Claude Code On-the-Go (granda.org)

A developer has successfully set up a system enabling him to run six Claude Code agents simultaneously from his phone, using a combination of cloud resources and secure connections. By leveraging Termius for SSH access and Tailscale for a private network, he connects to a Vultr virtual machine (VM) to manage coding tasks without needing a laptop or desktop. The developer’s setup features asynchronous development, allowing him to initiate tasks on Claude Code and receive push notifications for input, streamlining productivity on-the-go. This innovation is significant for the AI/ML community as it showcases the feasibility of mobile development in machine learning environments, breaking the dependency on traditional desktop setups. With lightweight infrastructure—including a pay-per-use cloud VM costing only $0.29/hour—developers can work on multiple projects concurrently, utilizing Git worktrees for feature development and achieving high efficiency in coding sessions. The integration of tools like mosh for resilience during network transitions and tmux for persistent sessions enhances the mobile coding experience, indicating a potential shift toward more flexible work styles in tech.