chore: finalize backend feature scope

app/services/session_service.py

@@ -15,7 +15,6 @@ from app.models.training import SessionSubmission, TrainingSession
 from app.repositories.case_repository import CaseRepository
 from app.repositories.session_repository import SessionRepository
 from app.schemas.session import (
-    ChatResponse,
     CreateSessionRequest,
     CreateSessionResponse,
     SessionStatusResponse,
@@ -80,34 +79,6 @@ class SessionService:
             patient_config=patient_config,
         )
 
-    async def chat(self, ctx: UserContext, session_id: int, message: str) -> ChatResponse:
-        """问诊对话：拼接病例上下文、短期记忆和用户输入后调用 Patient Agent。"""
-        session = self._get_session(session_id, ctx.user_id)
-        if session.status != "inquiry":
-            raise AppError("SESSION_STATUS_INVALID", "chat is only allowed in inquiry status", 400)
-        case = self.case_repo.get_active_case(session.case_id)
-        if not case:
-            raise AppError("CASE_NOT_FOUND", "case not found or inactive", 404)
-
-        start = time.perf_counter()
-        memory_messages = runtime_memory.get_messages(session.memory_key)
-        runtime_memory.add_message(session.memory_key or "", "doctor", message)
-        try:
-            response = await asyncio.wait_for(
-                self.orchestrator.patient_reply(session, case, memory_messages, message),
-                timeout=settings.llm_chat_timeout_seconds,
-            )
-        except TimeoutError as exc:
-            raise AppError("LLM_CALL_TIMEOUT", "AI 病人回复超时，请稍后重试或切换为普通问诊", 504) from exc
-        runtime_memory.add_message(session.memory_key or "", "patient", response.content)
-        self.audit.log(ctx, "session.chat", "training_session", str(session.id), session.id)
-        return ChatResponse(
-            reply=response.content,
-            latency_ms=response.latency_ms or int((time.perf_counter() - start) * 1000),
-            model=response.model,
-            fallback_used=response.model.startswith("mock-fallback"),
-        )
-
     async def stream_chat(self, ctx: UserContext, session_id: int, message: str) -> AsyncIterator[str]:
         """流式问诊：返回 SSE 格式的 AI 病人回复。"""
         session = self._get_session(session_id, ctx.user_id)
@@ -214,7 +185,7 @@ class SessionService:
         return f"event: error\ndata: {payload}\n\n"
 
     async def generate_hints(self, ctx: UserContext, session_id: int, payload: HintRequest) -> HintResponse:
-        """新手提示：基于当前会话上下文、已申请检查和病例信息生成提醒。"""
+        """练习提示：基于当前会话上下文、已申请检查和病例信息生成提醒。"""
         session = self._get_session(session_id, ctx.user_id)
         if session.mode != "practice":
             raise AppError("SESSION_STATUS_INVALID", "hints are only available in practice mode", 400)