang:
-        def __init__(self, cmd, **kwargs):
-            pass
-
-        def communicate(self, timeout=None):
-            raise subprocess.TimeoutExpired(cmd="x", timeout=timeout or 0)
-
-        def kill(self):
-            killed["n"] += 1
-
-    monkeypatch.setattr("subprocess.Popen", _Hang)
-    with pytest.raises(subprocess.TimeoutExpired):
-        backup_service._run_bounded(["pg_dump"], 1)
-    assert killed["n"] == 1

tests/test_celery_routing.py

@@ -1,20 +0,0 @@
-"""Queue routing — the long one-shot maintenance tasks run on a dedicated
-`maintenance_long` lane so a 30-min backup or a multi-chunk audit can't starve
-the quick recovery sweeps / vacuum on the concurrency-1 `maintenance` lane
-(operator-flagged 2026-06-07)."""
-from backend.app.celery_app import celery
-
-
-def test_long_one_shots_route_to_maintenance_long():
-    routes = celery.conf.task_routes
-    for prefix in (
-        "backend.app.tasks.backup.*",
-        "backend.app.tasks.admin.*",
-        "backend.app.tasks.library_audit.*",
-    ):
-        assert routes[prefix]["queue"] == "maintenance_long"
-
-
-def test_quick_maintenance_stays_on_maintenance():
-    routes = celery.conf.task_routes
-    assert routes["backend.app.tasks.maintenance.*"]["queue"] == "maintenance"

tests/test_cleanup_service.py

@@ -9,7 +9,6 @@ import pytest
 from sqlalchemy import func, select
 
 from backend.app.models import Artist, ImageRecord, Tag, TagKind
-from backend.app.models.series_chapter import SeriesChapter
 from backend.app.models.series_page import SeriesPage
 from backend.app.models.tag import image_tag
 from backend.app.services import cleanup_service
@@ -377,12 +376,7 @@ def test_reset_content_tagging_deletes_content_keeps_fandom_series(db_sync, tmp_
         {"image_record_id": img.id, "tag_id": c.id},
         {"image_record_id": img.id, "tag_id": s.id},  # series membership
     ]))
-    ch = SeriesChapter(series_tag_id=s.id, chapter_number=1)
-    db_sync.add(ch)
-    db_sync.flush()
-    db_sync.add(SeriesPage(
-        series_tag_id=s.id, chapter_id=ch.id, image_id=img.id, page_number=1
-    ))
+    db_sync.add(SeriesPage(series_tag_id=s.id, image_id=img.id, page_number=1))
     db_sync.commit()
 
     result = cleanup_service.reset_content_tagging(db_sync, dry_run=False)

tests/test_download_backends.py

@@ -1,23 +0,0 @@
-"""download_backends — the single predicate that routes a platform to the
-native ingester vs. gallery-dl. Pure, no DB."""
-
-from backend.app.services.download_backends import (
-    NATIVE_INGESTER_PLATFORMS,
-    uses_native_ingester,
-)
-
-
-def test_patreon_is_native():
-    assert uses_native_ingester("patreon") is True
-    assert "patreon" in NATIVE_INGESTER_PLATFORMS
-
-
-def test_gallery_dl_platforms_are_not_native():
-    # The five platforms still served by gallery-dl must NOT route to the
-    # native ingester — guards an accidental over-broad migration.
-    for platform in ("subscribestar", "hentaifoundry", "discord", "pixiv", "deviantart"):
-        assert uses_native_ingester(platform) is False
-
-
-def test_unknown_platform_is_not_native():
-    assert uses_native_ingester("nonsense") is False

tests/test_download_service.py

@@ -59,7 +59,7 @@ def _make_jpg(path: Path, split: str = "h"):
 def _make_fake_dl_result(
     *, success=True, written_paths=None, quarantined_paths=None,
     files_downloaded=0, error_type=None, error_message=None,
-    stdout="", stderr="", cursor=None, run_stats=None, posts_processed=0,
+    stdout="", stderr="",
 ):
     return SimpleNamespace(
         success=success,
@@ -78,22 +78,12 @@ def _make_fake_dl_result(
         duration_seconds=1.23,
         started_at="2026-05-20T14:00:00+00:00",
         completed_at="2026-05-20T14:01:00+00:00",
-        # plan #704: native ingester now returns the cursor + run_stats
-        # structurally (no more stderr `Cursor:` scraping).
-        cursor=cursor,
-        run_stats=run_stats,
-        posts_processed=posts_processed,
     )
 
 
 def _fake_gdl_with_result(result):
     fake = MagicMock()
     fake.download = AsyncMock(return_value=result)
-    # Real values for the attrs run_download's native branch reads off the gdl
-    # (the native pacing config, plan #703) — a MagicMock would break the
-    # arithmetic (max(0.5, rate_limit/4)).
-    fake._rate_limit = 3.0
-    fake._validate_files = True
     fake._compute_run_stats = lambda *a, **k: {
         "exit_code": 0, "downloaded_count": len(result.written_paths),
         "skipped_count": 0, "per_item_failures": 0,
@@ -104,28 +94,6 @@ def _fake_gdl_with_result(result):
     return fake
 
 
-def _stub_patreon_ingester(svc, result, resolved_campaign_id=None):
-    """Route the phase-2 dispatch (download_backends.run_download via
-    DownloadService._run_download, plan #707 A5) to a canned DownloadResult so
-    these tests exercise download_service's phase-2-result→phase-3 handling
-    (status mapping, backfill lifecycle, health) without the real ingester's
-    network/DB. The ingester itself is covered by test_patreon_ingester.py; the
-    native construction/resolution by the run_download tests below. Returns a list
-    capturing (ctx, source_config, skip_value, mode) per call so a test can assert
-    mode / resume_cursor threading."""
-    calls = []
-
-    async def fake(*, ctx, source_config, skip_value, mode):
-        calls.append({
-            "ctx": ctx, "source_config": source_config,
-            "skip_value": skip_value, "mode": mode,
-        })
-        return result, resolved_campaign_id
-
-    svc._run_download = fake
-    return calls
-
-
 @pytest.mark.asyncio
 async def test_download_source_attaches_written_files(
     db, db_sync, tmp_path, seed_artist_and_source,
@@ -144,13 +112,12 @@ async def test_download_source_attaches_written_files(
     _make_jpg(f1, split="h")
     _make_jpg(f2, split="v")
 
-    result = _make_fake_dl_result(
+    fake_gdl = _fake_gdl_with_result(_make_fake_dl_result(
         success=True,
         written_paths=[str(f1), str(f2)],
         files_downloaded=2,
         stdout=f"{f1}\n{f2}\n",
-    )
-    fake_gdl = _fake_gdl_with_result(result)
+    ))
 
     sync_settings = db_sync.execute(
         select(ImportSettings).where(ImportSettings.id == 1)
@@ -170,7 +137,6 @@ async def test_download_source_attaches_written_files(
         async_session=db, sync_session=db_sync,
         gdl=fake_gdl, importer=importer, cred_service=cred_service,
     )
-    _stub_patreon_ingester(svc, result)
     event_id = await svc.download_source(source.id)
 
     ev = (await db.execute(
@@ -248,19 +214,40 @@ async def test_in_flight_idempotency_returns_existing_event(
 
 
 @pytest.mark.asyncio
-async def test_patreon_resolved_campaign_id_is_cached(
-    db, db_sync, tmp_path, seed_artist_and_source,
+async def test_patreon_retry_caches_campaign_id(
+    db, db_sync, tmp_path, seed_artist_and_source, monkeypatch,
 ):
-    """plan #697: the native ingester resolves a Patreon vanity → campaign id
-    on the fly (replacing gallery-dl's reactive campaign-id retry). When phase 2
-    reports a freshly-resolved id, phase 3 caches it on the source so later runs
-    skip the lookup. Stub the ingester to report a resolved id; assert it lands
-    in config_overrides."""
     from backend.app.services.download_service import DownloadService
     from backend.app.services.importer import Importer
 
     _artist, source = seed_artist_and_source
 
+    failed = _make_fake_dl_result(
+        success=False, written_paths=[], stdout="",
+        stderr="[patreon][error] Failed to extract campaign ID for alice\n",
+    )
+    succeeded = _make_fake_dl_result(success=True, written_paths=[])
+
+    download_calls = []
+
+    async def fake_download(url, **k):
+        download_calls.append(url)
+        return failed if len(download_calls) == 1 else succeeded
+
+    fake_gdl = MagicMock()
+    fake_gdl.download = fake_download
+    fake_gdl._compute_run_stats = lambda *a, **k: {
+        "exit_code": 0, "downloaded_count": 0, "skipped_count": 0,
+        "per_item_failures": 0, "warning_count": 0, "tier_gated_count": 0,
+    }
+    fake_gdl._extract_errors_warnings = lambda *a, **k: ""
+    fake_gdl._truncate_log = lambda x, **k: x
+
+    monkeypatch.setattr(
+        "backend.app.services.download_service.resolve_campaign_id",
+        AsyncMock(return_value="99"),
+    )
+
     sync_settings = db_sync.execute(
         select(ImportSettings).where(ImportSettings.id == 1)
     ).scalar_one()
@@ -274,97 +261,19 @@ async def test_patreon_resolved_campaign_id_is_cached(
 
     svc = DownloadService(
         async_session=db, sync_session=db_sync,
-        gdl=_fake_gdl_with_result(_make_fake_dl_result(success=True)),
-        importer=importer, cred_service=cred_service,
-    )
-    _stub_patreon_ingester(
-        svc, _make_fake_dl_result(success=True, written_paths=[]),
-        resolved_campaign_id="99",
+        gdl=fake_gdl, importer=importer, cred_service=cred_service,
     )
     await svc.download_source(source.id)
 
+    assert len(download_calls) == 2
+    assert "id:99" in download_calls[1]
+
     overrides = db_sync.execute(
         select(Source.config_overrides).where(Source.id == source.id)
     ).scalar_one()
     assert overrides.get("patreon_campaign_id") == "99"
 
 
-@pytest.mark.asyncio
-async def test_run_download_native_resolves_vanity_and_runs(
-    db, db_sync, tmp_path, seed_artist_and_source, monkeypatch,
-):
-    """run_download (native branch, plan #707 A5): with no cached campaign id,
-    resolve the vanity and pass the resolved id straight to the ingester; report
-    it back so phase 3 can cache it."""
-    from backend.app.services import download_backends as db_mod
-    from backend.app.services.gallery_dl import SourceConfig
-
-    _artist, source = seed_artist_and_source
-
-    monkeypatch.setattr(
-        db_mod, "resolve_campaign_id_for_source",
-        AsyncMock(return_value=("4242", "4242")),
-    )
-    run_kwargs = {}
-
-    class _FakeIngester:
-        def __init__(self, **kw):
-            pass
-
-        def run(self, **kw):
-            run_kwargs.update(kw)
-            return _make_fake_dl_result(success=True, written_paths=[])
-
-    monkeypatch.setattr(db_mod, "PatreonIngester", _FakeIngester)
-
-    ctx = {
-        "platform": "patreon",
-        "source_id": source.id, "url": "https://patreon.com/alice",
-        "artist_slug": "alice", "cookies_path": None,
-        "config_overrides": {},
-    }
-    result, resolved = await db_mod.run_download(
-        ctx=ctx, source_config=SourceConfig.from_dict({}), skip_value=True,
-        mode="tick",
-        gdl=_fake_gdl_with_result(_make_fake_dl_result(success=True)),
-        sync_session_factory=MagicMock(),
-    )
-    assert result.success is True
-    assert resolved == "4242"
-    assert run_kwargs["campaign_id"] == "4242"
-    assert run_kwargs["mode"] == "tick"
-
-
-@pytest.mark.asyncio
-async def test_run_download_native_unresolvable_fails_loud(
-    db, db_sync, tmp_path, seed_artist_and_source, monkeypatch,
-):
-    """A campaign id we can't resolve is a loud NOT_FOUND failure, never a
-    silent empty success."""
-    from backend.app.services import download_backends as db_mod
-    from backend.app.services.gallery_dl import ErrorType, SourceConfig
-
-    _artist, source = seed_artist_and_source
-
-    monkeypatch.setattr(
-        db_mod, "resolve_campaign_id_for_source",
-        AsyncMock(return_value=(None, None)),
-    )
-    ctx = {
-        "platform": "patreon",
-        "source_id": source.id, "url": "https://patreon.com/alice",
-        "artist_slug": "alice", "cookies_path": None,
-        "config_overrides": {},
-    }
-    result, resolved = await db_mod.run_download(
-        ctx=ctx, source_config=SourceConfig.from_dict({}), skip_value=True,
-        mode="tick", gdl=MagicMock(), sync_session_factory=MagicMock(),
-    )
-    assert result.success is False
-    assert result.error_type == ErrorType.NOT_FOUND
-    assert resolved is None
-
-
 # --- FC-3d: finalize hook updates Source health columns -------------------
 
 
@@ -437,40 +346,6 @@ async def test_finalize_error_increments_failures_and_sets_error(db):
     assert row.last_checked_at is not None
 
 
-@pytest.mark.asyncio
-async def test_rate_limited_cooldown_honors_retry_after(db, monkeypatch):
-    """plan #708 B1: a RATE_LIMITED result with a Retry-After stamps the platform
-    cooldown at that duration, clamped to [60, 3600]; no hint → the flat default."""
-    from unittest.mock import AsyncMock, MagicMock
-
-    from backend.app.services import download_service as dl_mod
-    from backend.app.services.download_service import DownloadService
-
-    source_id, _event_id = await _seed_source_with_health(db, suffix="rl")
-    cooldown = AsyncMock()
-    monkeypatch.setattr(dl_mod, "set_platform_cooldown", cooldown)
-    svc = DownloadService(
-        async_session=db, sync_session=None,
-        gdl=MagicMock(), importer=MagicMock(), cred_service=MagicMock(),
-    )
-
-    async def _health(retry):
-        cooldown.reset_mock()
-        await svc._update_source_health(
-            source_id=source_id, status="error", error_message="rate limited",
-            error_type="rate_limited", retry_after_seconds=retry,
-        )
-
-    await _health(120.0)
-    assert cooldown.await_args.kwargs["seconds"] == 120  # honored as-is
-    await _health(5.0)
-    assert cooldown.await_args.kwargs["seconds"] == 60  # floored
-    await _health(99999.0)
-    assert cooldown.await_args.kwargs["seconds"] == 3600  # capped
-    await _health(None)
-    assert "seconds" not in cooldown.await_args.kwargs  # default cooldown
-
-
 @pytest.mark.asyncio
 async def test_finalize_skipped_preserves_failures_clears_error(db):
     from backend.app.services.download_service import DownloadService
@@ -497,20 +372,32 @@ async def test_finalize_skipped_preserves_failures_clears_error(db):
     assert row.last_checked_at is not None
 
 
-# --- Plan #693: backfill state machine (time-boxed chunks, run-until-done) ---
+# --- Plan #544: backfill lifecycle + PARTIAL → status=ok -------------------
 
 
-def _backfill_svc(db, db_sync, tmp_path, result):
-    """DownloadService wired with a real importer (so empty written_paths just
-    attaches nothing) whose Patreon phase-2 branch is stubbed to return `result`.
-
-    The seeded source is Patreon, so phase 2 routes to the native ingester
-    (plan #697); the stub returns the canned result + captures the per-call
-    (ctx, source_config, mode). Returns (svc, ingester_calls)."""
+@pytest.mark.asyncio
+async def test_backfill_decrements_after_run(
+    db, db_sync, tmp_path, seed_artist_and_source,
+):
+    """When backfill_runs_remaining > 0 going in, a non-clean / non-empty
+    run decrements by 1 — operator gets N runs to complete the deep scan
+    before tick mode resumes."""
     from backend.app.services.download_service import DownloadService
     from backend.app.services.importer import Importer
 
+    _artist, source = seed_artist_and_source
+    source.backfill_runs_remaining = 3
+    await db.commit()
+
     images_root = tmp_path / "images"
+    f1 = images_root / "alice" / "patreon" / "post" / "a.jpg"
+    _make_jpg(f1, split="h")
+
+    fake_gdl = _fake_gdl_with_result(_make_fake_dl_result(
+        success=True, written_paths=[str(f1)], files_downloaded=1,
+        stdout=f"{f1}\n",
+    ))
+
     sync_settings = db_sync.execute(
         select(ImportSettings).where(ImportSettings.id == 1)
     ).scalar_one()
@@ -519,280 +406,97 @@ def _backfill_svc(db, db_sync, tmp_path, result):
         thumbnailer=Thumbnailer(images_root=images_root), settings=sync_settings,
     )
     cred_service = CredentialService(db, CredentialCrypto(tmp_path / "key.b64", bootstrap_ok=True))
-    fake_gdl = _fake_gdl_with_result(result)
     svc = DownloadService(
         async_session=db, sync_session=db_sync,
         gdl=fake_gdl, importer=importer, cred_service=cred_service,
     )
-    calls = _stub_patreon_ingester(svc, result)
-    return svc, calls
-
-
-@pytest.mark.asyncio
-async def test_backfill_chunk_progress_advances_cursor(
-    db, db_sync, tmp_path, seed_artist_and_source,
-):
-    """A running backfill chunk that didn't finish but advanced (new cursor)
-    stays 'running', checkpoints the cursor, bumps the chunk counter, and
-    spends one safety-cap chunk."""
-    _artist, source = seed_artist_and_source
-    source.config_overrides = {"_backfill_state": "running"}
-    source.backfill_runs_remaining = 5
-    await db.commit()
-
-    svc, _ = _backfill_svc(db, db_sync, tmp_path, _make_fake_dl_result(
-        success=False, written_paths=[], files_downloaded=0,
-        cursor="03:PAGE2:xyz",
-    ))
     await svc.download_source(source.id)
 
-    remaining, co = (await db.execute(
-        select(Source.backfill_runs_remaining, Source.config_overrides)
-        .where(Source.id == source.id)
-    )).one()
-    assert co.get("_backfill_state") == "running"
-    assert co.get("_backfill_cursor") == "03:PAGE2:xyz"
-    assert co.get("_backfill_chunks") == 1
-    assert remaining == 4
-
-
-@pytest.mark.asyncio
-async def test_backfill_state_running_selects_backfill_mode_and_resumes(
-    db, db_sync, tmp_path, seed_artist_and_source,
-):
-    """state=='running' drives backfill mode (chunk budget) and threads the
-    stored cursor to the native ingester as the resume point."""
-    from backend.app.services.gallery_dl import BACKFILL_CHUNK_SECONDS
-
-    _artist, source = seed_artist_and_source
-    source.config_overrides = {"_backfill_state": "running", "_backfill_cursor": "03:RESUME:here"}
-    source.backfill_runs_remaining = 5
-    await db.commit()
-
-    svc, calls = _backfill_svc(db, db_sync, tmp_path, _make_fake_dl_result(
-        success=False, written_paths=[],
-        cursor="03:RESUME2:next",
-    ))
-    await svc.download_source(source.id)
-
-    assert len(calls) == 1
-    assert calls[0]["mode"] == "backfill"
-    assert calls[0]["source_config"].resume_cursor == "03:RESUME:here"
-    assert calls[0]["source_config"].timeout == BACKFILL_CHUNK_SECONDS
-
-    co = (await db.execute(
-        select(Source.config_overrides).where(Source.id == source.id)
+    remaining = (await db.execute(
+        select(Source.backfill_runs_remaining).where(Source.id == source.id)
     )).scalar_one()
-    assert co.get("_backfill_cursor") == "03:RESUME2:next"
+    assert remaining == 2
 
 
 @pytest.mark.asyncio
-async def test_backfill_clean_exit_marks_complete(
+async def test_backfill_auto_resets_on_clean_zero_files(
     db, db_sync, tmp_path, seed_artist_and_source,
 ):
-    """A clean rc=0 chunk = gallery-dl reached the bottom → state 'complete',
-    cursor cleared, returns to tick mode."""
+    """A clean run (rc=0) that downloaded zero files means the backfill
+    queue drained — reset to 0 immediately instead of wasting the rest of
+    the N-run budget on no-op walks."""
+    from backend.app.services.download_service import DownloadService
+    from backend.app.services.importer import Importer
+
     _artist, source = seed_artist_and_source
-    source.config_overrides = {"_backfill_state": "running", "_backfill_cursor": "03:NEAR:bottom"}
-    source.backfill_runs_remaining = 5
+    source.backfill_runs_remaining = 3
     await db.commit()
 
-    svc, _ = _backfill_svc(db, db_sync, tmp_path, _make_fake_dl_result(
+    fake_result = _make_fake_dl_result(
         success=True, written_paths=[], files_downloaded=0,
-    ))
+    )
+    fake_gdl = _fake_gdl_with_result(fake_result)
+
+    sync_settings = db_sync.execute(
+        select(ImportSettings).where(ImportSettings.id == 1)
+    ).scalar_one()
+    importer = Importer(
+        session=db_sync, images_root=tmp_path / "images",
+        import_root=tmp_path / "images",
+        thumbnailer=Thumbnailer(images_root=tmp_path / "images"),
+        settings=sync_settings,
+    )
+    cred_service = CredentialService(db, CredentialCrypto(tmp_path / "key.b64", bootstrap_ok=True))
+    svc = DownloadService(
+        async_session=db, sync_session=db_sync,
+        gdl=fake_gdl, importer=importer, cred_service=cred_service,
+    )
     await svc.download_source(source.id)
 
-    remaining, co = (await db.execute(
-        select(Source.backfill_runs_remaining, Source.config_overrides)
-        .where(Source.id == source.id)
-    )).one()
-    assert co.get("_backfill_state") == "complete"
-    assert "_backfill_cursor" not in co
+    remaining = (await db.execute(
+        select(Source.backfill_runs_remaining).where(Source.id == source.id)
+    )).scalar_one()
     assert remaining == 0
 
 
 @pytest.mark.asyncio
-async def test_backfill_cap_exhaustion_stalls(
+async def test_tick_mode_does_not_touch_backfill_counter(
     db, db_sync, tmp_path, seed_artist_and_source,
 ):
-    """A progressing chunk that spends the last safety-cap chunk without
-    reaching the bottom pauses as 'stalled' (doesn't loop forever)."""
-    _artist, source = seed_artist_and_source
-    source.config_overrides = {"_backfill_state": "running"}
-    source.backfill_runs_remaining = 1
-    await db.commit()
+    """When backfill_runs_remaining is already 0, downloads don't go
+    negative or otherwise mutate the counter."""
+    from backend.app.services.download_service import DownloadService
+    from backend.app.services.importer import Importer
 
-    svc, _ = _backfill_svc(db, db_sync, tmp_path, _make_fake_dl_result(
-        success=False, written_paths=[],
-        cursor="03:MORE:left",
+    _artist, source = seed_artist_and_source
+    assert source.backfill_runs_remaining == 0
+
+    fake_gdl = _fake_gdl_with_result(_make_fake_dl_result(
+        success=True, written_paths=[], files_downloaded=0,
     ))
+
+    sync_settings = db_sync.execute(
+        select(ImportSettings).where(ImportSettings.id == 1)
+    ).scalar_one()
+    importer = Importer(
+        session=db_sync, images_root=tmp_path / "images",
+        import_root=tmp_path / "images",
+        thumbnailer=Thumbnailer(images_root=tmp_path / "images"),
+        settings=sync_settings,
+    )
+    cred_service = CredentialService(db, CredentialCrypto(tmp_path / "key.b64", bootstrap_ok=True))
+    svc = DownloadService(
+        async_session=db, sync_session=db_sync,
+        gdl=fake_gdl, importer=importer, cred_service=cred_service,
+    )
     await svc.download_source(source.id)
 
-    remaining, co = (await db.execute(
-        select(Source.backfill_runs_remaining, Source.config_overrides)
-        .where(Source.id == source.id)
-    )).one()
-    assert co.get("_backfill_state") == "stalled"
+    remaining = (await db.execute(
+        select(Source.backfill_runs_remaining).where(Source.id == source.id)
+    )).scalar_one()
     assert remaining == 0
 
 
-@pytest.mark.asyncio
-async def test_backfill_stuck_guard_stalls_after_two_no_advance(
-    db, db_sync, tmp_path, seed_artist_and_source,
-):
-    """Two consecutive chunks that fail to advance the cursor → 'stalled',
-    cursor cleared, so a wedged walk can't re-strand forever."""
-    from unittest.mock import MagicMock
-
-    from backend.app.services.download_service import DownloadService
-
-    _artist, source = seed_artist_and_source
-    source.config_overrides = {"_backfill_state": "running", "_backfill_cursor": "stuck"}
-    source.backfill_runs_remaining = 5
-    await db.commit()
-
-    svc = DownloadService(
-        async_session=db, sync_session=db_sync,
-        gdl=MagicMock(), importer=MagicMock(), cred_service=MagicMock(),
-    )
-    ctx = {
-        "source_id": source.id, "platform": "patreon",
-        "config_overrides": {"_backfill_state": "running", "_backfill_cursor": "stuck"},
-        "backfill_runs_remaining": 5,
-    }
-    stuck = _make_fake_dl_result(success=False, cursor="stuck")
-
-    await svc._apply_backfill_lifecycle(ctx, stuck)
-    await db.commit()
-    co = (await db.execute(
-        select(Source.config_overrides).where(Source.id == source.id)
-    )).scalar_one()
-    assert co.get("_backfill_state") == "running"
-    assert co.get("_backfill_cursor_stalls") == 1
-
-    ctx["config_overrides"] = dict(co)
-    await svc._apply_backfill_lifecycle(ctx, stuck)
-    await db.commit()
-    co2 = (await db.execute(
-        select(Source.config_overrides).where(Source.id == source.id)
-    )).scalar_one()
-    assert co2.get("_backfill_state") == "stalled"
-    assert "_backfill_cursor" not in co2
-
-
-@pytest.mark.asyncio
-async def test_backfill_lifecycle_leaves_posts_to_ingester(
-    db, db_sync, tmp_path, seed_artist_and_source,
-):
-    """plan #704 (#5), revised: _backfill_posts is OWNED by the ingester now (it
-    writes a monotonic absolute live mid-walk), so the post-chunk lifecycle must
-    NOT touch it — it carries the ingester's committed value forward unchanged."""
-    from unittest.mock import MagicMock
-
-    from backend.app.services.download_service import DownloadService
-
-    _artist, source = seed_artist_and_source
-    source.config_overrides = {"_backfill_state": "running", "_backfill_posts": 12}
-    source.backfill_runs_remaining = 5
-    await db.commit()
-
-    svc = DownloadService(
-        async_session=db, sync_session=db_sync,
-        gdl=MagicMock(), importer=MagicMock(), cred_service=MagicMock(),
-    )
-    ctx = {
-        "source_id": source.id, "platform": "patreon",
-        "config_overrides": {"_backfill_state": "running", "_backfill_posts": 12},
-        "backfill_runs_remaining": 5,
-    }
-    await svc._apply_backfill_lifecycle(
-        ctx, _make_fake_dl_result(success=False, cursor="C1", posts_processed=10),
-    )
-    await db.commit()
-    co = (await db.execute(
-        select(Source.config_overrides).where(Source.id == source.id)
-    )).scalar_one()
-    assert co.get("_backfill_posts") == 12  # untouched — the ingester owns it
-
-
-@pytest.mark.asyncio
-async def test_backfill_timeout_chunk_reclassified_to_ok(
-    db, db_sync, tmp_path, seed_artist_and_source,
-):
-    """A backfill chunk that hit its time-box (TIMEOUT) but advanced is
-    reclassified to PARTIAL → event status 'ok' (progress, not error), and
-    consecutive_failures is not bumped."""
-    from backend.app.services.gallery_dl import ErrorType
-
-    _artist, source = seed_artist_and_source
-    source.config_overrides = {"_backfill_state": "running"}
-    source.backfill_runs_remaining = 5
-    await db.commit()
-
-    svc, _ = _backfill_svc(db, db_sync, tmp_path, _make_fake_dl_result(
-        success=False, files_downloaded=3,
-        error_type=ErrorType.TIMEOUT,
-        error_message="Download timed out",
-        cursor="03:NEXT:page",
-    ))
-    event_id = await svc.download_source(source.id)
-
-    ev = (await db.execute(
-        select(DownloadEvent).where(DownloadEvent.id == event_id)
-    )).scalar_one()
-    assert ev.status == "ok"
-    failures = (await db.execute(
-        select(Source.consecutive_failures).where(Source.id == source.id)
-    )).scalar_one()
-    assert failures == 0
-
-
-@pytest.mark.asyncio
-async def test_tick_mode_when_not_running_leaves_state_untouched(
-    db, db_sync, tmp_path, seed_artist_and_source,
-):
-    """No _backfill_state → not backfilling; the lifecycle is a no-op and the
-    counter/state aren't mutated."""
-    _artist, source = seed_artist_and_source
-    assert (source.config_overrides or {}).get("_backfill_state") is None
-
-    svc, _ = _backfill_svc(db, db_sync, tmp_path, _make_fake_dl_result(
-        success=True, written_paths=[], files_downloaded=0,
-    ))
-    await svc.download_source(source.id)
-
-    co = (await db.execute(
-        select(Source.config_overrides).where(Source.id == source.id)
-    )).scalar_one()
-    assert (co or {}).get("_backfill_state") is None
-
-
-@pytest.mark.asyncio
-async def test_recovery_mode_selected_and_flag_cleared_on_complete(
-    db, db_sync, tmp_path, seed_artist_and_source,
-):
-    """plan #697: `_backfill_bypass_seen` alongside a running backfill selects
-    recovery mode (ingester bypasses the seen-ledger). A clean rc=0 walk
-    completes the shared lifecycle AND clears the bypass flag so the next tick
-    honors the ledger again."""
-    _artist, source = seed_artist_and_source
-    source.config_overrides = {"_backfill_state": "running", "_backfill_bypass_seen": True}
-    source.backfill_runs_remaining = 5
-    await db.commit()
-
-    svc, calls = _backfill_svc(db, db_sync, tmp_path, _make_fake_dl_result(
-        success=True, written_paths=[], files_downloaded=0,
-    ))
-    await svc.download_source(source.id)
-
-    assert calls[0]["mode"] == "recovery"
-    co = (await db.execute(
-        select(Source.config_overrides).where(Source.id == source.id)
-    )).scalar_one()
-    assert co.get("_backfill_state") == "complete"
-    assert "_backfill_bypass_seen" not in co
-
-
 @pytest.mark.asyncio
 async def test_partial_error_type_maps_to_ok_status(
     db, db_sync, tmp_path, seed_artist_and_source,
@@ -830,7 +534,6 @@ async def test_partial_error_type_maps_to_ok_status(
         async_session=db, sync_session=db_sync,
         gdl=fake_gdl, importer=importer, cred_service=cred_service,
     )
-    _stub_patreon_ingester(svc, fake_result)
     event_id = await svc.download_source(source.id)
 
     ev = (await db.execute(
@@ -866,11 +569,10 @@ async def test_download_enqueues_thumbnail_and_ml_per_attached_image(
     _make_jpg(f1, split="h")
     _make_jpg(f2, split="v")
 
-    result = _make_fake_dl_result(
+    fake_gdl = _fake_gdl_with_result(_make_fake_dl_result(
         success=True, written_paths=[str(f1), str(f2)],
         files_downloaded=2, stdout=f"{f1}\n{f2}\n",
-    )
-    fake_gdl = _fake_gdl_with_result(result)
+    ))
 
     sync_settings = db_sync.execute(
         select(ImportSettings).where(ImportSettings.id == 1)
@@ -903,7 +605,6 @@ async def test_download_enqueues_thumbnail_and_ml_per_attached_image(
         async_session=db, sync_session=db_sync,
         gdl=fake_gdl, importer=importer, cred_service=cred_service,
     )
-    _stub_patreon_ingester(svc, result)
     await svc.download_source(source.id)
 
     # Two files attached → two thumbnail enqueues + two ML enqueues, IDs
@@ -918,11 +619,11 @@ async def test_download_enqueues_thumbnail_and_ml_per_attached_image(
 async def test_releases_db_connections_before_subprocess(
     db, db_sync, tmp_path, seed_artist_and_source, monkeypatch,
 ):
-    """Phase 1's DB connections must be released BEFORE the (multi-minute)
-    phase-2 fetch, so they don't idle-die and strand phase 3 with
-    ConnectionDoesNotExistError (Anduo #40014). Spy on the async session close
-    and assert it happened before the phase-2 work (here the native Patreon
-    ingester) runs; phase 3 must still finalize the event."""
+    """Phase 1's DB connections must be released BEFORE the (up to ~19.5-min
+    in backfill) gallery-dl subprocess, so they don't idle-die and strand
+    phase 3 with ConnectionDoesNotExistError (Anduo #40014). Spy on the
+    async session close and assert it happened before gdl.download runs;
+    phase 3 must still finalize the event."""
     from backend.app.services.download_service import DownloadService
     from backend.app.services.importer import Importer
 
@@ -938,9 +639,19 @@ async def test_releases_db_connections_before_subprocess(
     monkeypatch.setattr(db, "close", spy_close)
 
     seen = {}
-    fake_gdl = _fake_gdl_with_result(
-        _make_fake_dl_result(success=True, written_paths=[], stdout="")
-    )
+
+    async def fake_download(url, **k):
+        seen["closed_before_download"] = closed["async"]
+        return _make_fake_dl_result(success=True, written_paths=[], stdout="")
+
+    fake_gdl = MagicMock()
+    fake_gdl.download = fake_download
+    fake_gdl._compute_run_stats = lambda *a, **k: {
+        "exit_code": 0, "downloaded_count": 0, "skipped_count": 0,
+        "per_item_failures": 0, "warning_count": 0, "tier_gated_count": 0,
+    }
+    fake_gdl._extract_errors_warnings = lambda *a, **k: ""
+    fake_gdl._truncate_log = lambda x, **k: x
 
     sync_settings = db_sync.execute(
         select(ImportSettings).where(ImportSettings.id == 1)
@@ -957,15 +668,9 @@ async def test_releases_db_connections_before_subprocess(
         async_session=db, sync_session=db_sync,
         gdl=fake_gdl, importer=importer, cred_service=cred_service,
     )
-
-    async def fake_ingest(*, ctx, source_config, skip_value, mode):
-        seen["closed_before_phase2"] = closed["async"]
-        return _make_fake_dl_result(success=True, written_paths=[], stdout=""), None
-
-    svc._run_download = fake_ingest
     event_id = await svc.download_source(source.id)
 
-    assert seen["closed_before_phase2"] is True
+    assert seen["closed_before_download"] is True
     ev = (await db.execute(
         select(DownloadEvent).where(DownloadEvent.id == event_id)
     )).scalar_one()