cua-driver

Automate macOS app and browser interactions via OpenClicky's MCP server.

471|92|Updated Apr 23, 2026
One-click install
npx skills add https://github.com/jasonkneen/openclicky --skill cua-driver-jasonkneen
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: cua-driver
Source: https://github.com/jasonkneen/openclicky/tree/main/AppResources/OpenClicky/OpenClickyBundledSkills/cua-driver
Command: npx skills add https://github.com/jasonkneen/openclicky --skill cua-driver-jasonkneen

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill streamlines complex interactions with macOS applications and web browsers, enabling precise automation without compromising user experience.

Core Features & Use Cases

  • macOS App Automation: Automate tasks in native macOS applications using Accessibility and Screen Recording permissions.
  • Browser Control: Navigate and interact with web pages and applications within the browser.
  • Use Case: Automate repetitive tasks in an application like Finder, perform complex interactions in a web application like Slack, or navigate and interact with Google Calendar within a browser.

Quick Start

Use the cua-driver skill to open a Finder window and navigate to the 'Downloads' folder.

Frequently Asked Questions about cua-driver

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate macOS app interaction and browser control?

You can automate macOS app interaction and browser control by using the cua-driver Skill, which leverages the OpenClicky local computer-use MCP server to execute precise GUI tasks. It requires macOS Accessibility and Screen Recording permissions.

Do I need specific macOS permissions for GUI automation?

Yes, GUI automation on macOS requires explicit Accessibility and Screen Recording permissions. These allow the underlying OpenClicky server to interact with native applications and capture visible UI elements for structured routing.

Can I automate browser tasks like navigating Google Calendar on macOS?

Yes, you can automate browser tasks like navigating Google Calendar. The cua-driver Skill supports navigating and interacting with web pages and web applications directly within the browser on macOS.

How does macOS GUI automation verify actions were executed correctly?

macOS GUI automation verifies actions through a re-snapshot mechanism. After performing an interaction, it captures a new snapshot of the visible UI to confirm the structured route and action were successfully executed.

What is the best way to automate repetitive tasks in Finder or Slack?

The best way to automate repetitive tasks in Finder or Slack is using a computer-use MCP server. This approach enables structured routes over visible UI elements for both native macOS apps and browser-based web applications.

Why is my macOS automation not interacting with visible UI elements?

macOS automation may fail to interact with visible UI elements if Accessibility or Screen Recording permissions are missing. The OpenClicky local computer-use server requires both to map structured routes and verify actions via re-snapshot.