#Enable openclaw to autonomously click around and see screen
1 messages · Page 1 of 1 (latest)
First thing: what does openclaw --version print? Version matters a lot for browser/tooling issues.
For this setup, I would not try to make OpenClaw control a full Wayland desktop on a VPS. Wayland is intentionally hostile to synthetic input/screen control, and VPS desktop sessions add extra moving parts. The reliable path is OpenClaw’s browser control: a dedicated Chromium/Chrome profile that the agent can snapshot, click, type, and screenshot.
On Ubuntu VPS I’d do this:
# avoid snap Chromium for this if possible
wget https://dl.google.com/linux/direct/google-chrome-stable_current_amd64.deb
sudo dpkg -i google-chrome-stable_current_amd64.deb
sudo apt --fix-broken install -y
Then set browser config in ~/.openclaw/openclaw.json:
{
browser: {
enabled: true,
executablePath: "/usr/bin/google-chrome-stable",
headless: true,
noSandbox: true,
defaultProfile: "openclaw"
},
tools: {
alsoAllow: ["browser"]
}
}
Restart the gateway, then test: