MC1531932 High
Copilot Studio - Custom Graders Library and Ready-to-Use Microsoft Graders
Summary AI-generated
Copilot Studio introduces centralized custom graders library, enabling standardized agent evaluation. General availability starts November 15, 2026.
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- MC1472547 Microsoft Copilot Studio - Custom Grader Library
- MC1531924 Copilot Studio - Generated custom grader
- MC1530742 Copilot Studio - Task completion grader
- MC1261573 Microsoft Copilot Studio – Tool & Topic Invocation Grader
- MC1530748 Copilot Studio - Tool accuracy grader
- MC1518233 Copilot Studio - Enhanced support for Copilot Chat agents
Original post from Microsoft
We are announcing the ability to create, save, manager, discover, and reuse custom grades from a central library in Copilot Studio. This feature will reach general availability on November 15, 2026.
How does this affect me?
This feature enables you to standardize how agent quality is measured by building and reusing a shared set of evaluation criteria across agents and test sets. Makers can create custom graders or utilize out-of-the-box graders provided by Microsoft. This reduces duplication, improves consistency across teams and agents, and makes it easier to scale evaluation practices across the organization.
What action do I need to take?
This message is for awareness, and no action is require
How does this affect me?
This feature enables you to standardize how agent quality is measured by building and reusing a shared set of evaluation criteria across agents and test sets. Makers can create custom graders or utilize out-of-the-box graders provided by Microsoft. This reduces duplication, improves consistency across teams and agents, and makes it easier to scale evaluation practices across the organization.
What action do I need to take?
This message is for awareness, and no action is require