diff --git a/metagpt/reflect/__init__.py b/metagpt/reflect/__init__.py
index 6cc738dbe..6acb42c9e 100644
--- a/metagpt/reflect/__init__.py
+++ b/metagpt/reflect/__init__.py
@@ -3,8 +3,10 @@
 # @Desc   : reflection module
 
 from metagpt.reflect import agent_reflect
-
+from metagpt.reflect import ga_prompt_generator
 __all__ = [
     "agent_reflect",
     "LongTermMemory",
+    "ga_po"
+    "ga_prompt_generator"
 ]
diff --git a/metagpt/actions/ga_action_base.py b/metagpt/reflect/ga_prompt_generator.py
similarity index 95%
rename from metagpt/actions/ga_action_base.py
rename to metagpt/reflect/ga_prompt_generator.py
index 01bdca6b6..6b7bdbad7 100644
--- a/metagpt/actions/ga_action_base.py
+++ b/metagpt/reflect/ga_prompt_generator.py
@@ -12,7 +12,7 @@ from metagpt.llm import DEFAULT_LLM
 # 特殊指令加入Prompt生成
 
 
-def final_response(prompt, special_instruction, example_output=None):
+async def final_response(prompt, special_instruction, example_output=None):
     """
     通过将特殊指令加入Prompt生成最终的响应。
 
@@ -30,7 +30,7 @@ def final_response(prompt, special_instruction, example_output=None):
     if example_output:
         prompt += "Example output json:\n"
         prompt += '{"output": "' + str(example_output) + '"}'
-    return DEFAULT_LLM.ask(prompt)
+    return await DEFAULT_LLM.aask(prompt)
 
 # prompt填充模板
 
diff --git a/metagpt/reflect/reflect.py b/metagpt/reflect/reflect.py
index 84ac5c604..218df4fea 100644
--- a/metagpt/reflect/reflect.py
+++ b/metagpt/reflect/reflect.py
@@ -2,25 +2,35 @@
 # -*- coding: utf-8 -*-
 # @Desc   : base class of reflection
 
+import asyncio
 import json
 from metagpt.logs import logger
 import time
-from metagpt.actions.ga_action_base import final_response
+from ga_prompt_generator import final_response
 '''
 等待Agent和memory更新，保留相关引用但可以忽略。
 '''
 from metagpt.memory.ga_memory_storage import AgentMemory, MemoryBasic
 
+import json
+import time
 
-def agent_reflect(agent):
-    A = generate_focus_point(agent.memories_list)
+
+async def agent_reflect(agent):
+    """
+    代理反思函数：生成关注点并生成洞察和证据
+
+    """
+    A = await generate_focus_point(agent.memories_list)
 
     for i in A:
-        B = generate_insights_and_evidence(
-            agent, agent.memories_list, question=i)
+        B = await generate_insights_and_evidence(agent, agent.memories_list, question=i)
 
 
-def generate_focus_point(memories_list, n=3):
+async def generate_focus_point(memories_list: list[MemoryBasic], n=3):
+    """
+    生成关注点函数：根据记忆列表生成关注点
+    """
     wait_sorted_mem = [[i.accessed_time, i] for i in memories_list]
     sorted_memories = sorted(wait_sorted_mem, key=lambda x: x[0])
     memorys = [i for created, i in sorted_memories]
@@ -32,19 +42,22 @@ def generate_focus_point(memories_list, n=3):
     Given only the information above, what are {num_question} most salient high-level questions we can answer about the subjects grounded in the statements?
     '''
     example_output = '["What should Jane do for lunch", "Does Jane like strawberry", "Who is Jane"]'
-    out = final_response(prompt.format(statements=statements, num_question=n),
-                         "Output must be a list of str.", example_output)
+    out = await final_response(prompt.format(statements=statements, num_question=n),
+                               "Output must be a list of str.", example_output)
     try:
         poi_dict = json.loads(out)
-        return (poi_dict['output'])
+        return poi_dict['output']
     except ValueError:
         print(out)
         logger.error('无法返回正常结果')
         return out
 
 
-def generate_insights_and_evidence(agent, memories_list, question, n=5):
-    agent_retrive(agent, question, 20, 10)
+async def generate_insights_and_evidence(agent: Agent, memories_list: list[MemoryBasic], question: str, n=5):
+    """
+    生成洞察和证据函数：根据问题生成洞察和证据
+    """
+    await agent_retrieve(agent, question, 20, 10)
     statements = ""
     for count, mem in enumerate(memories_list):
         statements += f'{str(count)}. {mem.description}\n'
@@ -53,7 +66,7 @@ def generate_insights_and_evidence(agent, memories_list, question, n=5):
     {statements}
 
     What {n} high-level insights can you infer from the above statements?
-    You should return a list of list[str,list] . The first element is the insight you have found.The second element is the 
+    You should return a list of list[str,list]. The first element is the insight you have found. The second element is the 
     '''
 
     ret = final_response(prompt.format(
@@ -63,27 +76,22 @@ def generate_insights_and_evidence(agent, memories_list, question, n=5):
         for insight, index in insight_list:
             agent.memory_list.append(Memory_basic(
                 time.time(), None, insight, None, None))
-        return (insight_list)
+        return insight_list
     except:
-        Logger.error('我们无法获得想要的返回。')
+        logger.error('我们无法获得想要的返回。')
         return ret
 
 
-if __name__ == "__main__":
+""" if __name__ == "__main__":
     # 例子，构建John Agent，实现retrive
     John_iss = "John Lin is a pharmacy shopkeeper at the Willow Market and Pharmacy who loves to help people. He is always looking for ways to make the process of getting medication easier for his customers; John Lin is living with his wife, Mei Lin, who is a college professor, and son, Eddy Lin, who is a student studying music theory; John Lin loves his family very much; John Lin has known the old couple next-door, Sam Moore and Jennifer Moore, for a few years; John Lin thinks Sam Moore is a kind and nice man; John Lin knows his neighbor, Yuriko Yamamoto, well; John Lin knows of his neighbors, Tamara Taylor and Carmen Ortiz, but has not met them before; John Lin and Tom Moreno are colleagues at The Willows Market and Pharmacy; John Lin and Tom Moreno are friends and like to discuss local politics together; John Lin knows the Moreno family somewhat well — the husband Tom Moreno and the wife Jane Moreno."
     John = AgentMemory(
         "John", John_iss, memory_path="agent_memories/John_memory.json")
 
     # John的相关信息：{'Had a friendly chat with Yuriko about her garden.': 2.4992317730827667, 'Helped Mrs. Moore carry groceries into her house.': 1.957656720441911, 'Discussed local politics with Tom Moreno.': 1.9458268038234035}
-    A = generate_focus_point(John.memories_list)
-
-    for i in A:
-        B = generate_insights_and_evidence(
-            John, John.memories_list, question=A[0])
-        print(type(B))
-        print(B)
+    asyncio.run(agent_reflect(John))
     '''
     这里是输出,list形式，返回给记忆。
     [['The pharmacy is a friendly and helpful community.', [0, 2, 9, 12]], ['The pharmacy is a place where people come for more than just medication.', [3, 5, 13, 14]], ['The pharmacy is a place where people come for advice and conversation.', [0, 2, 6, 9, 12]], ['The pharmacy is a place where people come for assistance with daily tasks.', [3, 5, 13, 14]], ['The pharmacy is a place where people come for political discussions.', [1]]]
     '''
+ """