OpenClaw Browser Automation Guide

The default openclaw browser profile is a separate agent-controlled Chromium profile. It does not use your personal browsing profile or logged-in sessions unless you explicitly select an existing-session or extension-based profile.

Diagnose and start the managed profile

openclaw browser --browser-profile openclaw doctor
openclaw browser --browser-profile openclaw status
openclaw browser --browser-profile openclaw start
openclaw browser --browser-profile openclaw open https://example.com
openclaw browser --browser-profile openclaw snapshot

If the browser command is missing, inspect the browser plugin. If the CLI works but the agent cannot call it, inspect the effective tool profile. The normal coding profile does not automatically expose the full browser tool.

Pick the correct profile

• openclaw: isolated managed browser; safest default.
• user: attaches to a real signed-in Chromium session and requires local approval.
• chrome: extension-based access to signed-in tabs.

Use a real-session profile only when an existing login is necessary. A browser session can access sensitive account state, so narrow the target site and task.

Operate predictably

Keep one task tab, take a snapshot before acting, and take a new snapshot after navigation or a major UI change. Treat login, CAPTCHA, 2FA, camera, and microphone prompts as manual blockers rather than guessing.

Expected result

• Doctor identifies a usable Chromium binary or remote profile.
• Status reports the intended profile.
• The test page opens and returns a readable snapshot.
• The agent can call browser only where policy allows it.

Use Browser not working for launch, profile, display, or policy failures.