feat(rag): optimize RAG pipeline — JSON-Mode, CoT, Hybrid Search, Re-Ranking, Cross-Reg Dedup, chunk 1024

Phase 1 (LLM Quality):
- Add format=json to all Ollama payloads (obligation_extractor, control_generator, citation_backfill)
- Add Chain-of-Thought analysis steps to Pass 0a/0b system prompts

Phase 2 (Retrieval Quality):
- Hybrid search via Qdrant Query API with RRF fusion + automatic text index (legal_rag.go)
- Fallback to dense-only search if Query API unavailable
- Cross-encoder re-ranking with BGE Reranker v2 (RERANK_ENABLED=false by default)
- CPU-only PyTorch dependency to keep Docker image small

Phase 3 (Data Layer):
- Cross-regulation dedup pass (threshold 0.95) links controls across regulations
- DedupResult.link_type field distinguishes dedup_merge vs cross_regulation
- Chunk size defaults updated 512/50 → 1024/128 for new ingestions only
- Existing collections and controls are NOT affected

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

breakpilot-compliance-sdk/services/rag-service/config.py

@@ -23,8 +23,11 @@ class Settings(BaseSettings):
     llm_model: str = "qwen2.5:32b"
 
     # Document Processing
-    chunk_size: int = 512
-    chunk_overlap: int = 50
+    # NOTE: Changed from 512/50 to 1024/128 for improved retrieval quality.
+    # Existing collections (ingested with 512/50) are NOT affected —
+    # new settings apply only to new ingestions.
+    chunk_size: int = 1024
+    chunk_overlap: int = 128
 
     # Legal Corpus
     corpus_path: str = "./legal-corpus"