Anthropic starts rolling out Remote Control for Claude Code, letting users control a session begun in the terminal from the Claude mobile app or the web
New in Claude Code: Remote Control. Kick off a task in your terminal and pick it up from your phone while you take a walk or join a meeting. Claude keeps running on your machine, and you can control the session from the Claude app or https://claude.ai/code [video]
@claudeaiClaude
Context & Ripple Effects
Remote Control extends Claude Code beyond its terminal origin while keeping execution on the user’s machine. It follows Anthropic’s earlier web and iOS expansion for Claude Code, turning those surfaces into ways to resume an already-running local session rather than merely alternate entry points.
The move also establishes the session as the product’s durable unit of interaction. Later Claude Code additions—including Slack task relay and channels for Telegram and Discord—fit that same model of reaching one coding session from multiple interfaces.
First-order effects
Claude Code users can start work from a terminal and monitor or steer that same session from Claude’s mobile app or web interface, reducing the need to remain at the originating computer.
Anthropic makes its mobile and web clients control surfaces for local developer workflows while the underlying session continues to run on the user’s machine.
Developers may increasingly evaluate coding assistants on continuity across terminal, browser, and phone—not only on code-generation quality—pressuring competing tools to connect their local sessions to more client surfaces.
Third-order effects
If this pattern persists, coding agents will shift from single-interface tools toward persistent work surfaces that users supervise across devices and communications channels.
That shift raises the importance of a control layer for long-running local work: product differentiation will increasingly depend on reliable session handoff and clear user control, not just model access.
The trend: Claude Code is part of the broader shift toward agentic work surfaces that keep work running while users move among interfaces.
Announcing a new Claude Code feature: Remote Control. It's rolling out now to Max users in research preview. Try it with /remote-control Start local sessions from the terminal, then continue them from your phone. Take a walk, see the sun, walk your dog without losing your flow. […
If this is being correctly reported, it should be the lead story in the country. The Pentagon is legally prohibited from engaging in mass surveillance of Americans. It has a policy requiring human oversight of autonomous weapons. Is Hegseth threatening to disregard any of that?
fuck yeah now you can run your claude code agent from your phone! message and vibe-code from the clyde app. all run on your desktop while you go for walks. direct response to openclaw
Ok now this is exciting. I don't want to baby sit my terminal when tinkering on long tasks and I also don't want to give Claude code full access to roam free on my computer. Epic win here. The pace of development in AI tooling has grown tenfold the last 90 days. Truly a
Regardless of everything else, Anthropic's shipping speed is seriously impressive. Back in my PhD days, I had a terminal app on my phone so I could ssh into my server, monitor training runs, and launch experiments when I was away from my laptop. Now you can orchestrate AI
This feature just made at least 6 startups irrelevant overnight. Happy, Claude Remote, VibeTunnel, Superconductor, Claudia, CodeRemote. All launched to solve exactly this problem. Some raised money. Some built paid apps. Some had thousands of users. Claude Remote literally