Copilot code review: Analysis depth and efficiency updates

GitHub's Copilot code review now uses CLI-based file exploration tools, reducing costs by 20% while maintaining quality. Medium analysis depth public preview adds organization-level default settings and attribution labels for review runs.

Copilot code review: Analysis depth and efficiency updates Copilot code review now uses the built-in file exploration tools available in the Copilot CLI and SDK, significantly improving review cost efficiency with no change to your existing workflow. If you’re in the Medium analysis depth public preview, you’ll also see some new updates centered around configurability and visibility of review depth. Configure Medium analysis depth for your organization configure-medium-analysis-depth-for-your-organization If you’re opted into the Medium review effort level public preview https://github.blog/changelog/2026-06-02-shape-copilot-code-review-around-your-team/ , you now get two updates: : Copilot code review now labels medium analysis depth runs in its pull request overview comment so you can quickly confirm which level generated the review. Medium attribution in the pull request overview comment- Organization-level : Organizations can now set a default review level for unconfigured repositories. Repositories under an organization that has configured the default review level will continue to be able to override that default setting if desired. default level setting Behind the scenes: CLI-based file tools in Copilot code review behind-the-scenes-cli-based-file-tools-in-copilot-code-review Copilot code review now uses the grep , rg , glob and view tools from the Copilot CLI and SDK for exploring the source code in its review path. These replace custom tools previously used for file exploration. This capability, along with careful tuning of instructions behind the scenes, has resulted in a more focused review where Copilot finds the code that matters, quickly. These efficiency gains have reduced Copilot code review costs by about 20% while maintaining the same standard of review quality. This has been observed in both offline and online evaluation.