feat(ml): read suggestions + allowlist from image_prediction (#768 step 2)

Switch every prediction READER off the JSON column onto the normalized
image_prediction table. Parity by construction: each reader loads the same
{raw_name: {category, confidence}} dict it consumed before (via small
_load_predictions helpers), so all downstream threshold/alias/merge/consensus
logic is byte-identical — only the data source changed.

- suggestions.SuggestionService.for_image (and for_selection via it)
- ml.apply_allowlist_tags (iterates images that have prediction rows)
- importer re-import reset deletes the image's prediction rows
The tagger_predictions JSON column is still dual-written (step 1) so it stays
valid during transition; the backfill task's NULL check still works. Removing
the JSON write + DROP column + retiring the #764 prune is the cleanup
follow-up (needs a quiesced-worker window for the DROP lock).

Tests: shared tests/_prediction_helpers.seed_predictions seeds the table;
read-path tests (suggestions, bulk consensus, allowlist apply, API) seed there
instead of ImageRecord.tagger_predictions.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

tests/test_api_suggestions.py

@@ -15,14 +15,17 @@ def eager():
 
 
 async def _img(db, preds):
+    from tests._prediction_helpers import seed_predictions
+
     img = ImageRecord(
         path="/images/s.jpg", sha256="s" * 64, size_bytes=1,
         mime="image/jpeg", width=1, height=1,
         origin="imported_filesystem", integrity_status="unknown",
-        tagger_predictions=preds,
     )
     db.add(img)
     await db.commit()
+    await seed_predictions(db, img.id, preds)
+    await db.commit()
     return img