Agentic Architecture Now Generally Available

GitHub Copilot code review now runs on an agentic tool-calling architecture and is generally available for all users with Copilot Pro, Copilot Pro+, Copilot Business, and Copilot Enterprise. This upgrade builds on the public preview announced in October 2025.

Key Improvements

The agentic approach enables Copilot to gather broader repository context as needed, including:

• Relevant code patterns and dependencies across your codebase
• Directory structure and architectural decisions to understand the broader system design
• Code references to identify how changes integrate with existing implementations

This context-aware analysis delivers three main benefits:

• Higher-quality findings that prioritize correctness and architectural integrity
• Lower noise by filtering out trivial or overly broad comments and surfacing only meaningful feedback
• More actionable guidance that helps developers resolve issues quickly and confidently

Deployment and Configuration

Agentic Copilot code review runs on GitHub Actions. Most users require no setup, as the feature defaults to GitHub-hosted runners. However, if you have opted out of GitHub-hosted runners, you'll need to complete a one-time self-hosted runners setup to receive agentic code reviews on pull requests. Alternatively, you can use GitHub-hosted larger runners if preferred.

Getting Started

Users new to Copilot code review can get started immediately by consulting the official documentation. For details on the technical architecture powering these improvements, see GitHub's blog post on the topic.