MC1530748 Low
Copilot Studio - Tool accuracy grader
Summary AI-generated
The tool accuracy grader in Copilot Studio is now generally available, enabling evaluation of agent tool selection and execution for improved reliability.
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- MC1530742 Copilot Studio - Task completion grader
- MC1531932 Copilot Studio - Custom Graders Library and Ready-to-Use Microsoft Graders
- MC1261573 Microsoft Copilot Studio – Tool & Topic Invocation Grader
- MC1531924 Copilot Studio - Generated custom grader
- MC1531903 Copilot Studio - Evaluation for automation flows
- MC1218332 Power Platform - Track, manage and scale Copilot adoption in PPAC new functionality
Original post from Microsoft
We are announcing the tool accuracy grader in Copilot Studio. This feature reached general availability on October 15, 2026.
How does this affect me?
This feature enables you to evaluate whether your agents selected the correct tools or actions and used them with the expected inputs for a given user request. This helps you to identify execution issues earlier, reduce manual trace inspection, and increase confidence that agents are performing actions correctly and reliably.
What action do I need to take?
This message is for awareness, and no action is required.
How does this affect me?
This feature enables you to evaluate whether your agents selected the correct tools or actions and used them with the expected inputs for a given user request. This helps you to identify execution issues earlier, reduce manual trace inspection, and increase confidence that agents are performing actions correctly and reliably.
What action do I need to take?
This message is for awareness, and no action is required.