← All posts

How to Keep Claude Code Running While You're Away From Your Laptop

May 24, 2026

Claude Code is great at long, multi-step work - refactors, test runs, build fixes, research. But those runs take time, and they pause when they hit usage limits. If you start one and then leave your desk, you lose visibility: Is it done? Did it stall waiting for input? Can I resume now that my limit reset?

Here’s a setup that lets the agent keep running on your machine while you check in - and steer it - from your phone.

The usual non-answers

The setup

Keep the agent on your machine and reach it from any browser:

  1. Run Remote Terminal Bridge on the machine where Claude Code runs. It’s a Windows tray app that connects outbound to a relay - no port forwarding, no VPN - and can autostart and autoconnect.
  2. Start your Claude Code session in that terminal as you normally would.
  3. From your phone (or any browser), sign in with Microsoft Entra ID and open the live session.

Because it’s a real PTY, you see Claude Code’s actual interface - the spinner, the diff view, the prompts - exactly as on your desk.

What this unlocks

Why PTY fidelity matters for AI agents

AI coding CLIs are rich terminal UIs now, not plain text streams. A screenshot or a simplified web shell loses the interaction. A real pseudo-terminal means the agent behaves identically whether you’re at your desk or on your phone - so you can actually answer its prompts and approve its steps, not just read output.

The same setup works for Copilot CLI and opencode too.

Security

Outbound-only from your machine, Microsoft Entra ID sign-in on both ends, and an encrypted connection. No inbound ports to open.

New to the pattern? Start with Access your laptop’s terminal from your phone.

Try Remote Terminal Bridge

7-day free trial. Cancel anytime in the Microsoft Store.

Get it from Microsoft Store