Claude Code Goes Mobile With 'Remote Control' — Asynchronous Coding Is Here
Anthropic releases a mobile companion for Claude Code, letting developers command coding agents from their phone.
What Happened
Anthropic has released Claude Code Remote Control, a mobile app (iOS and Android) that lets developers issue commands to Claude Code sessions running on their desktop or cloud machines. Available exclusively to Claude Max subscribers ($100–$200/month), Remote Control supports project monitoring, task queuing, and real-time progress notifications — all from a smartphone.
The app doesn’t run Claude Code locally on the phone. Instead, it acts as a control plane: you can start tasks, review progress, approve changes, and receive alerts when the agent completes work or encounters blockers.
Why This Matters
Coding Decouples From the Desktop
The fundamental assumption of software development — that you need to be sitting at a computer — is being systematically dismantled. With Remote Control, the workflow becomes: queue tasks during your commute, let the agent work while you’re in meetings, review and approve results when convenient.
This is not about writing code on a tiny screen. It’s about managing an agent that writes code. The interface paradigm shifts from “editor” to “dashboard.”
The Async Development Loop
Remote Control enables a genuinely asynchronous development cycle:
- Morning commute: Queue today’s tasks from your phone
- At desk: Deep work alongside the agent, pair-programming style
- In meetings: Agent continues working, sends progress updates
- Evening: Review completed PRs, approve deployments
This mirrors how managers delegate to teams — except the “team” is an AI agent that works 24/7.
Pricing Signal
At $100–$200/month, Remote Control is clearly targeting professional developers and teams, not hobbyists. The willingness to gate mobile access behind premium pricing suggests Anthropic sees this as a high-value, high-retention feature — people who use it won’t downgrade.
What You Can Do
- Evaluate your workflow for tasks that could run unattended — tests, refactoring, documentation generation, and dependency updates are prime candidates.
- Set up proper CI/CD guardrails before using async agent workflows — automated tests and code review requirements prevent the agent from shipping unchecked code.
- Start with low-risk tasks — let the agent handle tech debt and documentation first, then graduate to feature development as you build trust in the async workflow.