osaurus-macos-use

Automate macOS app and Safari interactions via accessibility APIs.

22|5|Updated Dec 2, 2025
One-click install
npx skills add https://github.com/osaurus-ai/osaurus-macos-use --skill osaurus-macos-use
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: osaurus-macos-use
Source: https://github.com/osaurus-ai/osaurus-macos-use/tree/main
Command: npx skills add https://github.com/osaurus-ai/osaurus-macos-use --skill osaurus-macos-use

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

macOS accessibility automation is hard because manual UI navigation and form filling demand watching screens and guessing element IDs, so this skill ensures precise interactions via accessibility APIs without re-observing before each action.

Core Features & Use Cases

  • Open-Observe-Act Workflow: Launch an app, observe with get_ui_elements to capture element IDs, and then use click_element, set_value, or press_key to act confidently.
  • Smart Re-observation Rules: Only re-run get_ui_elements after UI changes such as navigation, dialogs, or tab switches, keeping token usage low without stale IDs.
  • Safari and Native App Control: Browse the web, fill forms, navigate menus, or press shortcuts in Safari and other apps using accessibility commands with optional screenshots when necessary.

Quick Start

Launch Safari, observe its UI to gather element IDs, and then use accessibility actions like click_element or type_text to complete the task.

Frequently Asked Questions about osaurus-macos-use

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

FAQPage Schema
How do I automate macOS UI tasks using accessibility APIs?

Automate macOS UI tasks by following an open-observe-act workflow: launch your target app, use get_ui_elements to capture accessibility element IDs, and then send actions like click_element or set_value to interact confidently.

Can I use accessibility automation to fill forms and browse in Safari on macOS?

Yes, Safari browsing and form filling are supported via macOS accessibility commands. You can navigate menus, enter text, and press shortcuts in Safari and other native apps, using optional screenshots when necessary to verify UI states.

What is the correct workflow for clicking buttons and entering text on native macOS apps?

The correct workflow is open-observe-act. Launch the native macOS app, observe its UI with get_ui_elements to gather precise element IDs, and then use accessibility actions like click_element or type_text to complete your task.

When should I re-observe UI elements during macOS accessibility automation?

Re-observe UI elements with get_ui_elements only after significant UI changes like navigation, dialog appearances, or tab switches. This smart re-observation rule keeps token usage low while preventing stale element IDs during automation.

Does macOS accessibility automation work without manually guessing UI element IDs?

Yes, it eliminates guessing by using get_ui_elements to capture exact accessibility element IDs before sending actions. This ensures precise interactions via accessibility APIs without relying on manual screen observation or hardcoded element selectors.