feat(fusion_accounting_ai): bank_rec_prompt for AI re-rank step

Provider-agnostic system + user prompt builder for the confidence
scoring pipeline's Pass 3 (AI re-rank). Output contract is JSON with
"ranked" array; works with OpenAI, Claude, and local OpenAI-compatible
servers (LM Studio, Ollama).

Made-with: Cursor

fusion_accounting_bank_rec/tests/__init__.py

@@ -9,3 +9,4 @@ from . import test_reconcile_engine_unit
 from . import test_reconcile_engine_property
 from . import test_factories
 from . import test_reconcile_engine_integration
+from . import test_bank_rec_prompt