megachangelog
Announcement

Copilot code review: Agent skills and MCP now generally available

Copilot code review support for agent skills and Model Context Protocol (MCP) servers is now generally available for all Copilot Pro, Pro+, Business, and Enterprise users, expanding code review capabilities.

Copilot code review support for agent skills and MCP servers is now generally available for all Copilot Pro, Pro+, Business, and Enterprise users. Previously announced in public preview, these capabilities let you bring your team’s tools, standards, and external context directly into every code review.

What’s included

  • Agent skills let Copilot code review invoke your team’s internal tools and coding standards during a review. Add a SKILL.md file under a skill subdirectory in .github/skills to extend Copilot’s analysis with context and instructions specific to your repository or organization.
  • MCP server connections pull context from third-party platforms your team already uses (e.g., issue trackers, documentation systems, service catalogs) directly into the review.
    • All MCP tool calls performed by Copilot code review will be limited to read-only.
    • Any MCP configurations you’ve already set up for Copilot cloud agent automatically apply to Copilot code review. Note that the GitHub and Playwright MCP will be turned on by default.

What’s new

  • Attribution on skills and MCP comments: Copilot code review now indicates when a comment was generated using agent skills or MCP context, so you can see your skills and MCP servers in action.

Getting started

If you configured agent skills or MCP servers during the public preview, no changes are needed. Your existing setup continues to work now that these features are generally available.

For new users:

  • MCP servers: Add your MCP configuration under repository settings → Copilot → MCP servers. Store authentication tokens under repository settings → Secrets and variables → Agents. See example MCP configurations to get started.
  • Agent skills: Under .github/skills, create a skill-specific directory, then add a SKILL.md file to that directory with the context and instructions you want Copilot code review to use. For more details, see the agent skills documentation.

The post Copilot code review: Agent skills and MCP now generally available appeared first on The GitHub Blog.

copilotcode-reviewaimcpagents

Source: original entry ↗

More from GitHub

Follow GitHub to get its new changes in your feed and email digest.

Announcement

Claude Haiku 5.5 now available in GitHub Copilot

Claude Haiku 5.5, Anthropic's lightweight model, is now generally available in GitHub Copilot for fast, high-volume tasks like subagents, quick edits, and terminal operations.

copilotaimodelsannouncement
Feature

Purpose-built model for leaked secret detection

GitHub introduced a new purpose-built model for detecting leaked secrets that provides context-aware detection to keep pace with modern software development practices, including AI-assisted coding.

securitysecret-detectionaiprotection
Announcement

Local sandboxing for GitHub Copilot now generally available

GitHub Copilot's local sandboxing feature is now generally available across GitHub Copilot CLI, the GitHub Copilot app, and VS Code sessions using Agent Host, providing developers with a secure environment for testing and development.

copilotaisecuritysandboxvscode
See all GitHub changes →