1

app/server/ChatbotAgent/agent_server.py

@@ -1,6 +1,7 @@
 import asyncio
 import json
 import logging
+import os
 import uuid
 from typing import List, Dict
 
@@ -33,7 +34,7 @@ class LCAgent(ls.LitAPI):
             key_prefix=settings.REDIS_HISTORY_KEY_PREFIX
         )
         self.vector_db = VectorDatabase(
-            vector_db_dir=settings.VECTOR_DB_DIR,
+            vector_db_dir=os.getenv('VECTOR_DB_DIR', '/app/app/db'),
             collection_name=settings.COLLECTION_NAME,
             embedding_model_name=settings.EMBEDDING_MODEL_NAME
         )
@@ -44,6 +45,16 @@ class LCAgent(ls.LitAPI):
         }
 
     async def decode_request(self, request: AgentRequestModel):
+        """
+        [自动生成文档]
+        该方法将 HTTP 请求体解码为模型所需的输入格式。
+
+        Args:
+            request: 包含 prompt, history, temperature 的 ChatBotRequest 对象。
+
+        Returns:
+            prompt, history, temperature
+        """
         logger.info(f"request: {request.model_dump()}")
         return request
 
@@ -81,10 +92,12 @@ class LCAgent(ls.LitAPI):
         history_messages = self.redis.get_history(user_id)
         input_message = "\n".join([f"{msg.role.value}: {msg.content}" for msg in history_messages])
         # 临时调用 LLM 或使用本地逻辑生成总结
-        summary = await self.llm.generate_response(history=[Message(role=Role.USER, content=input_message)], system_prompt=SUMMARY_PROMPT)
+        summary = await self.llm.generate_response(history=[Message(role=Role.USER, content=input_message)],
+                                                   system_prompt=SUMMARY_PROMPT)
         return summary
 
-    async def recommend_outfit(self, request_summary: str, stylist_name: str, start_outfit=None, num_outfits: int = 1, user_id: str = "test"):
+    async def recommend_outfit(self, request_summary: str, stylist_name: str, start_outfit=None, num_outfits: int = 1,
+                               user_id: str = "test"):
         """
         基于用户的对话历史和需求，推荐一套搭配。