feat(vision): immagini di riferimento, per comporre senza camera

VisionReferenceImage (migration 013, chained after 012_vision_results):
task_id, path, station_id, device_code, calibration_snapshot,
engine_version, note, expected_json, acquired_at. station_id/device_code/
calibration_snapshot/engine_version/expected_json stay nullable - nessun
endpoint di questo task li popola ancora, arrivano coi Piani 1b/2/4.

save_reference_image salva sotto settings.upload_path (non upload_dir,
come production_export_service). preview esegue il grafo e non salva
niente - comporre non e' misurare, lo asserisce il test.

POST /api/vision/reference-images e /api/vision/preview su require_maker
(non require_measurement_tec di execute): comporre e provare sono lavoro
del Maker, non misure.

Suite intera su Python 3.11.15 locale (uv run pytest -q, SQLite
in-memory per il backend): 1 failed, 373 passed, 4 skipped in 82.35s.
Il fallimento e' quello preesistente e non correlato di
test_no_first_party_script_calls_out (fabric-debug.js locale,
git-ignored) - 371 passed prima di questo task, +2 qui.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014BBnuACZSCJqXrMYC3LUMU

src/backend/models/orm/__init__.py

@@ -7,7 +7,7 @@ from src.backend.models.orm.access_log import AccessLog
 from src.backend.models.orm.setting import SystemSetting, RecipeVersionAudit
 from src.backend.models.orm.station import Station, StationRecipeAssignment
 from src.backend.models.orm.production import ProductionRun, ProductionEvent
-from src.backend.models.orm.vision import VisionResult
+from src.backend.models.orm.vision import VisionReferenceImage, VisionResult
 
 __all__ = [
     "User",
@@ -24,4 +24,5 @@ __all__ = [
     "ProductionRun",
     "ProductionEvent",
     "VisionResult",
+    "VisionReferenceImage",
 ]