turix-mac
Computer Use Agent (CUA) for macOS automation using TuriX. Use when you need to perform visual tasks on the desktop, such as opening apps, clicking buttons, or navigating UIs that don't have a CLI or API.
Why use this skill?
Automate your macOS desktop with the TuriX-Mac skill. Enable OpenClaw to visually control apps, navigate UIs, and perform complex tasks through a powerful multi-model agent.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/tongyu-yan/turix-cuaWhat This Skill Does
The turix-mac skill empowers the OpenClaw agent to act as a Computer Use Agent (CUA) specifically optimized for macOS environments. It leverages a sophisticated multi-model architecture consisting of a Brain, Actor, Planner, and Memory module to visually interpret the macOS interface. Unlike traditional automation tools that rely on fragile selectors or APIs, TuriX perceives the screen, allowing it to interact with any application—even those lacking CLI or API support—by performing actions like clicking buttons, typing text, and navigating menus just as a human would.
Installation
To integrate this skill into your environment, use the OpenClaw command line interface. Execute the following command from your terminal: clawhub install openclaw/skills/skills/tongyu-yan/turix-cua. Ensure your system meets the hardware requirements, as the agent initializes multiple AI models which may take 2-5 minutes during the first launch. The skill is maintained within the openclaw/skills repository and authored by tongyu-yan.
Use Cases
This skill is ideal for automating repetitive visual workflows that are typically difficult to script. Common use cases include:
- Interacting with legacy desktop applications that lack automation hooks.
- Automating cross-application data transfers, such as extracting information from an email and pasting it into a desktop spreadsheet.
- Performing complex desktop configurations, like toggling system settings or managing file structures in Finder.
- Executing multi-step sequences that require autonomous reasoning, such as finding a specific invoice file and uploading it to a browser-based company portal.
Example Prompts
- "Open Spotify, search for my 'Focus' playlist, and play the first track."
- "Open System Settings, find the Display settings, and set the resolution to the optimized default."
- "Open Finder, navigate to my Downloads folder, identify the most recent PDF invoice, and upload it to the company portal in Chrome."
Tips & Limitations
For optimal results, provide explicit, high-level instructions rather than low-level coordinate mappings, as the agent is designed to understand semantic UI elements. Always be specific about which application to target. Please note that the agent requires system-level interactivity, so ensure permissions are granted in your macOS security settings. Avoid vague commands like "fix this," as the agent functions best with clear, sequential objectives. If a task is interrupted, you can resume it using the --resume [agent_id] flag provided by the run script. Finally, note that logs are buffered and saved to .turix_tmp/logging.log for detailed troubleshooting of LLM interactions and reasoning steps.
Metadata
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 skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-tongyu-yan-turix-cua": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: file-write, file-read, code-execution