Switched over to GPT-5 Orchestration and my Agent sometimes brings up tool call data in the chat.
Does anyone know why this is happening and how to turn it off? This is happening in both Copilot Studio testing and live model.
This is happening for any topics. It also doesn't happen for every message, just sometimes.
We have placed this in the instructions as well, and it does not seem to be working
"When responding to the user:
- Only output the user-facing text.
- Do not include internal metadata such as explanation_of_tool_call, intent completion signals, or JSON objects.
- If the response is empty or invalid, send a fallback message like "Sorry, I couldn't find that information."
- Do not send message about "explanation_of_tool_call""
Thank you for your reply! To ensure a great experience for everyone, your content is awaiting approval by our Community Managers. Please check back later.
29
Share
Report
All responses (
Answers (
