The build, deploy and infrastructure-as-code tools teams live in.
GitHub has launched Code Quality in public preview, enabling developers to identify and resolve code maintainability and reliability issues directly within pull requests using CodeQL analysis and Copilot-powered suggestions.
GitHub's Copilot code review now combines large language model analysis with deterministic tools like CodeQL and ESLint, while enabling direct handoff of fixes to the Copilot coding agent.
Developers can now route Linear issues directly to GitHub's Copilot coding agent, which will autonomously tackle the work in a temporary environment and submit draft pull requests for review.
Developers can now invoke Copilot's autonomous coding agent to modify pull requests in progress by simply tagging it in a comment, with changes proposed in a separate PR for review.
Developers can now route Copilot's autonomous agent through self-hosted GitHub Actions runners, enabling access to internal resources and private packages.
Enterprise Cloud administrators can now manage Copilot Business seat assignments directly through their enterprise account, decoupling license allocation from organizational membership requirements.
Administrators can now manage which Model Context Protocol servers developers access through a new registry system and allowlist controls in preview builds of popular IDEs.
GitHub's autonomous Copilot coding agent can now be delegated tasks directly from JetBrains IDEs in public preview, handling background work while developers focus elsewhere.
