microsoft/ufo is a Windows desktop AgentOS for single-device GUI automation paired with Galaxy, a multi-device orchestration framework that coordinates intelligent agents across Windows, Linux, and Android.
Project overview
It bridges a deep Windows-native automation layer with cross-platform agent orchestration, addressing the gap of automating GUI-based tasks and complex workflows across heterogeneous devices within a single framework.
Project type
AI Agent · Workflow
Use cases
Documents & Office · Automation
Deployment
Refer to project documentation
License
MIT
Best for
Developers and researchers building AI agent automation solutions that require deep Windows OS integration or multi-device orchestration.
Key capabilities
Decomposes user requests into dynamic DAGs and coordinates asynchronous execution across heterogeneous platforms.
Deep Windows OS integration with hybrid GUI and API actions for single device task execution.
Template-driven toolkit with Model Context Protocol (MCP) integration for tool augmentation.
Uses visual and UIA detection for robust control identification.
Limitations and risks
UFO² standalone execution is focused on Windows OS only.
Getting started
Install dependencies via pip install -r requirements.txt, copy the config template to a YAML file and add your LLM API key, then launch the interactive environment using python -m galaxy --interactive.
Evidence and sources
GitHub project description: UFO³: Weaving the Digital Agent Galaxy