Copilot CLI Goes Remote

GitHub Copilot CLI sessions can now be managed remotely via web and mobile, enabling continuous workflows across devices.

2 min read
Illustration of a developer using GitHub Copilot CLI on a laptop and phone.
GitHub Copilot CLI remote control allows management across devices.· Github Blog

GitHub is extending the reach of its AI coding assistant, allowing developers to manage GitHub Copilot CLI remote control sessions from anywhere. Previously confined to a local machine, these AI-driven coding tasks can now be monitored and directed through the GitHub website and its mobile app.

The move aims to bridge the gap between local development environments and on-the-go access, enabling a more fluid workflow. Developers can initiate tasks like refactoring code or debugging tests in VS Code or the CLI and then track their progress or issue new commands from their phone.

Related startups

Seamless Multi-Surface Control

This new functionality, now generally available, allows users to start a Copilot session and then use the `/remote on` command to access it via github.com or the GitHub Mobile app. This provides real-time visibility into the AI's actions, including research, file modifications, and command execution.

Developers can intervene mid-session, providing follow-up instructions to redirect the AI or expand its scope. Permission requests can also be managed remotely, offering a continuous workflow from initiation to pull request review and merging, all from a mobile device.

The remote control extends to VS Code and JetBrains IDEs, solidifying GitHub Copilot's presence across multiple platforms. Sessions remain private by default, accessible only to the user initiating them.

This feature is a significant step towards GitHub's vision of an end-to-end agentic platform. To get started, users can install the latest GitHub Copilot CLI or ensure their VS Code extension is updated.

© 2026 StartupHub.ai. All rights reserved. Do not enter, scrape, copy, reproduce, or republish this article in whole or in part. Use as input to AI training, fine-tuning, retrieval-augmented generation, or any machine-learning system is prohibited without written license. Substantially-similar derivative works will be pursued to the fullest extent of applicable copyright, database, and computer-misuse laws. See our terms.