← All AI news
GitHubAgents

Copilot code review adds API settings and clearer attribution

Administrators can manage review settings through APIs and inspect reported skills, MCP servers and instructions behind comments.

THE UPDATE

What changed

GitHub announced on November 6 that Copilot code review settings can now be read and managed through REST and GraphQL APIs from enterprise policy to repository configuration. New attribution fields can identify Copilot reviews, review effort and the skills, MCP servers or custom instructions Copilot reports as sources for a comment.

One useful next step

Choose one non-critical repository and export its current review settings before changing anything. Run a small reviewed pull-request sample, retain human approval, and record false positives, missed issues and correction time by review configuration.

Plan a bounded review pilot ↗

Before you dive in

Availability depends on an eligible Copilot plan, role and token permissions. Enterprise-managed settings cannot be overridden below that level. GitHub Apps and fine-grained personal access tokens do not support the enterprise and organization settings endpoints described in the announcement.

GO TO THE ORIGINAL

Source & announcement date

GitHub ·

GitHub Changelog: Copilot code review API management and review overview updates ↗ (opens in a new tab)

This brief summarizes the linked announcement. “The AI Superpower take” and the next step are our editorial interpretation, not claims from the source.

KEEP YOUR CURIOSITY GOING

A few more worth knowing.

All AI news ↗