MC1532071 Low
Copilot Studio - Agent status evaluation warnings announcement
Summary AI-generated
Agent status evaluation warnings are now generally available in Copilot Studio to help makers identify when agents need reevaluation based on their history and model changes.
Suggested actions AI-generated
- Select 'Evaluate agent' when warning appears to run evaluation
Written by Azure OpenAI (gpt-4.1) from the text of the post below. It can be incomplete or wrong; the original post is authoritative.
Similar posts
Search for more like this- MC1402296 Microsoft Copilot Studio – See a unified view of errors, warnings, and governance notifications
- MC1518085 Microsoft Copilot Studio - Improvements to agent Evaluations experience
- MC1194567 Power Platform Monitor - Announcing Copilot Studio agent health metrics and alerts in Public Preview
- MC1403393 Microsoft Copilot Studio - Evaluate AI action outcome in workflows
- MC1190188 Microsoft Copilot Studio – Increase sense of security for makers inside Copilot Studio
- MC1396327 Microsoft Copilot Studio - Evaluate AI action outcome in workflows with confidence
Original post from Microsoft
Agent status evaluation warnings are now generally available in Copilot Studio as of October 15, 2026.
How does this affect me?
Makers will automatically see warnings in the Review panel when an agent has never been evaluated, or if its model has changed since the most recent evaluation. These warnings help makers decide when to reevaluate an agent before continuing development or deployment. They do not block publishing.
What action do I need to take?
No admin action or additional configuration is required. Warnings appear automatically based on the agent’s evaluation history and model changes. When a warning appears, select Evaluate agent to open the Evaluation page and run an evaluation with the agent’s current model.
How does this affect me?
Makers will automatically see warnings in the Review panel when an agent has never been evaluated, or if its model has changed since the most recent evaluation. These warnings help makers decide when to reevaluate an agent before continuing development or deployment. They do not block publishing.
What action do I need to take?
No admin action or additional configuration is required. Warnings appear automatically based on the agent’s evaluation history and model changes. When a warning appears, select Evaluate agent to open the Evaluation page and run an evaluation with the agent’s current model.