Back to rankings

GitHub Copilot: Now Available in JetBrains Air

GitHub CopilotView original changelog

GitHub made GitHub Copilot available inside JetBrains Air through the Agent Client Protocol (ACP). Developers can pick Copilot in Air's agent picker, run several sessions in parallel and review changes next to their code. The integration also supports cloud tasks that run in isolated environments and use the developer's Copilot plan instead of JetBrains AI credits.

Key Takeaways

  • Copilot is now a selectable agent in JetBrains Air, so developers can use their existing Copilot subscription without a separate purchase.
  • The integration is built on the Agent Client Protocol (ACP), the open protocol for connecting coding agents to development environments.
  • Cloud tasks bill against the Copilot plan, not JetBrains AI credits, and they run in a fresh Docker container on a new virtual machine.
  • Developers can run several agent sessions in parallel and review each session's changes next to their code.
  • Cloud tasks need an Air Teams-enabled business organization and an Air administrator who has enabled Copilot. Local tasks only need a connected Copilot account.
  • There are governance gaps for now: Air analytics do not show Copilot usage, and admins can disable Copilot but cannot set spending limits in Air.

Copilot as an agent in JetBrains Air

GitHub Copilot can now be used in JetBrains Air to delegate coding tasks, guide how they run and review the results. The integration is built on the Agent Client Protocol (ACP), an open protocol for connecting coding agents to development environments. It supports tasks that run in a JetBrains IDE and tasks that run in the cloud. It works alongside the existing GitHub Copilot plugin for JetBrains rather than replacing it.

What developers get

Agent picker

Developers select GitHub Copilot from the agent picker in Air and use their existing Copilot subscription for coding tasks. No separate subscription is needed.

Parallel work and review

Air tracks multiple agent sessions at once. Each session can use project context and IDE tools, and developers can review the changes it produces next to their own code.

Cloud tasks

Through Air Teams, Copilot can run in isolated cloud environments, started from the IDE or from a browser. These tasks consume the developer's Copilot plan, not JetBrains AI credits.

Getting started

For local tasks, a developer installs Air in their JetBrains IDE, selects GitHub Copilot and connects a GitHub Copilot account.

Cloud tasks have more requirements. The developer needs a JetBrains account in an Air Teams-enabled business organization and a GitHub Copilot subscription. An Air administrator must also have enabled Copilot. There are two ways to start a cloud task:

  • Browser: Open Air in the browser, connect GitHub under Settings > VCS Providers, choose a repository and remote branch, select GitHub Copilot and send the prompt.
  • IDE: Use the Air plugin and choose Cloud from the environment selector.

Where cloud tasks run

Copilot runs in a Docker container on a fresh virtual machine. All traffic passes through a proxy and no state carries over between tasks. Copilot can build, test and install packages without touching the developer's local environment. JetBrains has published more detail on cloud agent sandboxing in Air.

Things to know about cloud tasks

  • Subscription checks: Connecting Copilot to Air does not request Copilot permissions. Air may verify the subscription or credits when the cloud environment starts. If they are unavailable, the task fails and an error appears in the session.
  • Plan selection: Developers cannot choose whether a task uses their personal plan or their organization's plan.
  • Usage visibility: Air's governance and analytics features do not include Copilot usage.
  • Admin controls: Air administrators can disable Copilot through a governance policy. They cannot set Copilot spending limits within Air.