OpenClaw allows your agent to control a dedicated Chrome/Brave/Edge/Chromium profile. This provides a safe, isolated environment for browser automation and web scraping, distinct from your personal browsing session.

• Isolation: Keeps agent activity separate from your personal work.
• Safety: Prevents the agent from accidentally closing your tabs or leaking your session data.
• Capabilities: Enables full automation (snapshots, PDF generation, screenshots, complex interactions).
• Multi-Profile: Support for multiple isolated profiles (e.g., work, research, testing).

1. Check Status Verify if the browser service is ready.

   openclaw browser --browser-profile openclaw status

2. Start the Browser Launch the managed browser instance.

   openclaw browser --browser-profile openclaw start

3. Open a Page Direct the agent to a URL.

   openclaw browser --browser-profile openclaw open https://example.com

4. Take a Snapshot See what the agent sees.

   openclaw browser --browser-profile openclaw snapshot

To use the managed browser by default, set browser.defaultProfile: "openclaw" in your config.

Browser settings are located in ~/.openclaw/openclaw.json.

{
  browser: {
    enabled: true,                    // Master switch
    // cdpUrl: "http://127.0.0.1:18792", // Legacy override
    remoteCdpTimeoutMs: 1500,         // Timeout for remote CDP checks
    remoteCdpHandshakeTimeoutMs: 3000,
    defaultProfile: "chrome",         // Change to "openclaw" for isolation
    color: "#FF4500",                 // UI tint color
    headless: false,                  // Set true for background operation
    noSandbox: false,
    attachOnly: false,                // "true" = never launch, only attach
    executablePath: "/Applications/Brave Browser.app/Contents/MacOS/Brave Browser",
    profiles: {
      openclaw: { cdpPort: 18800, color: "#FF4500" },
      work: { cdpPort: 18801, color: "#0066CC" },
      remote: { cdpUrl: "http://10.0.0.42:9222", color: "#00AA00" }
    }
  }
}

If your system default browser is Chromium-based, OpenClaw uses it automatically. You can force a specific browser executable if needed.

openclaw config set browser.executablePath "/usr/bin/google-chrome"

{
  browser: {
    executablePath: "/Applications/Brave Browser.app/Contents/MacOS/Brave Browser"
  }
}

{
  browser: {
    executablePath: "C:\\Program Files\\BraveSoftware\\Brave-Browser\\Application\\brave.exe"
  }
}

{
  browser: {
    executablePath: "/usr/bin/brave-browser"
  }
}

OpenClaw supports flexible deployment models:

1. Local Control (Default): Gateway runs locally, launches browser locally.
2. Remote Control (Node Host): Gateway runs elsewhere, but uses a “Node Host” on your machine to control the browser.
3. Remote CDP: Gateway connects directly to a remote debugger (e.g., a browser in a Docker container or another server).

You can connect to a remote browser instance using a CDP URL.

• Query Token: https://provider.example?token=<token>
• Basic Auth: https://user:pass@provider.example

Browserless provides a hosted Chromium service, perfect for serverless deployments or when you don’t want to manage local browser resources.

{
  browser: {
    enabled: true,
    defaultProfile: "browserless",
    profiles: {
      browserless: {
        cdpUrl: "https://production-sfo.browserless.io?token=<BROWSERLESS_API_KEY>",
        color: "#00AA00"
      }
    }
  }
}

Drive your existing Chrome tabs (with your login sessions) using the Chrome Extension.

1. Install Extension Load the extension in “Developer Mode”.

   openclaw browser extension install

   Navigate to chrome://extensions, enable Developer Mode, and “Load Unpacked” from the path provided.

2. Attach to Tab Click the OpenClaw icon on the tab you want to control. The badge will show ON.

3. Control via Agent The agent can now use the chrome profile to interact with that tab.

   openclaw browser --browser-profile chrome tabs

For a full guide, see Chrome Extension.

Common issues and how to solve them.

See Browser CLI Reference for a complete list of commands.