From c097910f24cdc38154145b14ab62817e31bc0f46 Mon Sep 17 00:00:00 2001 From: Michael Date: Tue, 30 Jun 2026 22:07:34 +0200 Subject: [PATCH] feat: make issue closing rules stricter in CoordinatorAgent system prompt --- core/prompts.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/core/prompts.py b/core/prompts.py index 8a4f39a..7c43038 100644 --- a/core/prompts.py +++ b/core/prompts.py @@ -20,9 +20,10 @@ Based on the conversation state, you must choose and call exactly one of the fol - Or the agent answered, but the human replied with follow-up questions/clarifications. Provide a clear, helpful response. -4. `close_issue`: Choose this if the AI agent previously answered a question and the human has replied confirming they are satisfied or giving approval to close (e.g., "thanks", "looks good", "close it"). +4. `close_issue`: Choose this ONLY if the AI agent previously answered a question AND the human has explicitly replied with a message confirming they are fully satisfied or explicitly instructing the agent to close the issue (e.g., "thanks, this answers my question", "looks good, you can close this", "close it"). If the human's response is a follow-up question, is ambiguous, or does not explicitly approve closing, you must NOT call this tool (call `answer_question` or `take_no_action` instead). Provide a polite closing comment. + 5. `take_no_action`: Choose this if the issue is already resolved, or if we cannot proceed for another reason. CRITICAL INSTRUCTIONS: -- 2.52.0