Mininglamp-AIMininglamp-AIOfficialยท1 Agent Skills Included

mano-skill

Natural language desktop GUI automation with vision models

Automates desktop tasks by capturing screenshots, analyzing them with vision models, and executing clicks, typing, scrolling, and drags. Eliminates manual clicking and repetitive form filling when no API or CLI exists for the target app. Supports fully on-device local mode for privacy and cloud mode for complex multi-step reasoning.
npx skills add Mininglamp-AI/mano-skill --all -g -y

All Skills in This Repository (1)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install mano-skill?โ–ผ

Run `npx skills add Mininglamp-AI/mano-skill --all -g -y` in your terminal to install this skill globally.

How to automate desktop apps with plain English?โ–ผ

Describe your task in natural language, and mano-cua captures the screen, plans each step with a vision model, and performs the clicks and typing for you.

Does mano-cua work without sending data to the cloud?โ–ผ

Yes. Local mode runs the Mano-P model entirely on-device via MLX on macOS Apple Silicon, so no screenshots or data ever leave your machine.

What actions can mano-cua perform on screen?โ–ผ

It supports click, type, hotkey, scroll, drag, mouse move, screenshot, wait, and app launch, all bounded by your task description and step limits.

Is mano-cua safe for sensitive tasks?โ–ผ

Yes. Sensitive or irreversible actions trigger a confirmation prompt, and the agent stops immediately if you touch the mouse or keyboard.

Related Repositories in Software Engineering

View All in Software Engineeringโ†’