Microsoft Copilot Studio UX - Build 26.06.07-10

This article describes the fixes and improvements to existing functionality that are included in this build.

Visit the Microsoft Copilot Studio Blog to stay up-to-date on the latest news!

Version 26.06.07-10 updates

New features

• End-to-end workflow evaluation enhancements let makers preview latest runs in a new workflow preview page, select real runs in the new evaluation page, start evaluations directly from the workflow details page with a guided setup and prefilled name, choose which test methods to run, and use the Run button to start a flow evaluation.
• Maker Evaluation scoring now includes clearer visuals with score badges, multi-score details, per-metric explanations for General Quality metrics, and optional display of the grader name in results.
• Workflow analytics is now available with an overview page, a date range picker, summary cards for total runs, success rate, and actions performed, and a new consumption metric card in analytics.
• Makers can switch between inline-card and side-panel layouts in the flow designer, with improved stability for large, schema-heavy flows.
• Text and JSON output types now include descriptions to help makers choose the right output format.
• Microsoft Copilot Studio now supports English (New Zealand) (en-NZ) as a user interface language.

Fixes and improvements

• Accessibility improvements across Copilot Studio include correct screen reader announcements for required fields and placeholders, headings no longer exposed from hero and thumbnail card titles, clearer accessible names and keyboard focus management, properly labeled instructions and dialogs, visible and usable controls at small or reflow sizes, and information popovers that fit and scroll at high zoom.
• Layout and responsiveness upgrades improve usability at high zoom and small screens: evaluation results use available space, the Test chat send area and transcript remain visible, side panels and variable pickers fit within the viewport, dialog and content areas scroll properly, analytics uses full browser width, and tooltips and menus are positioned reliably.
• Analytics correctness and navigation are improved with accurate sentiment badges, totals, and percentages, better sessions metrics and filtering, correct “Learn more” links and drilldowns, consistent breadcrumbs, non-clickable preview charts when actions are disabled, and custom metrics visible only for published version 2 agents.
• Navigation, routing, and licensing behaviors are more predictable: deep links resolve environments more reliably, provisioning status is tracked via server polling, SharePoint and OneDrive links open in the browser viewer, classic experiences remain in classic routes, and Teams-only users are consistently routed to the legacy experience without prompts to switch when blocked by licensing.
• Maker Evaluation UI and configuration updates include refreshed results and summary cards, improved Elevate setup with clearer layout and labels, Content Safety grader setup that uses severity levels and clearer content-type labeling, category-specific thresholds that persist, simplified navigation and breadcrumbs, and appropriate hiding of workflow headers and configuration fields where not applicable.
• Continued improvements to performance and reliability.

Security

• No changes in this release.

Return to the Microsoft Copilot Studio Release Notes.