update summary op

experiencemaker/op/base_op.py

@@ -33,7 +33,8 @@ class BaseOp(PromptMixin, ABC):
         if self.op_config.prompt_file_path:
             prompt_file_path = self.op_config.prompt_file_path
         else:
-            prompt_file_path = Path(__file__).parent / f"{self.simple_name}_prompt.yaml"
+            prompt_name = self.simple_name.replace("op", "").strip("_") + "_prompt.yaml"
+            prompt_file_path = Path(__file__).parent / prompt_name
 
         # Load custom prompts from prompt file
         self.load_prompt_by_file(prompt_file_path=prompt_file_path)

experiencemaker/op/summarizer/simple_summarizer_prompt.yaml

@@ -1,22 +0,0 @@
-summary_prompt: |
-  # Role
-  You are a helpful assistant named BeyondAgent.
-
-  # User Question
-  {query}
-
-  # Execution Process
-  {execution_process}
-
-  # Answer
-  {answer}
-
-  # Task
-  Reflect on the strengths and weaknesses of the **Execution Process** and the **Answer** based on the **User Question**.  
-  Finally, summarize generalized experience from handling such problems to accumulate experience for future similar tasks.  
-  The experience should be broadly applicable, such as how to use tools effectively or approaches to solving certain types of problems.  
-  Also, specify the conditions or scenarios in which these experience are applicable.
-  
-  # Output Format
-  <condition> Output the scenarios or conditions in which applying this experience would be particularly effective... </condition>
-  <experience> Output generalized experience, concise content is required... </experience>

experiencemaker/op/summarizer/simple_summary_prompt.yaml

@@ -0,0 +1,41 @@
+summary_prompt: |
+  # Role
+  You are a helpful assistant.
+
+  # User Question
+  {query}
+
+  # Execution Process
+  {execution_process}
+
+  # Answer
+  {answer}
+
+  # Task
+  Reflect on what went well and what did not go well in the execution path based on the user's question.
+  The summarized experiences should be generalizable, providing assistance in solving similar types of problems in the future. 
+  They can include good suggestions or highlight areas to avoid and potential pitfalls.
+  The type of experience can be plain text or a piece of code that solves a specific problem.
+  If there are no experiences to summarize, output an empty list []. 
+  For each experience, first state the scenario when it is applicable (when to use), then provide the experience itself, with a maximum of three experiences summarized.
+  
+  # Output Format
+  {summary_example}
+
+summary_example: |
+  ``` json
+  [
+    {
+      "when_to_use": "...",
+      "experience": "..."
+    },
+    {
+      "when_to_use": "...",
+      "experience": "..."
+    },
+    {
+      "when_to_use": "...",
+      "experience": "..."
+    }
+  ]
+  ```