GUI-Agent-Harness
Autonomous desktop control via screenshots, clicks, and verification
All Skills in This Repository (1)
Pure Emerald Level IndicatorsFrequently Asked Questions
FAQPage SchemaHow to install GUI-Agent-Harness?▼
Run `npx skills add Fzkuji/GUI-Agent-Harness --all -g -y` in your terminal to install this skill globally.
How to automate desktop tasks with plain language?▼
Run `gui-agent --work-dir /tmp/my-task "your task description"` and the agent screenshots the screen, detects UI elements, clicks, types, and verifies until the task is done.
Can it control a remote virtual machine?▼
Yes. Pass the VM's HTTP API address with the --vm flag, and the agent operates the remote desktop the same way it operates a local one.
Which platforms and models does it support?▼
It runs on macOS, Windows, and Linux, and works with providers like OpenAI, Anthropic, and Gemini, selectable via the --provider and --model flags.
Do I need coding skills to use GUI-Agent-Harness?▼
No. You describe the task in plain English, and the agent handles screen perception, action planning, and execution automatically.
Related Repositories in Software Engineering
View All in Software Engineering→
openclawopenclaw
Run a personal AI assistant across your devices and chat apps
obrasuperpowers
Gives coding agents a disciplined workflow from idea to merged code
facebookreact
AI agent skills for building, testing, and porting React core