Five extension-miss findings from the 2026-06-02 audit, where a status
value was added on one side but a downstream consumer didn't pick it up.
- download_service._phase3_persist: explicit branches for
ImportResult.status in ('failed','refreshed'). For 'failed' (archive
probe crash from _import_archive), unlink the source file so the
filesystem scanner doesn't re-import and re-crash on the same
archive forever. 'refreshed' is currently unreachable from the
download path (no deep=True) but matches the importer's documented
contract; treat as 'attached'.
- gallery-dl backfill auto-complete now gates on dl_result.success +
no error_type, not just return_code==0 + files_downloaded==0.
VALIDATION_FAILED exits the subprocess with returncode=0 and
files_downloaded=0 when every file was quarantined, matching the
prior predicate exactly and zeroing the operator's armed backfill
budget on the FIRST quarantine run instead of decrementing.
- attach_in_place archive dispatch now threads artist + source_row
through _import_archive (and _import_media for archive members)
and _supersede. The path-walk fallback (_resolve_artist) is still
used by filesystem-import; the download path now binds
ImageProvenance to the explicit subscription Source instead of
rediscovering by (artist_id, platform).
- Three FE handlers now recognize status:'deferred' from
/api/sources/<id>/check: SubscriptionsTab.onCheck (was toasting
"event #undefined"), SubscriptionsTab.checkAll (was counting
deferred as queued), DownloadEventRow.onRetry (was saying
"re-queued" when nothing was). Pattern matches DownloadsTab.onRetryAll
which already had it.
- celery_signals._queue_for now maps backup/admin/library_audit
prefixes to 'maintenance' (matching task_routes). TaskRun.queue
was returning 'default' for those rows, so per-queue dashboard
filters and per-queue threshold overrides (added in G3) silently
missed them.
Plugs the FC long-running-entity discipline gaps the 2026-06-02 audit
flagged: every entity that can get stuck now has recovery + retention +
timeout, and the long-runners no longer collide with the FC-3i sweep.
Recovery sweeps (every 5 min):
- recover_stalled_backup_runs — flips BackupRun stuck in
running/restoring past 7h (covers the 6.5h images-backup hard
limit) to error. prune_backups docstring corrected — the FC-3i
TaskRun sweep never touched BackupRun rows.
- recover_stalled_library_audit_runs — flips LibraryAuditRun stuck
past 135 min (10-min buffer above scan_library_for_rule's 2h5m
hard limit) to error. Previously a SIGKILL'd row blocked all
future audits until manual DB surgery.
- recover_stalled_import_batches — finalizes ImportBatch rows
stuck running >2h whose child tasks are all terminal (orphan case
where the orchestrator crashed before the closing UPDATE). Uses
the same EXISTS predicate /api/system/stats already had.
Retention (daily):
- prune_library_audit_runs — 30-day window. Audit rows carry
matched_ids JSONB blobs that can hold tens of thousands of ids.
- prune_import_batches — 30-day window. Cascades to ImportTask via
the model relationship.
time_limits on five long-runners that previously had none (the
audit's headline finding — every one of these collided with the
recover_stalled_task_runs 5-min default and could be marked
'error' mid-flight):
- scan_directory: 60m soft / 70m hard
- verify_integrity: 60m / 70m
- backfill_phash: 30m / 35m
- apply_allowlist_tags: 30m / 35m
- recompute_centroids: 30m / 35m
QUEUE_STUCK_THRESHOLD_MINUTES now covers maintenance (75) and scan
(75) — above the longest task on each — with per-task overrides
for the outliers (backup_images_task 420, restore_images_task 420,
scan_library_for_rule 130).
start_audit_run guard is now age-aware: a 'running' row older than
the audit hard limit doesn't block a new run (the sweep will catch
it within 5 min). Previously a SIGKILL'd row blocked forever.
/api/import/status now uses the same EXISTS predicate
/api/system/stats does, so the two endpoints no longer disagree on
the active-batch question.
DownloadEvent.started_at resets on pending→running so a freshly-
promoted event from a busy queue isn't measured against its
original enqueue time (was racing recover_stalled_download_events
on heavy-queue days).
Extracts gallery.js's hand-rolled inflightId pattern into a new
useInflightToken composable; adopts in every store that previously
had no guard against late-response overwrites or wrong-image URL
interpolation.
Two operator-impacting bugs the audit (workflow wf_bbe3fdb1-e62)
flagged:
- modal.removeTag rolled back the chip rail unconditionally even
when only the secondary dismiss POST had failed — UI lied until
refresh. And all tag-mutation URLs interpolated currentImageId
AFTER an await, so a fast prev/next could route DELETE/POST to
the wrong image. Both fixed: split try/catch (dismiss failure
surfaces a warning, doesn't roll back the delete); imageId
captured at call-time and used in URLs throughout.
- suggestions.accept dereferenced currentImageId after the awaited
POST /api/tags, so the subsequent /suggestions/accept could
apply A's chosen tag to image B AND push it to B's allowlist.
Fixed by capturing imageId at click-time + inflight guard on
load().
Same shape across artist / downloads / artistDirectory /
tagDirectory / posts stores: rapid filter/nav changes used to
interleave responses (last-writer-wins). Now the late response is
discarded and the most-recent request wins. Filter-change-during-
search no longer drops the second fetch because the loading flag
was still true from the first.
gallery.js's inflightId removed in favor of the shared composable
so the pattern stays consistent.
Stale assertion pinned the buggy pre-G1.6 behavior (the test name
"upload invalidates the cache" literally describes the bug). The
audit's correction makes upload mirror the returned record into the
store so the card updates immediately — update the assertion to
match the corrected behavior.
- BACKFILL_TIMEOUT_SECONDS 1800→1170: keep the subprocess timeout
30s below Celery's hard time_limit=1200 so SIGKILL doesn't beat
TimeoutExpired (matched the tick 870s/900s rationale). Backfill
runs that hit the cap let the next tick continue via the archive.
- recover_interrupted_tasks orphan UPDATE now stamps finished_at;
without it cleanup_old_tasks' WHERE finished_at<cutoff never
reaped orphan-swept rows. recover_stalled_task_runs also now sets
duration_ms (matches celery_signals.finalize's millisecond math).
- ExtensionService.quick_add_source arms NEW_SOURCE_BACKFILL_RUNS=3
on Source creation, mirroring SourceService.create. Without it,
Firefox quick-add on a creator with >20 unsynced posts walked the
full feed until subprocess timeout. Renamed the constant from
_NEW_SOURCE_BACKFILL_RUNS so it can be imported cross-module.
- gallery_dl.verify() accepts TIER_LIMITED as auth-success alongside
NO_NEW_CONTENT — the download path (line 712) already does, and
TIER_LIMITED proves auth reached the post and was told it was
tier-gated. Verify endpoint previously showed red on this and
prompted operators to rotate working cookies.
- prune_unused_tags now runs a single DELETE with the NOT-IN
predicate find_unused_tags uses, instead of SELECT-ids →
DELETE-WHERE-IN. Removes the psycopg 65535-param cliff that
would have surfaced on a tag explosion (>65k unused tags).
- credentials.upload() reflects the returned record into the store
cache (`.set(platform, rec)`) instead of evicting it; previously
the card briefly rendered "no credential" between upload and
loadAll().
attach_in_place mirrored only the media flow, so gallery-dl-downloaded
zips/PDFs/audio bounced back as `skipped+invalid_image`, which
download_service counted as an ingest error and flipped runs to
status="error" despite N successful image attaches. Lustria patreon
event #38998 (21 images + 1 OST zip) went red for exactly this reason.
Now attach_in_place dispatches the same way as import_one: archives →
_import_archive (extracts media members, captures archive as
PostAttachment), non-media → _capture_attachment. Download_service
accepts the new `attached` result and treats non-duplicate skips as
soft skips, not ingest errors.
Also: ShowcaseView always loadInitial() on mount, not just when the
store is empty — Pinia persists across navigations and operator wants
a fresh shuffle every time the showcase loads.
/api/showcase returns a random sample; after enough scrolling, an
unlucky 3-item batch can be entirely in the `seen` Set, which was
flipping `exhausted=true` and surfacing "End." mid-scroll. The
showcase is endless by intent — only a genuinely empty API response
(library has 0 images) should mark it exhausted. Tiny-library
fallback: cap retries at 8 to avoid spinning forever.
Operator-flagged 2026-06-01: clicking Accept on a suggestion added
the tag to the database but the modal's chip rail kept showing the
old set. Suggestion would disappear from the suggestions list
(suggestionsStore drops it from byCategory) and the toast would
confirm "Tagged: …", but visually the modal looked unchanged
until you closed and reopened it.
Root cause: useSuggestionsStore.accept() POSTs to the backend and
updates its own state, but never tells useModalStore that the
backing image's tag set has changed. The addExistingTag and
createAndAdd flows in TagPanel already call modal.reloadTags()
after applying — the suggestions path needed the same refresh.
Two-line fix in SuggestionsPanel: after store.accept() and
store.aliasAccept(), call modal.reloadTags() so TagPanel's
`modal.current?.tags` rebinds.
Same 'change shared shape, miss pinned tests' trap as cfa4fb4 — the
prior commit updated only test_backfill_mixed_aggregate; three sibling
tests in the same file also pinned the old {enqueued,ok,regenerated}
shape without 'scanned' and CI bounced.
Two coupled problems, operator-flagged 2026-06-01: "missing thumbnails
but triggering backfill found nothing."
1. **Backfill UI was a black box.** `POST /api/thumbnails/backfill`
returned just `{celery_task_id}` and the admin card said
"Enqueued." with no counts. There was no way to tell whether the
scan found 0 candidates, 5000 candidates, or whether the worker
even picked up the task. "Found nothing" was indistinguishable
from a broken queue.
Fix: refactor the scan into a sync helper (`_run_backfill_scan`)
shared by the Celery task and the API endpoint. The API now runs
the scan in an executor and returns `{scanned, enqueued, ok,
regenerated}`. The actual thumbnail generation work still goes
to the thumbnail Celery queue per row via
`generate_thumbnail.delay()` — the scan itself is fast
(SELECT id+thumbnail_path + a file.stat() per row).
2. **`_thumb_is_valid` accepted header-only corrupt files.** The
magic-byte check passed for any 8-byte file starting with a JPEG
or PNG header, including empty/truncated/zero-pad files that
browsers render as broken. Backfill counted these as `ok` and
never regenerated.
Fix: also require file size ≥ MIN_THUMB_BYTES (256). Real
thumbnails are minimum ~2KB even on solid-color sources; header-
only corrupt files top out around 12 bytes. 256 is well above
the corrupt floor and well below any legitimate thumbnail.
Plus the admin card now shows the per-run counts instead of
"Enqueued.":
Scanned 5,432 · enqueued 3 (2 regenerated) · 5,429 ok
Operator-flagged 2026-06-01: downloaded images stayed at
thumbnail_path=NULL until a periodic backfill sweep picked them up,
surfacing as broken-thumbnail tiles in the gallery for hours after
the download landed.
Importer.attach_in_place deliberately skips inline thumbnail
generation (importer.py:591-592) so the import queue stays moving —
the CALLING task is responsible for the enqueue. tasks/import_file.py
already did this (line 228-239). tasks/download.py / download_service
did not — every gallery-dl-attached image landed un-thumbnailed.
Fix in download_service._phase3_persist: after each
`attach_in_place` returning status in (imported, superseded), fan out
`generate_thumbnail.delay()` + `tag_and_embed.delay()` for each
image_id. Lazy import avoids circular-import risk between
download_service and the celery task modules that depend on it.
Mirrors the existing pattern verbatim — single source of truth for
"what fires after a successful attach" remains a comment in two
places (filesystem-import task, download orchestrator) rather than
a shared helper, because the contexts differ enough (sync session vs
async orchestrator) that abstracting would obscure more than it'd
share.
Test covers the happy-path with two attached files: both get the
thumbnail enqueue AND the ML enqueue, with image_ids drawn from
ImportResult (so future supersede-on-attach paths stay covered).
The zebra striping in 717b601 was too subtle — inside the tonal error
card, surface-tint contrast washed out and rows still looked uniform.
Operator-flagged the same issue twice.
Replace the surface-tint approach with hard visual structure:
- Bottom border on each row (1px white/0.08) — clearly delineated rows
regardless of card background
- Hover darkens to black/0.25 — much more obvious than the previous
error-tint hover, gives the eye a clear horizontal track
- Slightly more vertical padding (8px vs 6px) for tap-target comfort
- Drop the gap; borders are the separator now
Operator-flagged 2026-06-01: with the failing-sources panel expanded,
all rows have the same flat low-opacity background, no visual track
from the artist name on the left to the Logs/Retry buttons on the
right. Hard to tell which row a button belongs to when scanning down
a list of 17.
Three pure-CSS changes (no DOM, no logic):
- 3px gap between rows (was 2px) — slightly more breathing room
- Even rows get a darker surface tint (zebra striping)
- Hover paints the whole row in a low-opacity error tint, so moving
the cursor toward Logs/Retry lights up the whole horizontal band
and the eye traces back to the artist name automatically
`test_set_backfill_runs_arms_source` was pinning the pre-auto-arm initial
value (0) when checking that the override works. Now that create() pre-arms
enabled sources to 3, the assertion is stale — the test was already
verifying the override path, the pre-assertion just decorated it.
Drop the pre-assertion; keep the override check. Other tests use raw
Source(...) constructors (default to 0 via the column server_default),
not SourceService.create(), so they're unaffected.
A freshly created subscription has no gallery-dl archive yet, so the
first poll in tick mode would walk forever — exit:20 doesn't trip
until 20 contiguous archive-hits, and there are none. The wall-clock
cap kicks in mid-walk, and the partial-success classifier (plan #544)
gracefully labels it status=ok, but the operator still wonders why
their new subscription isn't grabbing the back-catalog.
Pre-arm `backfill_runs_remaining = 3` on create() when `enabled=True`.
Same default as the manual "Deep scan" button, same auto-decrement
and auto-reset rules — once the queue drains (clean exit + zero new
files) or the budget runs out, tick mode resumes naturally.
Disabled sources (sidecar synthetics with `url='sidecar:<platform>:<slug>'`
that arrive disabled = False, or operator-added sources that start
disabled deliberately) skip the pre-arm — they're never polled, no
budget to waste.
Operator-flagged 2026-06-01 (DaferQ patreon, event #38919). Video posts
hosted on Mux carry a JWT that encodes a playback restriction policy
(`playback_restriction_id` in the token). The token signature alone is
not sufficient — Mux's CDN ALSO checks Referer/Origin on every fetch.
gallery-dl's HEAD probe to `stream.mux.com/<id>.m3u8?token=...` returns
200 (token is valid, HEAD sends no Referer that Mux's policy rejects),
but when yt-dlp follows up with the actual manifest GET it sends its own
default Referer and Mux 403s. yt-dlp retries 4 times, gives up, the
single video fails — every other image in the same post still downloads.
Static `downloader.ytdl.raw-options.http_headers` block now pins Referer
and Origin to `https://www.patreon.com` so yt-dlp's manifest fetch
clears the policy check.
Headers-only restrictions are now handled. Mux IP-range restrictions
(if a creator opts into them) remain unfixable from our worker — those
would need a Patreon-region IP. Per-video PARTIAL classifier (shipped
in plan #544 last commit) already handles the residual case: a run that
grabs all images and fails 1 video classifies as status=ok rather than
flipping the whole event red.
Routine subscription polls walked the entire post history every tick
even when nothing had changed, because gallery-dl's default `skip: True`
continues iterating archived posts. A creator with ~550 archived posts
(Knuxy patreon) saturates the 870s wall-clock cap before completing,
even with zero downloads needed. Plus, a tier-limited run that
downloaded hundreds of files but ran out the clock should be a
warning, not an error.
Two coupled changes, both operator-flagged 2026-06-01:
* **Tick mode (default, cron polls).** New `TICK_SKIP_VALUE = "exit:20"`
asks gallery-dl to exit after 20 contiguous archived items. Fresh
subscriptions + new-content cases still walk normally; established
subscription with zero new content exits in ~30s of HEAD requests
instead of pegging the timeout. 20 (not 5) gives headroom against
paywall warnings interleaving with archived items.
* **Backfill mode (explicit, operator-triggered).** Sticky for N runs
via new `Source.backfill_runs_remaining` (alembic 0031). While > 0,
downloads use `skip: True` + 1800s timeout. Auto-decrements per run
with early-reset to 0 when a clean run finds zero files (queue
drained). N defaults to 3 — multiple runs give the system enough
budget to finish a deep walk across timeout boundaries. New
`POST /api/sources/{id}/backfill` arms the source; "Deep scan"
button on each SourceRow (chip shows remaining count) wires it.
Plus partial-success classifier: non-zero gallery-dl exit + ≥1 file
downloaded + no source-level error fires `ErrorType.PARTIAL`, which
download_service maps to `status=\"ok\"`. The run did real work; the
next tick continues via gallery-dl's archive. No more red events for
"timed out mid-walk after downloading 300 files."
Retires `SourceConfig.skip_existing` — skip value is now derived from
the source state and passed as a separate `skip_value` parameter
through download() / _build_config_for_source(). `GD_DEFAULTS` drops
the now-dead key (was inert data after this refactor).
Tests cover:
* tick + backfill skip-value emission in _build_config_for_source
* PARTIAL classifier branch + TIER_LIMITED-wins-over-PARTIAL ordering
* SourceService.set_backfill_runs validation + persistence
* /api/sources/{id}/backfill 200/400/404 paths
* download_service auto-decrement / auto-reset / tick-mode-no-touch
* PARTIAL → status=ok in the orchestrator (no consecutive_failures bump)
gallery_service._provenance_clause artist branch was correlating its bare
Post reference to the outer query's primary_post_id outer-join, so the
artist filter silently matched zero rows for images with no primary post.
Alias Post inside the EXISTS subquery so SQLAlchemy adds it to the inner
FROM rather than treating it as a correlated outer table.
Five sidecar/import tests still asserted that a synthetic Source row
appears after a filesystem import. Alembic 0030 retired that behavior;
the Post sits null-source and the artist linkage lives on Post.artist_id.
Updated test_sidecar_creates_provenance, test_reimport_same_post_idempotent,
test_sidecar_artist_used_when_no_folder_artist, test_supersede_applies_new_file_sidecar,
and test_apply_sidecar_recovers_from_integrity_error to assert
post.source_id IS NULL + post.artist_id linkage instead.
Operator-asked 2026-06-01 after the Dymkens orphan investigation
(Scribe plan #540). The pre-2030 sidecar-synthetic Source pattern
(`sidecar:<platform>:<slug>` enabled=false rows) existed solely to
satisfy `Post.source_id NOT NULL`, and leaked into the Subscriptions
UI as phantom subscriptions. Now the data model says what's true:
filesystem-imported content with no live subscription has NULL
source_id, full stop.
## Schema (alembic 0030)
- `post.artist_id` — NEW NOT NULL FK to artist (CASCADE). Backfilled
from source.artist_id in the migration. Indexed for the artist-filter
queries.
- `post.source_id` — NOT NULL → nullable; FK ondelete CASCADE → SET
NULL. Deleting a Source detaches its Posts instead of destroying
archived content (subscription ends, archive stays).
- `image_provenance.source_id` — same nullable + SET NULL.
- Partial unique index `uq_post_artist_external_id_null_source` on
(artist_id, external_post_id) WHERE source_id IS NULL — guards
filesystem-import dedup since the existing source-bound unique
ignores NULLs (Postgres treats NULL != NULL).
- Sidecar synthetic Sources deleted: NULL out FKs in post,
image_provenance first, then DELETE FROM source WHERE url LIKE
'sidecar:%'. The Dymkens cleanup.
## Model + service changes
- `Post.source_id` → `Mapped[int | None]`; new `Post.artist_id`
denormalized.
- `ImageProvenance.source_id` → `Mapped[int | None]`.
- Importer: `_source_for_sidecar` (synthetic-creating) →
`_lookup_source_for_sidecar` (returns None when no subscription).
`_find_or_create_post` takes required `artist_id`; matches on
(source_id, external_post_id) for source-bound posts or
(artist_id, external_post_id) for NULL-source posts.
- Service queries switched off the Source detour to use Post.artist_id
directly: post_feed_service.scroll/around/get_post (LEFT JOIN to
Source so NULL-source posts surface); artist_service date_row/
activity/post_count; provenance_service.for_image/for_post (LEFT
JOIN); gallery_service._provenance_exists_where_artist via
Post.artist_id instead of ImageProvenance.source_id → Source.
- `_to_dict` and provenance dict-builders emit `"source": null` for
NULL-source rows.
## Frontend
- `ProvenancePanel.vue` + `PostCard.vue`: render `e.source?.platform
?? 'filesystem import'` so NULL-source posts get a clear
"filesystem import" affordance instead of a NaN crash.
## Tests
- `test_importer_upsert_helpers`: removed the four synthetic-anchor
tests; added `_find_or_create_post_idempotent_with_null_source`
(dedup via the partial unique index) and
`_lookup_source_for_sidecar_returns_*` (existing-subscription +
none cases). The existing `_find_or_create_post_idempotent` now
also passes `artist_id` and asserts it.
- 8 other test files updated: every direct `Post(...)` construction
gains `artist_id=<artist>.id`. The `_seed_post` helper in
`test_post_feed_service` looks up artist_id from the source row so
callsites stay one-arg.
## Verification on deploy
After alembic 0030 runs:
- `SELECT COUNT(*) FROM source WHERE url LIKE 'sidecar:%'` → 0.
- `SELECT COUNT(*) FROM post WHERE source_id IS NULL` → count of
filesystem-imported posts (Dymkens + any other historical).
- Every `post.artist_id` non-null; consistent with source.artist_id
for source-bound rows.
- Subscriptions tab: no Dymkens phantom row.
- Artist detail → Posts/Gallery: Dymkens's content still reachable
via Post.artist_id.
- Provenance panel renders "filesystem import" chip for NULL-source
posts; PostCard same.
## Out of scope
- UI to manage/delete orphan NULL-source Posts. Data model is right;
UI follows if operator wants it.
Operator-asked 2026-06-01: each row in the "X sources are failing"
rollup needs a button to surface the last run's stdout/stderr/error
without leaving the Downloads tab to find the matching event row.
Wired:
- `downloadsStore.loadLastForSource(sourceId)` two-steps via
`GET /api/downloads?source_id=N&limit=1` (most recent event) then
the existing `loadOne(id)` for the full detail payload. Returns
null if no events exist (toast warns).
- `FailingSourcesCard` row: new text button `Logs` with
`mdi-text-box-search-outline`, per-row spinner via `logLoadingIds`,
emits `view-logs` with the source record.
- `DownloadsTab.onViewFailingLogs` is the handler — same
`DownloadDetailModal` instance the event-row clicks use.
Operator-flagged 2026-06-01: clicking a Provenance post title from the
view modal opens the post in the posts feed scoped to that artist
(`/posts?post_id=N&artist_id=A`), but the older/newer infinite-scroll
loaded UNFILTERED global posts instead of staying in the artist's
stream. Root cause: the posts store's loadAround/loadOlder/loadNewer
sent only `{around, cursor, direction}` — never the `artist_id` /
`platform` filters. Backend `/api/posts` accepts them on every path
(post_feed_service.scroll filters via `Source.artist_id`); the
frontend just wasn't passing them.
Fix:
- `posts.js` `loadAround(postId, filters)` now takes the filter
snapshot and stores it. `_aroundParams(extra)` mixes the active
filters into around/cursor calls.
- `PostsView.vue` `loadAroundAndAnchor` passes `artist_id` +
`platform` from the route query.
Two operator-flagged UX gaps from 2026-06-01:
1. **Esc trapped inside the tag-entry field**
The autofocused TagAutocomplete input made Esc-to-close unreachable
because ImageViewer's keydown handler bailed early on
`isTextEntry(ev.target)` for every key including Escape. Now Escape
closes the modal from text inputs too — except when a Vuetify
overlay is open (FandomPicker, autocomplete dropdown, suggestion
3-dot menu), in which case that overlay's own Esc-handling fires
instead of closing the whole modal mid-interaction. Detected via
`.v-overlay--active`. Arrow keys still gate on isTextEntry so the
tag input handles typing without navigating images.
2. **Provenance cards expanded the side panel unbounded**
When an image had many ImageProvenance entries the cards stack
pushed the Tags section below the fold. Wrapped the cards in a
`.fc-prov__cards` container with max-height 270px (≈2.5 cards) and
thin overflow scrollbar. Title stays anchored at top; Tags panel
sits at a consistent position below regardless of provenance count.
`${GITHUB_SHA:0:7}` substring expansion is bash-only; the runner
executes the step via dash/BusyBox sh and errored with
`Bad substitution` (act_runner workflow shell, observed on the
65386f0 main-push run). Switched to
`SHORT_SHA=$(printf '%s' "$GITHUB_SHA" | cut -c1-7)` which works in
both shells. Both build-web and build-ml tag-determination steps
updated.
Two operator-asked modal UX changes (Scribe plan #514):
1. **Post title click → artist-scoped posts feed**
- The bold post title in ProvenancePanel is now the primary click
target. Click navigates to /posts?post_id=N&artist_id=A; the
PostsView already filters on `artist_id`, so the user lands in
that creator's stream, not the global one.
- The redundant "View post" link is removed. "Show description"
stays as the only action link below the meta line.
- Title is styled as a button-link: accent color, hover underline,
focus ring for keyboard nav (family rule 24 — UI quality bar).
2. **Suggestion rows look like buttons**
- SuggestionItem becomes a chip-card: visible border, hover/focus
background, unified container for name + score + actions
(operator's "nothing visual to unify the buttons to their object"
complaint).
- Accept is an explicit tonal pill button labeled "Accept"
(operator's pick over whole-row-clickable). 3-dot menu retains
alias/dismiss, now `variant="outlined"` so it reads as a button.
- Score is a fixed-width monospace pill on the right.
- "+ new" badge upgraded to a pill chip with accent border so
it's visibly an annotation, not part of the tag name.
Four coupled operator-asked changes to the view modal (Scribe plan #509):
1. **Autofocus tag entry on modal open** — TagAutocomplete grabs focus
in onMounted/nextTick so the caret is in the input the moment the
modal renders. No click needed to start typing.
2. **General suggestions expanded by default** — SuggestionsPanel's
general-category group now mounts with `:default-open="true"`.
Operator can collapse if too noisy, but the v1 frame shows them.
3. **Lower general threshold default 0.95 → 0.50** — MLSettings.
suggestion_threshold_general default matches character. Alembic
0029 also bumps the existing singleton row's value if it's still
at the old 0.95. Operator can re-tune from Settings → ML.
4. **Retire `copyright` + `artist` as ML suggestion categories** —
neither feeds a Tag.kind (`artist` retired in FC-2d-vii-c, never
really existed as a copyright tag-kind). They were surfaced in the
suggestions pipeline + threshold settings UI but had no follow-
through. Drop from SURFACED_CATEGORIES, suggestions._threshold_for,
ml_admin GET/PATCH allowlist, MLSettings columns (alembic 0029
drops the two columns), frontend CATEGORY_ORDER + CATEGORY_LABELS,
SuggestionsPanel.peopleCats, AliasPickerDialog kind-check, and
MLThresholdSliders rows.
Out of scope (intentional): `tag_kind` Postgres enum still includes
`artist` for historic Tag row queryability (per the model comment);
no operator pain reported, no enum-shrink needed.
Tests:
- test_surfaced_categories asserts {character, general}, excludes
artist + copyright.
- test_threshold_for_artist_is_unsurfaced extended to cover copyright.
- test_get_and_patch_settings asserts new 0.50 default and the absent
artist + copyright keys in the GET payload.
Adds an immutable per-commit docker tag to every main-push build:
`git.fabledsword.com/bvandeusen/fabledcurator{,-ml}:c-<short_sha>`,
alongside the existing floating `:main` + `:latest`. Implements the
new family release-posture rule "Tags are milestones, not gates —
commit-SHA images are the rollback unit" so rollback to any commit
on main is `docker pull …:c-<sha>` with no release ceremony required.
Behavior change summary:
- main-push: was {:main, :latest} → now {:main, :latest, :c-<short_sha>}
- tag-push (opt-in vYY.MM.DD only, no .N): unchanged
- safety-net dev: unchanged
No code changes; the rule is about how the tag list is constructed.
Tag-push workflows stay as-is — vYY.MM.DD milestone cuts can still
fire them when the operator wants a labeled checkpoint.
Operator-asked 2026-05-31 (during sidecar synthetic anchor cleanup):
"the add source/subscription button idea to the firefox extension so
it can tell me if a source/artist is added or not and offer an option
to add it if it isn't." Plan tracked in Scribe task #507.
## Backend
- `ExtensionService.probe(url)` — read-only resolution. Reuses
`_derive` for platform+slug, then 2 SELECTs. Returns one of:
- `source_match` (exact (artist, platform, url) Source exists)
- `artist_match` (artist exists, this URL isn't a Source yet;
collapses the sidecar-synthetic-only case from v26.06.01.0)
- `new` (neither exists)
- `unknown_platform` (URL didn't match any artist-page regex)
- `GET /api/extension/probe?url=...` route with `X-Extension-Key`
auth posture matching `/quick-add-source`. Read-only, side-effect
free.
- 6 backend tests in tests/test_api_extension.py covering each state
+ auth + invalid URL.
## Extension
- `api.js`: `probeSource(url)` mirroring `quickAddSource` shape.
- `background.js`: `PROBE_SOURCE` + `OPEN_ARTIST_PAGE` handlers. The
latter strips the `/api` suffix from configured `apiUrl` (placeholder
format per options.html) and opens `${base}/artist/{slug}` in a new
tab via `browser.tabs.create`.
- `content-script.js`: probe-first render — on page-load and SPA
navigation, asks the backend for the URL's state and renders the
chip in the matching color/copy on FIRST paint instead of flashing
generic "Add" and updating after. Click handler branches:
`source_match` → OPEN_ARTIST_PAGE; `artist_match`/`new` → existing
ADD_AS_SOURCE flow (then re-probes so the chip flips green
immediately, no wait for next nav).
- `content-script.css`: three state-color modifiers
(--new, --artist-match, --source-match) on the FC parchment-on-slate
palette. Sage for already-added, amber for artist-exists, accent
orange for new.
## Versioning
- `extension/manifest.json` + `extension/package.json` → 1.0.6.
build.yml's sign-extension job will fire on push to main since no
`ext-1.0.6` Forgejo/Gitea release exists yet — exercises the
regenerated AMO keys end-to-end.
## Behavior on the sidecar-synthetic case
Filesystem-imported "Dymkens"-style artist with only a sidecar
synthetic Source: probe returns `artist_match` (not `new`), so the
chip reads "+ Add Patreon source to Dymkens" rather than offering to
recreate the artist. Clicking adds the real Source; existing
`_source_for_sidecar` preference logic (v26.06.01.0) routes future
gallery-dl Posts to the real one.
Three coupled operator-reported pains from the 2026-05-31 download
event audit:
1. `[patreon][warning] Not allowed to view post N` was bubbling up as
an error event, bumping consecutive_failures and parking the source
in "needs attention." The classifier's tier-gated branch was gated
on `return_code in (1, 4)`. Gallery-dl returns a different exit
code for mixed-failure runs (e.g. paywall warnings + a missing
yt-dlp dep flipping the exit bits), so the branch never fired and
the path fell through to UNKNOWN_ERROR. Widen the gate: when no
source-level error fired AND tier-gated warnings are present,
classify as TIER_LIMITED regardless of return code.
2. Knuxy event #38275 (2026-05-31) ran 30 min and finalized with
"stranded by recovery sweep (no terminal status after time_limit)"
+ empty stdout/stderr. Root cause: subprocess.run timeout (900s)
and Celery soft_time_limit (900s) raced; when Celery won, SIGKILL
wiped the in-memory captured output and the DownloadEvent ended up
empty-logged 18 minutes later when the sweep finalized it. Drop
gallery-dl's default subprocess timeout to 870s — a 30s margin
shy of Celery's soft limit — so subprocess.TimeoutExpired always
wins the race and captures the partial stdout/stderr via the
existing handler.
3. `[downloader.ytdl][error] Cannot import yt-dlp or youtube-dl` was
firing on every video attachment, causing per-item download
failures that masked legitimate tier-gated classification.
Add yt-dlp>=2025.1 to requirements.txt. Once it's in the image,
video posts download normally and the per-item failure noise
disappears.
Tests added:
- pure tier-gated stderr with exit code 128 → TIER_LIMITED + success
- mixed tier-gated + yt-dlp + per-item failures → still TIER_LIMITED
Two coupled bugs surfaced 2026-05-31 by the Subscriptions UI showing
"phantom" subscriptions like `sidecar:patreon:dpmaker`:
1. `SourceService.list()` returned every Source, no filter on URL.
alembic 0022 (2026-05-26) consolidated old per-post-URL Sources into
one canonical row per (artist, platform); when no real campaign URL
was salvageable it rewrote the canonical to `sidecar:<plat>:<slug>`
enabled=false as a disabled anchor. The UI then listed those
anchors as if they were polls — disabled, but visible. Fix: `list()`
excludes `url LIKE 'sidecar:%'` by default; `include_synthetic=True`
opts back in for admin tooling.
2. `importer._source_for_sidecar` picked the lowest-id Source for
(artist, platform). When alembic 0022 had rewritten a per-post row
into a synthetic anchor (lower id) AND the operator later added the
real subscription (higher id), every gallery-dl download silently
attached its Post to the SYNTHETIC instead of the real Source. Fix:
prefer a non-`sidecar:%` URL when one exists; fall back to the
synthetic; only create a new synthetic when nothing exists for
(artist, platform).
alembic 0028 is the data half: for every (artist, platform) with both
a synthetic AND a real Source, pre-merge Post+ImageProvenance
collisions on the canonical, bulk-repoint Posts/ImageProvenance/
DownloadEvent.source_id onto the real Source, and delete the
synthetic. Lone synthetics (no real twin) are left intact — they
anchor real imported content the operator may still want; the
list-filter hides them so they no longer surface as phantoms.
Today's platform-cooldown commit (61ce1ce) only filtered the scan tick
— manual /api/sources/<id>/check still bypassed it. Operator-flagged
2026-05-30: clicked "Retry failed" on a Patreon failure pile and saw
every one go 'queued' without realising the cooldown wasn't in the
loop. Bulk retry with N sources on a cooled-down platform bowls right
back into the rate limit the cooldown is trying to prevent.
**Backend (`/api/sources/<id>/check`):**
- Reads optional `?force=true` query flag.
- Without force: queries `active_platform_cooldowns` (renamed from the
private `_platforms_in_cooldown` since it's now a cross-module API).
If the source's platform is in cooldown, returns **202** with
`{status: 'deferred', platform, cooldown_until}` — no event created,
no dispatch.
- With force: cooldown skipped entirely.
- In-flight guard always applies (no point creating duplicate pendings).
**Frontend (`sourcesStore.checkNow(id, {force=false})`):** new optional
`force` flag → adds `?force=true` to the URL.
**Frontend (`DownloadsTab`):**
- `onRetrySource` (single-source RETRY click): passes `force: true` →
explicit operator override, useful for rapid auth-fix testing.
- `onRetryAll` (RETRY ALL + MaintenanceMenu "Retry failed"): no force →
cooldown respected. Tallies `deferred` alongside `queued` /
`already_running`; toast reads e.g. *"5 queued, 12 deferred
(cooldown), 3 already running"*. That count is the operator's
diagnostic answer for "is rate-limit the cause of most failures?"
(12-of-20 deferred → yes; 0 deferred → no).
**Auto-resume:** no new sweep needed. Deferred sources still have stale
`last_checked_at`, so the next scan tick after the cooldown AppSetting
expires picks them up via `select_due_sources` (which already filters
on `active_platform_cooldowns`).
Tests: two new — deferred-on-cooldown returns 202 with the right body
and no dispatch; force=true overrides the cooldown and creates the
event normally.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Operator-flagged 2026-05-30 (round 3): the all-parallel fetch was fast
but could let later chunks arrive ahead of earlier ones — even when
each chunk is a random sample, that "later chunk loads first" risk made
the load order non-deterministic. And the original goal behind asking
for batching was a faster first-image-on-screen, which neither sequen-
tial nor parallel really addressed cleanly.
Switched the loadInitial flow to a PIPELINE:
- Only one fetch in flight at any moment (in-order arrival, no race).
- The NEXT fetch kicks off as soon as the current one resolves (NOT
after its trickle finishes), so the next RTT overlaps the visible
trickle window — round-trips are hidden behind the animation cadence.
- PAGE 5 → 3 + INITIAL_BATCHES 12 → 20 (total still 60). Smaller chunk
→ first chunk's items appear sooner (a chunk of 3 trickles in 240ms,
well within one RTT, so by the time chunk 2 is in-hand the first
trickle is just finishing).
Trickle, sequence-token guard, and infinite-scroll behaviour unchanged.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Operator-flagged 2026-05-30 (round 2): the batched-loads change improved
consistency but still felt "chunky" — 5 items appeared together, then a
~200 ms API round-trip pause, then another 5. And infinite-scroll fired
too late when a single tall image (long manga page) made one masonry
column much taller than its siblings.
**Cadence (showcase store):**
- Fire all INITIAL_BATCHES fetches in PARALLEL — collapses the per-
batch round-trip gap so all the data arrives in ~1 RTT instead of 12
sequential RTTs.
- Trickle each response's items into images.value one at a time with
APPEND_DELAY_MS = 80ms between each (≈ the MasonryGrid stagger
animation, 70ms). User sees a smooth steady stream.
- fetchPage (the infinite-scroll path) uses the same trickle so its
5-item appends also cascade one-by-one instead of popping together.
- Sequence token guards against a fast shuffle / mount-then-shuffle
interleaving two trickles into the same images.value.
- Dropped useAsyncAction here — the parallel-fetch-then-trickle flow
doesn't fit its single-wrap-call shape cleanly; inline loading/error
state is clearer.
**Infinite-scroll trigger (MasonryGrid):**
- Pass `rootMargin: '2400px'` to useInfiniteScroll (was the 600px
default). The masonry sentinel sits at the bottom of the container,
whose height = MAX(column heights). A tall image in one column pushes
the sentinel ~2× viewport below where the user is actually reading
(the bottom of the SHORTER columns). 2400px ≈ 2-3 screen-heights of
pre-emptive trigger, comfortable for typical tall manga heights.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Every archive import was failing immediately with "AssertionError:
daemonic processes are not allowed to have children" (operator-flagged
2026-05-30 — import_archive_file crashes in 49ms with the assertion).
Celery's prefork pool runs tasks in daemon processes; Python's
multiprocessing module refuses to let daemons spawn children, which is
exactly what probe_archive was doing via mp.get_context("spawn")
.Process. The Layer-3 crash-isolation feature added 2026-05-28 was
effectively a hard-blocker on the very import path it was meant to
protect.
Switched probe_archive to subprocess.run — no daemon restriction, still
isolates the probe (a probe segfault/OOM exits non-zero, doesn't kill
the worker). The probe body lifted to a tiny runner module
(_archive_probe_runner) that imports the unchanged _run_probe helper
and prints a single JSON line; parent parses stdout, returns the
ProbeResult exactly as before (timeout, signal, OOM, clean-rejection,
ok — all preserved).
cwd for the subprocess is the repo root derived from __file__ parents
so `python -m backend.app.utils._archive_probe_runner` resolves both in
the Celery container and pytest, regardless of where the worker was
launched.
Test refactor: test_archive_probe_target_bomb_guard →
test_run_probe_bomb_guard. Same in-process call to the (renamed) probe
body so the monkeypatched cap still takes effect; the real subprocess
path is exercised by the existing test_probe_archive_valid_zip /
test_probe_archive_corrupt_zip_clean_rejection tests.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The showcase/gallery/artist/series/post-feed APIs were constructing
thumbnail URLs from (sha256, mime). The MIME-based extension predicate
("png if image/png or image/gif else jpg") DISAGREED with the
thumbnailer's actual on-disk extension predicate ("png if alpha else
jpg"). Result: every PNG source without transparency 404'd (URL asked
.png, disk had .jpg); every WebP/AVIF source with transparency 404'd
(URL asked .jpg, disk had .png) — despite the thumbnail file existing
on disk.
The backfill task couldn't catch these because backfill checks the
ACTUAL thumbnail_path stored on the record (correct), not the URL the
browser fetches (broken derivation). So records with valid on-disk
thumbnails kept showing as broken in the UI no matter how many times
backfill ran.
Operator-flagged 2026-05-30: "the generate thumbnails function appears
to not catch all of the failed thumbnail cases" — turned out to not be
a backfill bug at all.
Fix: thumbnail_url now takes (thumbnail_path, sha256, mime) and returns
the stored path verbatim — Quart serves /images/* 1:1 from the volume
(frontend.py:20-36), so the URL IS the disk path. Falls back to the old
sha256+mime derivation only when thumbnail_path is NULL (thumbnailer
hasn't run yet); that URL will 404 in the browser until backfill catches
it, same as before the path was tracked.
All 8 callers updated: showcase_service, gallery_service (2 sites),
artist_service, series_service, post_feed_service, tag_directory_service,
artist_directory_service. The four sites whose query was raw-tuple now
also SELECT ImageRecord.thumbnail_path.
Net effect: every record that has a valid on-disk thumbnail will now
render correctly, regardless of which extension the thumbnailer chose,
without any DB migration or backfill rerun needed.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The showcase-store change in adeee64 didn't actually land — only
gallery.js + ShowcaseView.vue made it into the commit. Without this,
ShowcaseView mounts loadInitial() which doesn't exist yet, so the
showcase blows up. Landing the matching store edit now.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Showcase and gallery were fetching the full initial batch (60 / 50
items) in a single API call. Operator-flagged 2026-05-30: items should
stream in as small batches so they render progressively rather than
blocking on one big response.
- showcase store: PAGE 60 → 5, INITIAL_BATCHES = 12 (5 × 12 = 60, same
total). loadInitial() fetches PAGE chunks in sequence; shuffle()
delegates. fetchPage() still returns one PAGE-sized chunk so
infinite-scroll also pulls 5 per trigger.
- gallery store: loadMore() limit 50 → PAGE (5), INITIAL_BATCHES = 10.
loadInitial() loops loadMore() up to INITIAL_BATCHES times, stopping
early when nextCursor becomes null (end of data).
- ShowcaseView: onMounted calls loadInitial() instead of fetchPage().
Gallery/setTagFilter/setPostFilter already routed through loadInitial.
Net visual effect: each ~5-item batch lands and renders independently;
combined with MasonryGrid's animateFromIndex logic, the showcase now
cascades batches in as they arrive instead of one big pop-in.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Operator-flagged 2026-05-30: "the fail state of timeouts doesn't show
anything other than that the task timedout and was cleaned up. I can't
tell why it ran over or if it was stuck failed or there was just that
much to get."
The TimeoutExpired branch was returning a DownloadResult with no stdout,
no stderr, no files_downloaded, and a generic "Download timed out after
N seconds" message — even though subprocess.TimeoutExpired carries the
partial output gallery-dl emitted before being killed.
Now:
- Capture e.stdout / e.stderr (coerced str if bytes; "" if None).
- Count files_downloaded from partial stdout via _count_downloaded_files.
- Surface a tail-of-stderr hint in error_message so the UI summary tells
the operator at a glance whether it was "lots of content" (high count,
clean stderr), "stuck retrying" (any count, 429-spam stderr), or "hung
silent" (zero count, "no stderr output").
- Promote error_type to RATE_LIMITED when the partial stderr matches
RATE_LIMIT_PATTERNS — gallery-dl spinning on retries through the whole
900s window is the timeout-shaped tail of a real rate limit, and the
platform cooldown should kick in for the same reason.
Existing test_download_timeout strengthened to also assert empty-partial
case stays correctly TIMEOUT-classified with no preserved output.
New test_download_timeout_preserves_partial_output_and_classifies covers
the rich-partial-output → RATE_LIMITED promotion path.
DownloadEvent.metadata already flows stdout/stderr/run_stats from
DownloadResult via _phase3_persist — no UI change needed; the existing
DownloadDetailModal will surface the captured output automatically once
the build redeploys.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
select_due_sources returned rows in undefined order (Postgres-determined,
typically PK). At tick rates that outpace download-queue throughput, a
freshly-rerun source could keep getting re-queued ahead of one that's
still waiting for its first attempt this cycle. Operator-flagged
2026-05-30:
> if there are 8 hours before a source is due again and 40 full time
> downloads can happen in that period that means that there's a chance
> the first one to fire gets back into the download queue before item 41
> has a chance to get downloaded.
Added `ORDER BY last_checked_at ASC NULLS FIRST, id` to the due-source
SELECT. Never-checked sources go first, then longest-since-checked, then
ties broken by id. Combined with Celery's FIFO `download` queue, the
oldest-overdue source in each tick now reaches a worker before any
fresher one.
Test pins the ordering: a NULL-last_checked source, a 4-hour-overdue
source, and a 2-min-overdue source come back in that exact order from
select_due_sources.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
61ce1ce added platform_cooldowns to scheduler_status(), and two pinned
key-set assertions tied to the old shape failed in CI. Updated:
- tests/test_api_sources.py::test_schedule_status_shape — direct
/api/sources/schedule-status response.
- tests/test_api_system_activity.py::test_summary_returns_rollup_shape
— nested under body["scheduler"] in the summary endpoint.
[[feedback-plan-grep-pinned-tests]] miss — should have grepped tests/
for `last_tick_at` / `auto_sources` before adding the new key.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The scan tick fired download_source.delay() for every due source without
grouping by platform; with multiple download workers, N due Patreon
sources could all hit Patreon's API in parallel and rate-limit each
other. Per-source consecutive_failures backoff REACTS to that (slows the
offender across cycles) but didn't PREVENT the first-tick burst.
When DownloadService._update_source_health sees a source error
classified as ErrorType.RATE_LIMITED, it now stamps an AppSetting row
`platform_cooldown:<platform>` with the cooldown expiry (now + 15 min,
PLATFORM_RATE_LIMIT_COOLDOWN_SECONDS). select_due_sources queries every
platform_cooldown:* key at the start of each tick and excludes every
source whose platform is in active cooldown. scheduler_status surfaces
active cooldowns as platform_cooldowns: {platform: expires_iso} so the
TopNav pipeline chip / activity summary can display them.
INSERT...ON CONFLICT DO UPDATE for the upsert so two workers racing
RATE_LIMITED responses on the same platform don't let one's
IntegrityError roll back the other's event-finalize transaction
(stranding the event for the recovery sweep). Atomic at the SQL level.
Tests cover: select_due_sources skips a platform in cooldown; other
platforms unaffected during single-platform cooldown; expired cooldown
rows don't filter; set_platform_cooldown is upsert-safe under repeated
calls.
Operator-flagged 2026-05-30 ("running multiple workers I don't know how
we'd keep the downloader from hitting a rate limit on a source").
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
SourceConfig.timeout defaulted to 3600s (1 hour), but download_source's
Celery task has soft_time_limit=900s and hard time_limit=1200s. So
gallery-dl never hit its own subprocess.run timeout — Celery always
killed it first. The hard SIGKILL leaves no terminal flip on the
DownloadEvent, which then sat pending/running until the recovery sweep
flipped it to error at 30 min from start. From the operator's seat that
was a ~30–40 min "hang" on every retry of a broken source.
Pinning the default to 900s (matching Celery's soft_time_limit) lets
subprocess.run raise TimeoutExpired cleanly inside Celery's window, and
the existing `except subprocess.TimeoutExpired` branch
(gallery_dl.py:635) captures it as a clean error_type='timeout' with a
real message. The DownloadEvent flips to error in ~15 min instead of
waiting on the recovery sweep at 30.
Per-source bumps still live in source.config_overrides for legitimately
long first-syncs. The new _DEFAULT_GDL_TIMEOUT_SECONDS constant carries
the rationale and the Celery-soft-limit dependency in its comment.
Operator-flagged 2026-05-30 on 59-source strand pile retries.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The maintenance dropdown in Subscriptions → Downloads was wired to the
filesystem-import pipeline (POST /api/import/retry-failed +
POST /api/import/clear-stuck) — the subtitles even said so ("Re-enqueue
every failed import task"), but it was contextually misplaced. From the
Downloads view "Retry failed" queued nothing the operator could see
because the action operated on import_task rows, not download_event
rows. Import-pipeline maintenance is already reachable from Settings →
Imports (ImportTaskList.vue), so removing the import wiring loses
nothing.
Rewired:
- "Retry failed" → bulk-retries the failing-sources list, same loop as
FailingSourcesCard's RETRY ALL (sourcesStore.checkNow per source).
Subtitle now matches: "Re-queue every currently failing source".
- "Force recovery sweep" → triggers recover_stalled_download_events on
demand via a new POST /api/downloads/recover-stalled endpoint. The
sweep also runs every 5 min on Beat; this is the manual fallback so
the operator doesn't have to wait for the next tick to clear newly
stranded events.
MaintenanceMenu is now stateless — emits retry-failed and recover-
stalled. DownloadsTab owns the handlers (reuses the existing
onRetryAll; new onRecoverStalled with a delayed refresh so swept rows
land in the failing rollup).
Operator-flagged 2026-05-29 — "the retry failed button in the
maintenance dropdown doesn't appear to queue anything but manual
requeues works."
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
104cac5's override sits at the same specificity as Vuetify's default
(`.v-tooltip > .v-overlay__content` on both sides). The fix relied on
source order — app.css imported after vuetify/styles in main.js — but
Vite's production bundler reorders node_modules CSS into the final
stylesheet unpredictably, so source order isn't a reliable winner.
!important on the two contrast properties (background + color) forces
the slate-on-parchment pair regardless of stylesheet load order. The
cosmetic border + shadow don't need it (Vuetify doesn't set them, so
nothing's competing).
Operator re-flagged 2026-05-29: tooltips still rendered light-on-light
on the deployed :latest after PR #33 — 104cac5 was in the bundle but
not winning. Also updated the file header so the source-order claim
isn't carried forward as gospel.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The scan tick (scan.py:_tick_due_sources_async) inserts
DownloadEvent(status='pending') and fires download_source.delay(). If the
task dies before finalizing the event — worker OOM/SIGKILL, lost task, or
a gallery-dl that didn't unwind on the 1200s hard time_limit — the event
stays in-flight forever. Every later tick then skips the source via the
in-flight guard (scan.py:168), so Source.last_checked_at is never written
and the operator sees "last check never" in the Subscriptions health
column, permanently.
cleanup_old_download_events only prunes terminal events (by design); no
existing sweep covered the pending/running case. Operator confirmed
2026-05-29 with a diagnostic query: all 43 "never checked" sources were
stranded behind stale in-flight events (eligible_stuck_inflight = 43,
every other bucket zero).
New recover_stalled_download_events task (Beat every 5 min):
- Flips DownloadEvent rows pending/running > 30 min (10 min past the
download_source 1200s hard kill, so legitimately-running tasks are
never touched) to status='error' with a sentinel message.
- Bumps each affected Source's consecutive_failures ONCE per source —
backoff is 2^N on that counter so per-event bumps would needlessly
inflate the next interval — sets last_error, stamps last_checked_at.
UPDATE...RETURNING source_id avoids a SELECT-then-UPDATE-WHERE-IN that
would hit the psycopg 65535-param ceiling on a large strand pile.
Net: the 43 currently-stranded sources unstick on the first sweep after
deploy, their health dots flip amber instead of unchecked, and the next
scan tick re-queues them.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The plain-JS frontend re-declares two backend enum value-sets by hand:
download-event statuses (downloadStatus.js) and platform keys
(platformColor.js). With no TS codegen, drift is silent — the same class
as the last_verified_at field mismatch.
tests/test_fe_be_contract.py parses both JS mirrors and asserts they equal
the backend canon (downloads._ALLOWED_STATUSES, platforms.known_platform_keys())
so a change on either side fails CI on whichever moved. Backend is the
source of truth. Documented the invariant in both JS file headers.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The IntersectionObserver-on-a-sentinel infinite-scroll pattern was copy-
pasted in 7 places. New composables/useInfiniteScroll(sentinelRef, cb)
owns the observer lifecycle (attach on mount, re-attach when the ref
changes, disconnect on unmount). Migrated GalleryGrid, MasonryGrid
(gallery+showcase), ArtistPostsTab, ArtistsView, TagsView, SeriesManageView.
PostsView left manual on purpose — its anchored mode does bidirectional
scroll-position preservation that doesn't fit the simple composable.
I4(a) (timestamps) is effectively already done: the UTC displays were
converted earlier; the remaining toLocaleString uses already render local
time, so they're left as-is rather than churned for format-only consistency.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
New /api/system/activity/summary aggregates scheduler health + per-queue
pending depths + running count + 24h failure count in one cached call (safe
to poll app-wide). PipelineStatusChip lives in the TopNav on every page: a
compact running/queued/failing chip with a scheduler-health dot that expands
to a popover (scheduler, busy queues, counts, link to downloads). Polls the
summary every 8s, paused when backgrounded. Reuses the queue-cache read via
_queues_cached(). + API test for the summary shape.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The row template renders status + platform (PlatformChip) + time + counts;
the artist isn't displayed there. Assert on 'Completed'/'Patreon' instead.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds @vue/test-utils + happy-dom and mounts CredentialCard, DownloadEventRow,
PostCard, ActiveDownloadsPanel, QueueStatusBar, asserting they render without
throwing and surface key content — catching the dead-binding / render-error
class (e.g. the last_verified_at regression, guarded explicitly). Vuetify
components are left unresolved (Vue renders unknown elements + slots), so no
Vuetify-plugin setup is needed; only RouterLink is stubbed. Per-file
happy-dom env via docblock keeps the existing node-env specs untouched.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
ruff is pre-installed in the ci-python image, so a new `lint` job runs it
with no dependency install and fails in seconds — surfacing the common lint
bounce class without waiting on the backend job's ~30-60s wheel install.
Dropped the now-redundant ruff step from backend-lint-and-test (same job
name, required-checks unchanged).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Provenance "View post" deep-links to /posts?post_id=X, which now opens the
feed centered on that post with infinite load in BOTH directions.
Backend: PostFeedService.scroll gains a direction (older|newer); new
around(post_id) returns a window of newer + the post + older with a cursor
for each end. /api/posts accepts ?around= and ?direction=. + API tests.
Frontend: posts store gains loadAround/loadOlder/loadNewer (older appends,
newer prepends) with per-end cursors; PostsView's anchored mode scrolls to
the post, observes top + bottom sentinels, and preserves scroll position on
upward prepend so the page doesn't jump. Normal feed mode unchanged.
Closes the remaining half of the post-navigation work.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Each maintenance button (ML backfill, centroid recompute → ml queue;
thumbnail backfill → thumbnail queue) now shows a status bar with the live
pending count for its Celery queue, so the operator can see work is already
queued/running before re-triggering and piling on. MaintenancePanel polls
/api/system/activity/queues every 4s; QueueStatusBar reads the depth and
turns warning-colored when the queue is busy.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
New ActiveDownloadsPanel pinned at the top of the Downloads tab shows what's
running (pulsing dot + live mm:ss timer counting from started_at) and what's
queued, so the operator can see activity at a glance without the running
filter. Backed by store.loadActive() which fetches running + pending
independent of the feed filter; refreshed every 4s by the existing live poll.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Download event times showed raw UTC wall-clock (iso.slice). Added
formatDateTime()/formatLocalDate() (local tz, robust to naive vs tz-aware
ISO) and applied them to the download row + detail modal datetimes and the
credential/artist date displays. formatPostDate stays UTC (date-only,
locale-stable, unit-tested).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Replaced the subtle 12px fade with a perspective rotateX(-28deg) tilt that
flips up and settles flat with a slight overshoot, staggered 70ms so tiles
cascade in one at a time. Makes the showcase read as an experience rather
than a quiet fade. Tunable (tilt/stagger/duration); reduced-motion safe.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
FC already matched IR's staggered entry fade-in; this adds the missing
piece — hover zoom (scale 1.03) + brighten (1.1) on each thumbnail, with
overflow:hidden so the scaled image clips to the rounded card. Disabled
under prefers-reduced-motion.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- Post titles arrived as stored HTML (e.g. <strong>…</strong>) rendering as
literal markup. New toPlainText() strips tags; titles render plain + bold
(ProvenancePanel and PostCard).
- Removed "View original post" from the provenance panel (modal) — the
open-original button lives on the PostCard (the post view).
- Provenance "View post" now navigates to the /posts feed (post_id query),
not the gallery image grid. (Feed in-context landing lands next.)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The whole view scrolled instead of just the subscription list. Made the
hub a viewport-height flex column (tabs stay fixed) with the v-window as
the single internal scroll container; the per-tab sticky control bars now
pin to the window top (top:48px -> 0). Operator-flagged 2026-05-28.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
ArtistView and SubscriptionsView both two-way-bound a tab ref with the
?tab= query param via identical tab->URL and URL->tab watchers. Extracted
useTabQuery(validTabs, defaultTab) — defaultTab accepts a string or a
function (ArtistView's default is postCount-dependent), and resolve() is
exposed so ArtistView can re-apply it after the artist loads. Both views
drop their now-orphaned ref/useRouter imports.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
_find_or_create_source, _source_for_sidecar, and _find_or_create_post each
repeated the SELECT → savepoint-INSERT → on-IntegrityError rollback+re-SELECT
pattern. Extracted _get_or_create(stmt, factory): the statement is reused for
the scalar_one_or_none lookup and the scalar_one post-conflict re-fetch, so
all three are reproduced exactly. Centralizing the race-safe pattern in one
place also reduces the risk of the copies drifting (the bug class banked
2026-05-26).
Left _upsert_artist (no savepoint by design) and the ImageProvenance void
ensure-exists block (no return / no re-select) alone — they don't fit.
The rest of the ingest pipeline was already DRY: sidecar parsing lives in
utils/sidecar.py, per-platform quirks in the platforms package, and
_safe_ext/_categorize_error/_build_config are each single-instance.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The `select(ImportSettings).where(id == 1)).scalar_one()` singleton load was
repeated 15× across services, API, and 5 task modules. Added async load() +
sync load_sync() classmethods on the model and migrated all 15 full-row sites
(callers already imported ImportSettings, so no new imports; dropped download's
now-orphaned select import). Left maintenance.py's deliberate column-select
(import_scan_path only) as-is.
Rest of the service layer was already adequately DRY — the Record/to_dict
pattern is only 2 instances and the savepoint find-or-create recovery is
correctly per-entity, so neither was forced into a shared abstraction.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
download/migration/scan each defined an identical _async_session_factory()
(fresh per-invocation async engine — async connections are event-loop-bound
so each asyncio.run() task needs its own engine, unlike the process-wide
_sync_engine). Moved it to tasks/_async_session.py; the 3 files import it
and drop their now-orphaned sqlalchemy.ext.asyncio / get_config imports
(migration keeps AsyncSession for a type hint). Call-site try/finally
dispose left as-is to avoid re-indenting the critical task bodies.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
8 blueprints each defined an identical _bad() (two variants: with/without
detail). Extracted error_response() into api/_responses.py; each blueprint
now imports it `as _bad` so call sites are unchanged. The detail-aware
canonical subsumes both variants. Left settings.py's distinct _bad_int and
the inline jsonify error sites (not duplicated helpers).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
New composables/useAsyncAction.js owns the loading/error/try-finally
lifecycle. Migrated 11 stores: credentials, downloads, sources, posts
(error=raw) + ml, artistDirectory, tagDirectory, showcase, suggestions,
seriesReader, modal (error=message). The errorAs option preserves each
store's existing error shape so store.error keeps the same type for
components (~50 consumption sites unchanged). Stores whose catch also
reset data (suggestions/seriesReader/modal) clear it upfront instead.
Deliberately NOT migrated (special control flow, would change behavior):
artist (conditional 404 catch + dual loading states), migration (rethrows),
gallery (inflight-id stale-response guard), and the Shape-B no-catch /
loaded-guard / keyed-cache stores.
Net -77 lines.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Removing the create_app import left 2 blank lines before pytestmark in 9
files where ruff isort wants 1. Also stripped two pre-existing
whitespace-only blank lines surfaced by the file change.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Removed the app/client fixtures duplicated across 36 test files (two
variants: separate app + client(app), and a self-contained client() that
called create_app inline) and the now-unused create_app imports. Both
fixtures now live once in conftest.py. test_suggestions_bulk keeps its
import (builds the app inline in two tests); test_health drops its local
client + unused pytest_asyncio.
Net -415 lines.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
D1 scheduler visibility: AppSetting last-tick stamp on every Beat tick +
GET /api/sources/schedule-status (last_tick_at/next_due_at/due_now/auto_sources)
+ SchedulerStatusBar on the Subscriptions tab (re-polled every 30s).
D2 failing-source rollup: ?failing=true on the sources list + FailingSourcesCard
on Downloads with per-source and bulk "retry" (re-runs the feed via /check).
D3 activity sparkline: GET /api/downloads/activity hourly buckets + CSS bar
chart by the stat chips (failures stacked in error color); refreshes on live poll.
D4 credential staleness: surface last_verified age + "re-verify recommended"
warning past 30d; also fixes the dead last_verified_at field-name mismatch so
the verification row renders at all.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Operator-flagged 2026-05-28: the Subscriptions/Downloads dashboards were
full-bleed and thin on filtering. Chosen via AskUserQuestion.
**Layout**
- SubscriptionsView capped at a centered max-width 1600px (covers all
three subtabs) so rows aren't a mile wide on ultrawide monitors.
- Sticky control headers on both tabs (top: 48px, below the top nav,
opaque bg) so filters/stat-chips stay reachable while scrolling a
long list.
**Downloads filters (all four requested)**
- Stat chips are now clickable filters: click Queued/Running/Completed/
Failed/Skipped to filter the list to that status; the active chip is
outlined + elevated; re-click clears.
- Free-text search box over the loaded events (artist / platform /
error substring).
- Artist filter: the filter popover's numeric "Source ID" field is
replaced with an artist autocomplete (sources.autocompleteArtist →
artist_id, which /api/downloads already supports). Pill shows the
artist name.
- "Show no-change scans" toggle (default OFF): hides status=ok/skipped
rows with 0 files (the scheduled scans that found nothing) so real
downloads + failures stand out.
**Subscriptions**
- "Needs attention" quick-filter chip: one click to show only artists
with sources that have errors OR have never been checked; chip shows
the count and disables the status dropdown while active.
Frontend-only — backend filter params (status/artist_id/date) and the
/api/downloads endpoint already supported everything.
Operator-flagged 2026-05-28: the download event detail modal had no way
to copy the error / stdout / stderr text for researching a failure
(parity with the ErrorDetailModal Copy button shipped in v26.05.26.0).
Added per-block Copy buttons (Error, Errors & warnings, Raw stdout, Raw
stderr — the stdout/stderr ones sit in the expansion-panel title with
@click.stop so they don't toggle the panel) plus a "Copy all
diagnostics" button in the footer that assembles a single block: header
line (event id / platform / artist / status / timestamps) + error +
errors&warnings + full stdout + stderr, ready to paste into an issue.
All routed through utils/clipboard.js copyText() — the navigator.clipboard
→ execCommand fallback that works on the plain-HTTP homelab origin
(per feedback_no_secure_context_apis). Each copy shows a confirmation
toast.
Operator-flagged 2026-05-28, two asks.
**1. Credential Verify (was missing vs GS — and now actually verifies).**
GS's Verify was a stub (`TODO: implement actual verification` — just
stamped last_verified). FC does a real check, which matters given the
recent auth pain (subscribestar age cookie, HF host-only PHPSESSID):
- GalleryDLService.verify(url, platform, cookies_path, auth_token) runs
gallery-dl in `--simulate --range 1-1` mode (no download) against the
URL with the materialized credentials, then reuses _categorize_error:
returncode 0 / NO_NEW_CONTENT → valid; AUTH_ERROR → invalid; other →
inconclusive (reason surfaced). 45s timeout.
- POST /api/credentials/<platform>/verify picks an enabled Source for
the platform to probe, runs verify, and on success stamps
credential.last_verified (new CredentialService.mark_verified).
Returns {valid: bool|null, reason, last_verified?}. valid=null means
untestable (no credential, or no enabled source to point at).
- CredentialCard gains a Verify button (on credentialed cards) + a
result chip (Verified ✓ / Failed / Untestable) and a toast with the
reason. SettingsTab reloads on @verified so last_verified refreshes.
**2. Live download-activity feedback.** The Downloads tab was static —
no way to tell if downloads were succeeding without manually hitting
Refresh. It now auto-polls: stats every 4s, and the event list too
while anything is queued/running. Polling pauses when the tab is
backgrounded (document.hidden) and the list reload is skipped on idle
ticks to stay light. A pulsing "● live" indicator next to the stat
chips shows when auto-refresh is active (queued+running > 0); honors
prefers-reduced-motion.
Tests: verify endpoint — untestable with no credential, untestable with
no enabled source, valid+stamped on success (gallery-dl mocked), and
auth-failure reported without stamping.
Operator-confirmed 2026-05-28: the BlenderKnight:* tags are `archive`
kind (caught by the kind purge), but the `source:patreon`-style tags
are IR's old `source` kind that fell back to `general` during migration
(FC's enum has no `source` kind) — so they can't be matched by kind.
Broadened the purge to a two-rule match and renamed it for accuracy
(all dev-only, unreleased):
- cleanup_service.purge_tags_by_kind → purge_legacy_tags. Predicate is
now `kind IN (archive, post, artist) OR name LIKE 'source:%'`
(LEGACY_NAME_PREFIXES). Preview classifies each row into by_kind OR
by_prefix (source:* counts once under the prefix bucket regardless of
its general kind).
- endpoint /tags/purge-retired-kinds → /tags/purge-legacy. dry-run
returns by_kind + by_prefix + count + sample.
- store purgeRetiredKindTags → purgeLegacyTags.
- Tag Maintenance card copy + breakdown updated to show both buckets;
button reads "Preview/Delete legacy tags".
Tests updated + extended: dry-run reports by_kind {archive,post,artist}
AND by_prefix {source:*}, plain general/character tags survive; commit
deletes both the kind-matched and source:*-matched rows and leaves the
rest.
Operator-flagged 2026-05-28, two asks.
**1. Provenance posts ate the panel.** Each post card rendered its full
description (180px scroll box) inline, so a few posts pushed everything
else off-screen. ProvenancePanel now collapses the description by
default behind a per-post "Show description ▾ / Hide ▴" toggle (state
keyed by provenance_id, reset when the viewed image changes). Cards stay
compact — platform/date/title/meta/actions — and the operator expands
only the descriptions they want.
**2. Purge tags of retired/system kinds.** The IR migration left
`archive`/`post`/`artist`-kind tags (e.g. `BlenderKnight:Hannah_BJ_Loops`)
that FC no longer creates — the tag input only makes
character/fandom/series/general, and provenance + artists are their own
systems now. (meta/rating were already hard-deleted by alembic 0023.)
- cleanup_service.purge_tags_by_kind(kinds=PURGEABLE_TAG_KINDS) — counts
(dry_run) or deletes tags whose kind ∈ (archive, post, artist).
CASCADE clears the image_tag / alias / allowlist / etc. rows.
- POST /api/admin/tags/purge-retired-kinds (Tier-A, dry-run preview
returns per-kind counts + sample names — the preview IS the
verification of exactly what'll be deleted before committing).
- Tag Maintenance card gets a second section: "Preview retired-kind
tags" → per-kind breakdown + sample → "Delete N retired-kind tag(s)".
Tests: dry-run counts by kind (general survives), commit deletes only
the retired kinds (general + character survive, retired count → 0).
NOTE: a dry-run preview will show exactly which kinds/counts are
present. If the operator's noisy tags turn out to be `general` (e.g. an
IR `source:patreon` that fell back to general during migration), they
won't be caught by the kind purge — the preview makes that visible so
we can decide on a name-based pass separately.
bvandeusen
