browser-extension-clicker

What This Skill Does

The browser-extension-clicker is a specialized automation skill designed to interact with browser user interfaces at the system level. By utilizing GUI automation libraries like pyautogui and Pillow, this skill bypasses standard browser security restrictions that prevent scripts from directly interacting with extension icons. It captures the screen, performs image template matching to locate specific icons—such as the OpenClaw Browser Relay—and triggers precise mouse clicks on those coordinates. This enables seamless control over browser plugins that are otherwise locked away from standard web-based automation tools.

Installation

To integrate this skill into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/binbinhihi/openclaw-browser-extension-clicker. Ensure that you have the necessary system permissions, particularly on macOS, where the terminal and Python process require Accessibility/Screen Recording permissions under System Settings to function correctly. You will also need to ensure the required Python packages are present by running pip install pyautogui Pillow.

Use Cases

This skill is perfect for scenarios requiring persistent interaction with browser toolbars. It is most commonly used to activate the OpenClaw Browser Relay, toggle developer extensions, or access context menus within browser plugins. It is particularly valuable for automated workflows where you need to refresh an extension's connection, clear cached data in a popup, or initiate a task via a plugin interface that doesn't provide a public API.

Example Prompts

1. "Click the OpenClaw browser extension icon to activate the relay for my current task."
2. "Open the browser extension menu and click the OpenClaw icon using the default settings."
3. "Verify the position of my browser extension icon by taking a screenshot using the clicker tool."

Tips & Limitations

Because this tool relies on visual coordinate detection, it is resolution-sensitive. If your monitor setup or browser window size changes, the hardcoded coordinates may fail; in these cases, rely on the automatic detection logic rather than manual overrides. Always use the --screenshot flag initially to calibrate the tool for your specific display configuration. Note that this skill simulates physical mouse clicks, meaning your mouse cursor will move during execution—it is recommended to avoid moving your mouse while the automation is running. It operates best when the target icon is visible and not hidden inside the browser's hidden extensions overflow menu.