The eval (#1130) proved the frozen-embedding + trained-head spine; this lands
its production form (the first of three slices that make heads the suggestion
source, replacing Camie + centroid).
- tag_head: one logistic-regression head per general/character concept with
enough labelled positives. Weights (pgvector), honest CV-derived suggest
threshold + earned-auto-apply point, and per-concept quality metrics.
- head_training_run: persisted batch lifecycle (mirrors tag_eval_run) so the
admin card shows live + historical status across navigation.
- services/ml/heads.py: TRAIN (sync, ml worker, reuses tag_eval's proven data
loaders + metric math so production heads match measured eval numbers) and
SCORE (async, API worker — numpy via pgvector, no scikit-learn): score one
image's embedding against all heads → the rail's suggestions, cached on
(count, max trained_at) so a retrain invalidates without per-request loads.
- tasks.ml.train_heads (ml queue, commits per head so a kill leaves progress)
+ recover_stalled_head_training_runs sweep + retention(20) + 5-min beat
(rule 89).
- api/heads.py: POST /api/heads/train (one run at a time, 409 guard) + GET
/api/heads (count, graduated, last-trained, running, per-concept table,
recent runs).
- ml_settings: head_min_positives + head_auto_apply_precision, tunable via
/api/ml/settings.
Scoring isn't wired into the rail yet (slice C) and the admin UI is slice B —
this slice makes training + scoring exist and CI-verifiable. 'precision' column
stored as precision_cv (SQL reserved word). Migration 0058.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa
"Keep" on a doubted positive was a no-op, so the same confirmed-correct images
came back in "head doubts" every run (operator-flagged: reinforcement keeps
surfacing the same images). Add tag_positive_confirmation (mirror of
tag_suggestion_rejection): keep → POST /images/<id>/tags/<tag_id>/confirm, and
the eval excludes confirmed positives from the doubts list — exactly as rejected
items already drop out of the suggest list. The tag stays a positive either way
(confirmation is a "reviewed" marker, not a training change).
- model TagPositiveConfirmation + migration 0057; confirm endpoint (idempotent).
- tag_eval: _confirmed_ids + exclude from head_doubts_positive examples.
- store.confirmTag + card "keep" calls it.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Slice 1 of milestone #114 (tagging v2). Proves the frozen-embedding + trained-
head spine on the operator's own data, reusing the SigLIP embeddings already
stored on image_record — no re-embedding, no GPU.
Per concept: train a logistic-regression HEAD (positives + negatives = explicit
rejections + sampled unlabeled) vs the old single-CENTROID baseline; report
cross-validated precision/recall/AP for both, a LEARNING CURVE (AP/F1 as tagged
positives grow 10→30→100→300), and example image ids (head-would-suggest /
head-doubts-positive) to eyeball.
Persisted so the report SURVIVES navigation (operator-flagged): the run + full
report live in a new tag_eval_run row (mirrors library_audit_run); the admin
card will rehydrate from GET on mount, not transient state.
- models.TagEvalRun + migration 0056; runs on the ml queue (only worker with
numpy/sklearn) — numpy/sklearn lazy-imported so the API can still enqueue.
- services/ml/tag_eval (compute + start helper, one-running guard), tasks.ml
.tag_eval_run, api/tag-eval (POST create, GET history light / detail w/ report).
- recover_stalled_tag_eval_runs sweep + retention (keep last 20) + 5-min beat
(rule 89). scikit-learn added to requirements-ml.
- tests: param normalization + the rehydrate read-path + create/conflict.
Frontend admin card (trigger + render persisted report) follows next.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Phase 1, step 1 of moving SubscribeStar off gallery-dl onto the native core
ingester (milestone: SubscribeStar native). Mirror of the Patreon ledger:
SubscribeStarSeenMedia (skip already-ingested media on routine walks; recovery
bypasses) and SubscribeStarFailedMedia (dead-letter so persistently-failing
media stops re-burning backfill chunks). Per operator decision, dedicated
per-platform tables (not a generalized shared ledger).
filehash is String(128): a CDN content hash when the URL carries one, else a
synthesized <post_id>:<filename> key. UNIQUE (source_id, filehash) upsert key.
Registered in models/__init__; migration 0054 creates both tables (down 0053).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Capture off-platform links (mega/gdrive/mediafire/dropbox/pixeldrain) embedded
in post bodies so they're never silently dropped, and surface them in the post
view. The download worker (Phase 4) walks these rows.
- link_extract.py: pure extractor — <a href> + bare URLs, unwraps Patreon
redirect shims, PRESERVES the full url incl. #fragment (mega's key), dedups.
Reusable by every platform (runs off Post.description).
- external_link model + migration 0049: post_id/artist_id/host/url/label/status
/attempts/last_error/attachment_id/timing; CHECK whitelists (full enum incl.
worker statuses up front) + (post_id,url) unique.
- importer._sync_external_links: insert-missing on both import paths
(_apply_sidecar + upsert_post_record) so a re-import never resets a link's
status; runs for all platforms.
- post_feed_service.get_post: returns external_links (detail-only).
- PostCard: renders the links (host chip + label + status) once expanded.
- tests: extractor (5 hosts, fragment, shim unwrap, dedup), importer (record +
no-dup on reimport), serializer.
Refs FC #830.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Normalize tagger predictions out of the image_record.tagger_predictions JSON
blob into a queryable per-prediction table. Step 1 of the cutover (expand):
additive + low-risk — reads still use the JSON, this just adds the table and
keeps it populated.
- ImagePrediction(image_record_id, raw_name, category, score) — stores the
RAW tagger vocab name (not tag_id) so read-time alias→canonical resolution
is unchanged. Indexed for per-image reads + by (raw_name, score).
- Migration 0045: create table + set-based backfill from the JSON via
json_each (fast post-#764-prune). The old column stays (vestigial) and is
dropped in a later follow-up — DROP needs an ACCESS EXCLUSIVE lock on the
hot image_record table, so it waits for a quiesced-worker window.
- tag_and_embed dual-writes the rows (delete-then-insert, idempotent);
tagger_store_floor already applied in infer().
Next: switch suggestion + allowlist reads to the table, then drop the JSON
write. Plan-task #768.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Confirm-only "this post may continue this series" matcher.
- series_suggestion table (post_id, series_tag_id, score, signals jsonb, status
pending|added|dismissed, UNIQUE(post,series)); migration 0041 + two settings
knobs (series_suggest_enabled, series_suggest_threshold).
- series_match_service: weighted additive score (title-stem / same-artist /
page-continuity / shared-distinctive-tags), no single signal gating. The title
"pattern" is derived on the fly from the post titles already in a series, so it
sharpens as more are confirmed (no persisted state to drift). Candidates are
bounded to the post's artist. match_post upserts pending suggestions (UNIQUE +
on-conflict, respecting prior added/dismissed decisions).
- accept reuses add_post_as_chapter then marks 'added'; dismiss marks 'dismissed'.
- rescan_series_suggestions_task: settings-gated, time-boxed + self-resuming from
a post-id cursor (maintenance_long lane), like normalize_tags_task.
- API: GET /series/suggestions, POST .../<id>/accept|dismiss, POST .../rescan.
- Settings: enabled + threshold exposed via /settings/import.
- Tests: pure scoring helpers + matcher/accept/dismiss/rescan lifecycle + UNIQUE
dedup.
Frontend (Suggestions tab + settings card) lands next.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adds an ordered chapter layer to series. Reading order becomes
(series_chapter.chapter_number, series_page.page_number); a chapter may be a
placeholder reserving a slot, and carries an optional parsed stated-page range
used to flag missing-page gaps. An image still lives in at most one series ⇒ one
chapter (image_id stays UNIQUE).
- models: series_chapter; series_page gains chapter_id (NOT NULL, cascade) +
stated_page. Migration 0040 backfills every existing series into one
auto-chapter holding its current flat pages — no data loss.
- SeriesService: chapter CRUD (create/update/reorder/delete/merge), page→chapter
assignment, reorder_pages, chapter-aware set_cover; list_pages now returns
chapters[] + gaps[] alongside a back-compat flat pages[]. Legacy series-wide
reorder operates on the single default chapter and rejects multi-chapter series.
- API: chapter endpoints under /api/series/<tag>/chapters; POST pages accepts an
optional chapter_id.
- TagService.merge now repoints series_chapter too, so a merged series' chapters
(and their pages) survive the source tag's deletion instead of cascading away.
- Tests: new chapter suite; updated the 4 direct SeriesPage(...) constructions to
supply chapter_id.
Frontend (chapter-aware manage view + reader) lands next; until then the
existing UI keeps working via the flat pages[] + single default chapter.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A media that fails every walk (404'd CDN, deleted post, geo-blocked Mux,
persistently-corrupt bytes) used to re-error forever and re-burn chunks.
New `patreon_failed_media` table (alembic 0038, chains 0037) records
per-media attempts; once attempts reach DEAD_LETTER_THRESHOLD (3) the
ingester skips it on routine tick/backfill walks (tier-1.5, folded into the
seen/skip predicate). Recovery BYPASSES it (the operator's "try everything
again" re-attempts dead media). A clean download clears the row (recovered);
errors/quarantines upsert-increment it. Surfaced as
run_stats.dead_lettered_count.
- New PatreonFailedMedia model + migration; ingester _dead_keys /
_record_failures (on_conflict increment) / _clear_failures.
- skip = seen | dead (empty in recovery); failures recorded post-fetch on
short sessions (same pattern as the seen-ledger).
Tests: a media erroring 3× is dead-lettered + skipped (no download attempt);
recovery re-attempts a dead media and clears it on success; a clean download
clears a sub-threshold failure.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
patreon_seen_media(source_id, filehash, post_id, seen_at), UNIQUE(source_id,
filehash) — our own queryable replacement for gallery-dl's archive.sqlite3.
Routine walks skip seen media; recovery mode bypasses the ledger. filehash is
a 32-hex CDN MD5 or a video:<post>:<media> sentinel (String(128)). alembic
0037 (← 0036). Integration test covers dedup + savepoint recovery.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Per-artist "+N" accent pill on the artists directory and a "N new since
last visit" banner inside ArtistView. Counts new IMAGES (not posts) so
multi-image posts increment correctly.
- alembic 0034: artist_visit (artist_id PK, last_viewed_at NOT NULL).
Seeds every existing artist with last_viewed_at=NOW() so the badge
starts at 0 across the board — no noisy "5000 unseen images" on
first deploy.
- ArtistService.find_or_create autoseeds a visit row alongside new
artists, so freshly imported content doesn't read as unseen.
- ArtistService.overview reads pre-visit last_viewed_at, counts images
created since, then atomically UPSERTs last_viewed_at=NOW() via
postgres ON CONFLICT DO UPDATE (no SELECT-then-INSERT race per
reference_scalar_one_or_none_duplicates). Returns the pre-update
count as `unseen_count_at_visit` so the banner has data.
- ArtistDirectoryService.list_artists adds an `unseen_count` aggregate
to each card via LEFT JOIN artist_visit + conditional COUNT. NULL
last_viewed_at (artist created before this code shipped) defensively
counts as "never visited" → all images unseen.
- Frontend: ArtistCard renders an accent pill in the preview-strip
corner when unseen_count > 0 (capped at 99+); ArtistView shows a
closable v-alert banner on initial load when
unseen_count_at_visit > 0, re-arms on slug change.
Single-row-per-artist (no user_id) — rule #47 multi-user ACL is
aspirational; widens to (user_id, artist_id) PK when User lands, per
rule #22.
Scribe plan #597.
Adds Tag.kind enum (artist/character/fandom/general/series/archive/post/meta/rating),
Tag.fandom_id FK with CHECK constraint (only valid for kind='character'), and a
kind-aware uniqueness index so the same name can exist across kinds and the same
character name can exist in different fandoms.
Adds ImportBatch + ImportTask state-machine tables for scan tracking, plus a
single-row ImportSettings table (CHECK id=1) holding the importer's filter knobs.
Adds image_record.integrity_status column (defaults to 'unknown'); FC-2e
populates this via the integrity verifier.
Drops the unused tag.namespace column from FC-1 — superseded by kind.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Implements the data model from spec §3 in one go so FC-2/FC-3 don't need
schema-adding migrations of their own. Artist is the unified entity for
both gallery 'artist:' tags and GallerySubscriber Subscriptions
(is_subscription flag). ImageProvenance is many-to-one, enabling the
enrich-on-duplicate rule for downloaded content that pHash-matches an
existing record.
The SigLIP embedding column uses pgvector(1152) for SigLIP-so400m;
swapping models in FC-2 will require a column-width migration.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Configures stable constraint naming so autogeneration produces clean diffs.
Alembic uses the sync psycopg driver while the runtime app uses asyncpg.
Also fixes a .gitignore bug caught during this task: the bare 'models/'
rule for the ML weights volume was matching backend/app/models/ (Python
package). Anchored all volume rules to repo root (/images/, /import/,
/downloads/, /models/, /postgres_data/, /redis_data/).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
bvandeusen