From a437a1e8c351a8ae6ab287a25cfb4e95ea67c48b Mon Sep 17 00:00:00 2001 From: AI Christianson Date: Mon, 10 Mar 2025 13:46:08 -0400 Subject: [PATCH] prompt improvements --- ra_aid/prompts/reasoning_assist_prompt.py | 1 + 1 file changed, 1 insertion(+) diff --git a/ra_aid/prompts/reasoning_assist_prompt.py b/ra_aid/prompts/reasoning_assist_prompt.py index 39f2e92..075dad5 100644 --- a/ra_aid/prompts/reasoning_assist_prompt.py +++ b/ra_aid/prompts/reasoning_assist_prompt.py @@ -88,6 +88,7 @@ REMEMBER, IT IS *IMPERATIVE* TO RECORD KEY INFO SUCH AS BUILD/TEST COMMANDS, ETC WE DO NOT WANT TO EMIT REDUNDANT KEY FACTS, SNIPPETS, ETC. WE DO NOT WANT TO EXCESSIVELY EMIT TINY KEY SNIPPETS --THEY SHOULD BE "paragraphs" OF CODE TYPICALLY. IF THERE IS COMPLEX LOGIC, COMPILATION ERRORS, DEBUGGING, THE AGENT SHOULD USE ask_expert. +IF ANYTHING AT ALL GOES WRONG, CALL ask_expert. Given the available information, tools, and base task, write a couple paragraphs about how an agentic system might use the available tools to implement the given task definition. The agent will be writing code and making changes at this point.