Update app/core/response_handlers.py

app/core/response_handlers.py

@@ -4,6 +4,7 @@ Response handlers for streaming and non-streaming responses
 
 import json
 import time
+import re
 from typing import Generator, Optional
 import requests
 from fastapi import HTTPException
@@ -459,6 +460,8 @@ class NonStreamResponseHandler(ResponseHandler):
                         
                         if upstream_data.data.phase == "thinking":
                             content = transform_thinking_content(content)
+                            # 移除单层引用，但保留多层引用
+                            content = re.sub(r'^\s*>\s*(?!>)', '', content, flags=re.MULTILINE)
                             
                             # 处理思考内容的分块格式
                             if not self.in_thinking_phase: