Implement Phase 4~14: LangGraph Agent, RAG pipeline, Gradio Web UI, voice interface

- Upgrade LLM to Qwen3-14B-4bit with Thinking mode (MlxChatModel as LangChain BaseChatModel)
- Add LangGraph ReAct agent with tool calling loop (search_documents, web_search, get_current_date, remember/recall_user_info)
- Add RAG pipeline: BAAI/bge-m3 embeddings + Qdrant vector store + semantic chunking (SemanticSplitter via cosine similarity)
- Replace fixed-size RecursiveCharacterTextSplitter with meaning-based SemanticSplitter (numpy only, no extra deps)
- Add Gradio Web UI (app.py): chat, document ingestion, document management tabs
- Add multi-user support (user_id isolation in DB + per-user agent cache + dropdown selector)
- Add conversation history restore from MySQL on agent init (Phase 11)
- Add UserProfileRepository for persistent user profile (remember/recall tools)
- Add thread-local DB connections to fix pymysql thread-safety with LangGraph ToolNode
- Add Phase 14 voice interface: Whisper STT (microphone → text) + macOS TTS (say -v Yuna)
- Enforce search_documents-first policy in system prompt and tool descriptions
- Update ROADMAP2.md: Phase 14 완료, Phase 13 청킹 부분 완료

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

container.py

@@ -2,14 +2,20 @@ from dependency_injector import containers, providers
 
 from config import Config
 from services.model.mlx_model import MlxModelService
+from services.model.mlx_chat_model import MlxChatModel
 from services.chat.history_service import HistoryService
 from services.chat.chat_service import ChatService
 from services.chat.compact_service import CompactService
 from services.db.mysql_service import DatabaseService
 from services.db.conversation_repository import ConversationRepository
+from services.db.user_profile_repository import UserProfileRepository
 from services.ui.cli_service import CliUiService
 from services.events.event_bus import EventBus
 from services.events.handlers import StreamTokenHandler, StreamEndHandler
+from langchain_huggingface import HuggingFaceEmbeddings
+from services.rag.ingestion_service import IngestionService
+from services.rag.retriever_service import RetrieverService
+from services.agent.agent_service import AgentService
 
 
 class Container(containers.DeclarativeContainer):
@@ -22,6 +28,14 @@ class Container(containers.DeclarativeContainer):
         model_id=providers.Callable(lambda c: c.model_id, config),
     )
 
+    # LangGraph 에이전트용 BaseChatModel (Phase 1)
+    chat_model = providers.Singleton(
+        MlxChatModel,
+        model_id=providers.Callable(lambda c: c.model_id, config),
+        max_tokens=providers.Callable(lambda c: c.max_tokens, config),
+        enable_thinking=providers.Callable(lambda c: c.enable_thinking, config),
+    )
+
     compact_service = providers.Singleton(
         CompactService,
         model=model_service,
@@ -41,6 +55,11 @@ class Container(containers.DeclarativeContainer):
         db=db_service,
     )
 
+    user_profile_repository = providers.Singleton(
+        UserProfileRepository,
+        db=db_service,
+    )
+
     history_service = providers.Factory(
         HistoryService,
         system_prompt=providers.Callable(lambda c: c.system_prompt, config),
@@ -62,3 +81,42 @@ class Container(containers.DeclarativeContainer):
 
     stream_token_handler = providers.Singleton(StreamTokenHandler)
     stream_end_handler = providers.Singleton(StreamEndHandler)
+
+    # Phase 2 — RAG 파이프라인
+    embeddings = providers.Singleton(
+        HuggingFaceEmbeddings,
+        model_name=providers.Callable(lambda c: c.embedding_model_id, config),
+        model_kwargs=providers.Callable(lambda c: {"device": c.embedding_device}, config),
+    )
+
+    ingestion_service = providers.Singleton(
+        IngestionService,
+        embeddings=embeddings,
+        qdrant_url=providers.Callable(lambda c: c.qdrant_url, config),
+        collection_name=providers.Callable(lambda c: c.qdrant_collection, config),
+        breakpoint_threshold_type=providers.Callable(
+            lambda c: c.semantic_breakpoint_threshold_type, config
+        ),
+    )
+
+    retriever_service = providers.Singleton(
+        RetrieverService,
+        embeddings=embeddings,
+        qdrant_url=providers.Callable(lambda c: c.qdrant_url, config),
+        collection_name=providers.Callable(lambda c: c.qdrant_collection, config),
+        top_k=providers.Callable(lambda c: c.rag_top_k, config),
+    )
+
+    # Phase 3 — LangGraph Agent
+    agent_service = providers.Singleton(
+        AgentService,
+        chat_model=chat_model,
+        retriever_service=retriever_service,
+        system_prompt=providers.Callable(lambda c: c.system_prompt, config),
+        rag_verbose=providers.Callable(lambda c: c.rag_verbose, config),
+        rag_show_sources=providers.Callable(lambda c: c.rag_show_sources, config),
+        langgraph_verbose=providers.Callable(lambda c: c.langgraph_verbose, config),
+        think_verbose=providers.Callable(lambda c: c.think_verbose, config),
+        user_profile_repository=user_profile_repository,
+        conversation_repository=conversation_repository,
+    )