AI

GitHub Copilot now works in JetBrains Air with cloud task support

GitHub Copilot integrates with JetBrains Air, letting developers delegate coding tasks to run locally or in isolated cloud environments.

2 min readSource: GitHub Changelog

Curated by Kandala Guruprasad. Drafted with AI and checked against the source.

GitHub Copilot now works in JetBrains Air with cloud task support

TL;DR

  • GitHub Copilot integrates with JetBrains Air through the Agent Client Protocol, supporting both local and cloud-based coding tasks
  • Use your existing GitHub Copilot subscription for tasks in Air; cloud tasks run in isolated Docker containers on fresh virtual machines
  • Track multiple agent sessions in parallel and review changes in context without modifying your local environment
  • Cloud tasks are available to JetBrains Air Teams members with a GitHub Copilot subscription and Air administrator approval
Integration protocol
Agent Client Protocol (ACP)
Task environments
Local IDE or isolated cloud
Subscription required
GitHub Copilot plan

What happened

GitHub Copilot is now available within JetBrains Air, GitHub's cloud-based development environment. The integration uses the Agent Client Protocol (ACP) and allows developers to delegate coding tasks either within their JetBrains IDE or through cloud environments. Developers can choose between using Copilot in Air or the existing GitHub Copilot plugin, depending on their preferred workflow.

Local and cloud task workflows

For local tasks, developers install Air in their JetBrains IDE, select GitHub Copilot as the agent, and connect their GitHub Copilot account. For cloud tasks, Air Teams members within a business organization can run Copilot in isolated environments through either the browser interface or the IDE plugin. Cloud tasks execute in Docker containers on fresh virtual machines, with all traffic routed through a proxy and no state carried between sessions. This isolation means Copilot can build, test, and install packages without affecting the local development environment.

  • Access cloud tasks via browser: connect GitHub under Settings > VCS Providers, create a task, choose repository and branch, select GitHub Copilot, and send a prompt
  • Use the Air plugin in JetBrains IDE: select Cloud from the environment selector to run tasks from the IDE

Key capabilities and limitations

GitHub Copilot in Air supports parallel work across multiple agent sessions, allowing developers to track several tasks simultaneously while using project context and IDE tools. Changes can be reviewed alongside existing code. Cloud tasks consume your GitHub Copilot subscription plan rather than JetBrains AI credits. However, there are important considerations: subscription verification occurs only when the cloud environment starts, so unavailable credits or subscriptions will cause the task to fail with an error. Developers cannot choose whether a task uses a personal or organizational GitHub Copilot plan, and Air's governance and analytics features do not currently include Copilot usage tracking. Administrators can disable GitHub Copilot through governance policies but cannot set spending limits within Air.

Getting started and prerequisites

To use local tasks, install Air in your JetBrains IDE and connect your GitHub Copilot account. For cloud tasks, you need a JetBrains account within an Air Teams-enabled business organization, an active GitHub Copilot subscription, and Copilot enabled by your Air administrator. Once configured, developers can delegate coding work and review results without needing to run tasks locally, which can be useful for resource-intensive builds or testing workflows.

Source

GitHub Changeloghttps://github.blog/changelog/2026-11-03-github-copilot-is-now-available-in-jetbrains-air

AI-assisted, verified against the source. This post was drafted with AI from the primary source linked above and checked automatically against it. For anything you depend on, read the original.

  • #githubcopilot
  • #jetbrains
  • #ide
  • #aicoding
  • #clouddevelopment

More in AI