Announcements
I would like to report a rather strange bug that is causing me headaches.
I have a system of connected agents that interact with each other to solve a task, governed by an orchestrator agent that uses the generative orchestration feature.
However, generative orchestration often ‘bugs out’ and stops working when I make minor changes to any part of the agent's configuration (for example, deleting a comma or space in the instructions or disabling a topic).
Is there a known way to make orchestration work robustly? Thanks in advance.
Here is an example of before and after disabling a topic that is irrelevant to the task.
BEFORE (calls the agent)
AFTER (ignores the agent)
Agent orchestration is still in preview, and this kind of inconsistent behavior can happen because the feature is not fully stabilized yet. Even small updates—like removing a space, editing instructions, or disabling a topic—can change how the orchestrator interprets the flow. When that happens, it may stop calling the connected agent and try to complete the request by itself. This is a common issue reported by users working with multi-agent setups.
Since the feature is still evolving, the best next step is to report the bug directly in Copilot Studio. This helps the product team review the behavior, reproduce the issue, and include a fix in future updates.
Under review
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.
Congratulations to our 2026 Super Users!
Congratulations to our 2025 community superstars!
These are the community rock stars!
Stay up to date on forum activity by subscribing.
Valantis 713
chiaraalina 170 Super User 2026 Season 1
deepakmehta13a 117