ClawKit Logo
ClawKitReliability Toolkit
Back to Registry
Official Verified productivity Safety 2/5

mac-control

Control Mac via mouse/keyboard automation using cliclick and AppleScript. Use for clicking UI elements, taking screenshots, getting window bounds, handling coordinate scaling on Retina displays, and automating UI interactions like clicking Chrome extension icons, dismissing dialogs, or toolbar buttons.

Why use this skill?

Automate macOS desktop interactions with the OpenClaw mac-control skill. Use precise mouse clicking, keyboard input, and screen analysis to manage apps.

skill-install — Terminal

Install via CLI (Recommended)

clawhub install openclaw/skills/skills/easonc13/mac-control
Or

What This Skill Does

The mac-control skill provides OpenClaw with the ability to physically interact with your macOS desktop environment by simulating human mouse and keyboard inputs. By leveraging high-precision automation tools like 'cliclick' and 'osascript', the agent can see, move, click, and type within any graphical interface, even those that do not provide native APIs or accessibility hooks. This allows the AI to interact with closed-source applications, browser extensions, system dialogs, and native menu bars effectively.

Installation

To integrate this skill into your environment, use the OpenClaw command-line interface: clawhub install openclaw/skills/skills/easonc13/mac-control Once installed, ensure that your user account has the necessary 'Accessibility' and 'Screen Recording' permissions granted in macOS System Settings for the terminal or shell executing the agent, as these are required to allow the agent to move the cursor and capture the screen.

Use Cases

  • Browser Automation: Interacting with specific Chrome extension icons that are not reachable via standard web automation libraries.
  • Desktop App Management: Navigating legacy software that lacks a command-line interface or JSON-based configuration.
  • Workflow Automation: Automating repetitive GUI tasks such as dismissing recurring OS dialogs, clicking 'Save' buttons in non-scriptable apps, or filling out desktop forms.
  • Visual Testing: Using screenshot analysis to verify that buttons or UI elements have appeared on screen after a specific set of actions.

Example Prompts

  1. "Open the browser, click on the extension icon in the top right corner, and click the 'Start Recording' button inside the dropdown."
  2. "Take a screenshot, identify the 'Submit' button in the application window, and click it."
  3. "Dismiss the system update prompt by pressing the Escape key and then click on the 'Remind Me Later' button if it appears."

Tips & Limitations

  • Coordinate Calibration: The skill is calibrated for a 1920x1080 display with 1:1 scaling. If you move your agent to a Retina display, remember to divide your screen coordinates by two, or run the provided calibrate-cursor.sh script to ensure accuracy.
  • Visual Verification: Always use the 'Screenshot -> Analyze -> Click -> Verify' workflow. Since the AI cannot 'feel' the UI, taking a post-action screenshot is the only way to confirm success.
  • System Permissions: If actions aren't working, verify that your terminal application is enabled in 'System Settings -> Privacy & Security -> Accessibility'.

Metadata

Author@easonc13
Stars2387
Views0
Updated2026-03-09
View Author Profile
AI Skill Finder

Not sure this is the right skill?

Describe what you want to build — we'll match you to the best skill from 16,000+ options.

Find the right skill
Add to Configuration

Paste this into your clawhub.json to enable this plugin.

{
  "plugins": {
    "official-easonc13-mac-control": {
      "enabled": true,
      "auto_update": true
    }
  }
}

Tags(AI)

#mac-automation#gui-testing#cliclick#macos#desktop-control
Safety Score: 2/5

Flags: file-write, file-read, code-execution