Anthropic documented Remote Control, which connects claude.ai/code or the Claude mobile apps to a Claude Code session running on your own machine. A task started at a desk can be picked up from a phone or another browser, and the important detail is what does not move: Claude keeps running locally the entire time, so code execution and filesystem access stay on the machine that started the session. The full local environment travels with it, including MCP servers, tools, project configuration and @ autocomplete over local file paths.
The access model is worth reading carefully. Remote Control is available on all plans, but on Team and Enterprise it is off by default until an Owner enables the toggle in Claude Code admin settings. That ordering is the right way round: a feature that turns a personal laptop into a remotely addressable execution surface is an administrative decision before it is a convenience.
Separating the control surface from the execution boundary is the distinction behind local coding agents becoming shared workspaces, and it is what the agent lock-in test being a full session export measures. The coding agents hub tracks both halves.
This is product documentation: it describes availability and the default-off enterprise toggle without publishing the session transport's audit trail. The condition to watch is whether remote-initiated actions are distinguishable in logs from local ones, because an execution boundary that stays on the laptop is only meaningful if the record shows who reached it and from where.