feat(fusion_accounting_assets): asset anomaly persisted model

- 3 anomaly types: behind_schedule, ahead_of_schedule, low_utilization
- 3 severity levels: low, medium, high
- expected / actual / variance_pct (mirrors anomaly_detection service output)
- 4-state lifecycle: new -> acknowledged -> resolved (or dismissed)
- action_acknowledge / action_dismiss / action_resolve transitions
- ondelete='cascade' on asset_id (anomalies follow the asset)
- 4 new tests (63 total)

Made-with: Cursor

fusion_accounting_assets/__manifest__.py

@@ -1,6 +1,6 @@
 {
     'name': 'Fusion Accounting Assets',
-    'version': '19.0.1.0.9',
+    'version': '19.0.1.0.10',
     'category': 'Accounting/Accounting',
     'summary': 'AI-augmented asset management with depreciation schedules.',
     'description': """