Claude Code on a Mac mini: an always-on server you drive from your phone

Why Claude Code needs a computer that stays on

The quickest start is the computer you already have. The Claude desktop app has Remote Control built in: on the computer you want to control, open Settings, turn on Connect new sessions to Remote Control, and open the session from the Claude app on your phone. Claude keeps running on that computer; the phone is a window into it.

The catch is that it only works while that computer is on. Close the laptop and your phone can't reach Claude Code until it wakes up again. A Mac mini that stays on at home removes that limit, which is what the rest of this guide sets up.

The setup at a glance

• A Mac mini at home, switched on and set not to sleep. It's the server: Claude Code, your repositories, logins and files live on it.
• Claude Code with Remote Control turned on, in the Claude desktop app or the CLI.
• The Claude app on your phone (iOS or Android), or claude.ai/code in any browser, to open those sessions from anywhere.
• Tailscale plus Screen Sharing, so you can open the Mac mini's desktop from a laptop as if you were sitting at it.

Why a Mac mini and not a Linux server

A rented Linux server stays on too, and that setup is in Claude Code on a VPS. A Mac adds things Linux can't do:

• Computer use, where Claude operates apps and the screen. In the CLI it's macOS only. In Claude Desktop it's a research preview on macOS and Windows, on Pro, Max or Ultra (not Team or Enterprise), and the app has to be running. See Anthropic's computer use docs.
• iOS apps. Xcode and the iOS Simulator run only on macOS, so if Claude Code works on an iPhone app, the server has to be a Mac.
• The full desktop app. Claude Desktop on Linux is a beta for Ubuntu and Debian, without computer use, Cowork or dictation.
• Your data at home. Repositories, browser logins and files stay on a machine you own instead of a provider's server.

How to set up Claude Code on a Mac mini

1. Keep the Mac mini awake

   In System Settings > Energy, turn on Prevent automatic sleeping when the display is off, so the Mac mini keeps running with its display off. Turn on Wake for network access as well, which lets it wake for shared services. See Apple's sleep and wake settings.

2. Install Claude Code and sign in

   Install the Claude desktop app, the Claude Code CLI or both, and sign in with your Claude account. Remote Control needs a Pro, Max, Ultra, Team or Enterprise plan. API keys don't work with it, and on Team or Enterprise an Owner has to turn it on first.

3. Turn on Remote Control

   In the desktop app, open Settings > Code (or Settings > Claude Code) and turn on Connect new sessions to Remote Control, so every new session can be opened from your phone. For a single session, type /remote-control or /rc in the Code tab. In the CLI, set remoteControlAtStartup to true in ~/.claude/settings.json, or run claude remote-control as a server that hosts several sessions at once.

   # Claude Code CLI: connect every new interactive session
   # (add this to ~/.claude/settings.json)
   { "remoteControlAtStartup": true }
   
   # Or run a Remote Control server that hosts several sessions at once;
   # --chrome turns on Chrome integration in its sessions
   claude remote-control --chrome

4. Open it from your phone

   Open the Claude app on iOS or Android, or claude.ai/code in a browser, and pick the session. Claude keeps running on the Mac mini; your phone sends messages and shows the results. If the network drops, Claude Code reconnects automatically when the Mac mini is back online. Setup details and fixes are in the Remote Control guide.

5. Put the Mac mini and your laptop on Tailscale

   Install Tailscale on the Mac mini and on the laptop you'll work from, and add both to the same tailnet. Tailscale creates a private peer-to-peer network between your devices with encrypted WireGuard connections, and only devices on your tailnet can reach each other. See what Tailscale is.

6. Turn on Screen Sharing

   On the Mac mini, open System Settings > General > Sharing and turn on Screen Sharing. It can't be on at the same time as Remote Management, so turn that off first if it's on. See Apple's Screen Sharing guide.

7. Work on it from your laptop

   From your laptop, connect with Screen Sharing to the Mac mini's name or address on your tailnet. You get its desktop: sign in to a website in its browser, open Xcode and the Simulator, or watch what Claude is doing, from home or anywhere else. Our founder works this way from a MacBook Air, and the files never leave the Mac mini.

Source: Anthropic's Remote Control documentation

What keeps it running, and what stops it

• The Claude Code process has to keep running. Quit the desktop app or close the terminal and the session goes offline, so leave it open on the Mac mini.
• Sleep and network drops are recoverable: Claude Code reconnects automatically when the machine is back online. Server mode (claude remote-control) is less patient. After about 10 minutes without a network it exits, so start it again.
• Remote Control drives a session that's already running. It doesn't start work on its own: when a task is done, Claude waits for your next message.
• Permission prompts and usage limits stop long runs too. How to run Claude Code 24/7 covers permission modes, /goal and plan limits.

Give the always-on Mac mini a standing job

An always-on Mac mini answers whenever you message it, and waits when you don't. Crewmark turns a Claude Code conversation on it into an AI employee with a role, goals, working hours and a heartbeat: during working hours the crewmark CLI wakes the employee to work through its issues on a shared board. The computer it runs on has to stay on, and the Mac mini does.

1. Hire an employee

   Pick a role from the AI employee roles, such as software engineer or SEO specialist, and write its goals and KPIs.

2. Connect Claude Code on the Mac mini

   Paste the setup link into a Claude Code conversation on the Mac mini. It installs the crewmark CLI and links that conversation to the employee. Codex works too.

3. Set working hours and a heartbeat

   Choose the hours, weekdays only or every day, and how often the employee wakes: every 15, 30, 60, 120 or 240 minutes. Hours can run past midnight, so on a machine that never sleeps a night shift works too.

4. Assign issues from your phone

   Create and assign issues on the board at crewmark.ai from any browser; assigning one wakes the employee right away. It keeps each issue's status current, and when it needs a review or a decision it sets the issue to Needs action and asks you there. Jobs to give it are in Claude Code use cases.

FAQ