Deploy OpenClaw to a server or keep it running reliably when your laptop sleeps.

Read this when your laptop sleeping breaks scheduled automations, or when you want your agent available from anywhere.

The problem with laptop deployments

OpenClaw runs on your machine. When your Mac sleeps, the Gateway pauses. Cron jobs don't fire. Alerts don't send. For occasional use this is fine. For daily briefings or real-time monitoring, it's a problem.

Three solutions, in order of increasing permanence:

Prevent sleep (simplest — keep your current setup)
Deploy to a always-on server (most reliable)
Use a small VPS or home server (good balance)

Option 1: Prevent sleep on Mac

For a dedicated machine that's always on:

# Keep the display off but prevent sleep
caffeinate -d &

Or in System Preferences → Battery → "Prevent your Mac from automatically sleeping when the display is off".

Simple. Works if you have a machine that's always on (Mac mini, older MacBook plugged in).

Option 2: Deploy to a VPS

A $5-6/month VPS (DigitalOcean Droplet, Hetzner CX11, Linode Nanode) is enough to run OpenClaw 24/7.

1. Provision a server running Ubuntu 22.04+

2. Install OpenClaw:

curl -fsSL https://openclaw.ai/install.sh | bash
openclaw onboard --install-daemon

3. Copy your config:

scp ~/.openclaw/openclaw.json user@your-server:~/.openclaw/
scp ~/.openclaw/SOUL.md user@your-server:~/.openclaw/
scp ~/.openclaw/USER.md user@your-server:~/.openclaw/

4. Start the daemon:

ssh user@your-server "openclaw gateway start --daemon"

The daemon registers as a systemd service and starts automatically on boot.

5. Verify:

ssh user@your-server "openclaw gateway status"

Using systemd directly

If you prefer to manage the service yourself:

# /etc/systemd/system/openclaw.service
[Unit]
Description=OpenClaw Gateway
After=network.target

[Service]
Type=simple
User=ubuntu
ExecStart=/usr/local/bin/openclaw gateway
Restart=on-failure
RestartSec=10

[Install]
WantedBy=multi-user.target

sudo systemctl enable openclaw
sudo systemctl start openclaw
sudo systemctl status openclaw

Keeping your server config in sync

When you update SOUL.md, USER.md, or openclaw.json locally, push changes to the server:

# Quick sync script
rsync -avz ~/.openclaw/ user@your-server:~/.openclaw/ --exclude workspace/
ssh user@your-server "openclaw gateway restart"

Add this to a shell alias for convenience.

Monitoring your server

Check if the Gateway is alive:

ssh user@your-server "openclaw gateway status && openclaw cron list"

View logs:

ssh user@your-server "openclaw logs --tail 50"

Set up a health check cron that alerts you if the Gateway stops:

{
  "id": "self-check",
  "schedule": "*/30 * * * *",
  "prompt": "Self-check: confirm you're running and responding. Reply with 'ok' only.",
  "channel": "telegram",
  "suppressIfOk": true
}

Security for server deployments

On a VPS, add these protections:

Firewall — only allow known ports:

ufw allow 22/tcp   # SSH
ufw allow 443/tcp  # HTTPS (if using reverse proxy)
ufw deny 18789     # Block Gateway port from the internet
ufw enable

The Gateway should NOT be directly accessible from the internet. Use a reverse proxy (nginx/Caddy) with authentication if you need external webhook access.

Keep the Gateway on localhost:

The default config binds to 127.0.0.1:18789. Don't change this to 0.0.0.0 unless you've added authentication.

Home server option

A Raspberry Pi 4 or any always-on home machine works just as well as a VPS — and costs nothing monthly. The setup is identical to the VPS approach.

Advantage: your data stays completely on your home network. Disadvantage: you need to handle dynamic DNS and port forwarding if you want external webhook access.

Run It 24/7