switchboard
M

Microsoft UFO

by Microsoft

Automates Windows desktop apps through natural-language tasks with the UFO2 Desktop AgentOS, and orchestrates agents across Windows, Linux, and Android devices with UFO3 Galaxy.

4
Skills
None
Auth
No
Streaming
No
Push

Skills

Windows Desktop Automation

Completes multi-app Windows tasks from a natural-language request using a HostAgent and AppAgents, launched with python -m ufo.

Hybrid GUI Actions

Combines GUI clicks with native API calls through Windows UIA, Win32, and WinCOM, using visual plus UIA detection to find controls.

Speculative Multi-Action

Batches predicted actions into a single step to cut the number of LLM calls needed to finish a desktop automation task.

Galaxy Orchestration

Decomposes a request into a task DAG and assigns steps to Windows, Linux, and Android device agents that run in parallel.

Browser & Computer UseAutonomous Agentscomputer-usewindows-automationdesktop-automationgui-automationmulti-deviceopen-sourcemicrosoft-researchmicrosoft
Visit Agent
microsoft-ufo
Automates Windows desktop apps through natural-language tasks with the UFO2 Desktop AgentOS, and orchestrates agents across Windows, Linux, and Android devices with UFO3 Galaxy.
fields
nameMicrosoft UFO
providerMicrosoft
urlhttps://github.com/microsoft/UFO
categoriesbrowser-computer · autonomous-agents
accesscli
authnone
streamingfalse
pushfalse
verifiedtrue
tagscomputer-use, windows-automation, desktop-automation, gui-automation, multi-device, open-source, microsoft-research, microsoft
skills
windows-desktop-automationWindows Desktop AutomationCompletes multi-app Windows tasks from a natural-langua…
hybrid-actionsHybrid GUI ActionsCombines GUI clicks with native API calls through Windo…
speculative-multi-actionSpeculative Multi-ActionBatches predicted actions into a single step to cut the…
galaxy-orchestrationGalaxy OrchestrationDecomposes a request into a task DAG and assigns steps …