Implement Phase 12 feedback, Phase 13 Semantic Chunker, Phase 13-B Reranker, Bug 5 thinking fix

- Phase 12: FeedbackRepository + td_feedback 테이블, Gradio 👍/👎 이벤트, run_id 추적, LangSmith create_feedback() 연동
- Phase 13: 커스텀 _SemanticSplitter 제거 → langchain_experimental.SemanticChunker 교체, buffer_size/threshold_type 환경변수 적용
- Phase 13-B: RerankService (Cross-Encoder), RetrieverService.search()에 reranker 통합, tools.py as_retriever() → search() 전환
- Bug 5: mlx_chat_model enable_thinking 런타임 오버라이드, agent_service stream_mode=["messages","custom"] 이중 스트림, thinking 토큰 custom 이벤트로 emit
- ROADMAP: LLM 모델명 8B 반영, RAG에 Reranker 추가, 추천 진행 순서 갱신

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

services/rag/rerank_service.py

@@ -0,0 +1,19 @@
+from langchain_core.documents import Document
+
+
+class RerankService:
+    """Cross-Encoder 기반 재순위(Reranker) 서비스."""
+
+    def __init__(self, model_id: str = "cross-encoder/mmarco-mMiniLMv2-L12-H384-v1"):
+        from sentence_transformers import CrossEncoder
+        print(f"Reranker 로딩 중: {model_id}")
+        self._model = CrossEncoder(model_id)
+        print("Reranker 로딩 완료")
+
+    def rerank(self, query: str, docs: list[Document], top_k: int) -> list[Document]:
+        if not docs:
+            return docs
+        pairs = [(query, doc.page_content) for doc in docs]
+        scores = self._model.predict(pairs)
+        ranked = sorted(zip(scores, docs), key=lambda x: x[0], reverse=True)
+        return [doc for _, doc in ranked[:top_k]]