Skip to main content
AgentOS is the runtime layer that gives AI agents direct control over a machine’s operating system — keyboard, mouse, screen, and system-level functions. It sits between your agent code and the OS, handling the low-level interactions so your agents can focus on tasks. Designed for everyone, not just developers. AgentOS is built so that non-technical users can install it and get it running without writing code or configuring system internals. The UI installer guides you through setup, the SDK auto-detects the service, and sensible defaults mean you don’t need to touch configuration files to get started. Platform independent. AgentOS runs on Windows, macOS, and Linux. In Companion Mode, it can also control Android devices, iOS devices (planned), and any machine reachable through hardware interfaces — regardless of what OS the target runs.

How It Works

AgentOS operates in two control modes:
  • Host Mode — AgentOS runs directly on the target machine as software. It controls the desktop via OS-level APIs with full access to screenshots, input, windows, and processes.
  • Companion Mode — AgentOS runs on an external device (e.g. Raspberry Pi) and controls the target machine through hardware interfaces (USB, Bluetooth, HDMI capture). No software installation on the target required.
Both modes expose the same capabilities to your agents — the building blocks for screen reading, input simulation, window management, and more.

Why AgentOS?

vs. Open Source (PyAutoGUI, xdotool, etc.)

Open-source desktop automation libraries run in the user’s session only. That works for simple scripting but breaks down for real-world agent deployments:
  • No service mode — when the user logs off or the RDP session disconnects, automation stops.
  • No logon screen control — you can’t automate Windows logon, lock screens, or UAC prompts.
  • No CTRL+ALT+DEL — Secure Attention Sequence requires a kernel-level driver. User-space tools cannot send it.
  • No session resilience — if an RDP session drops, the desktop locks and screenshots go black.
  • User-level privileges only — no access to SYSTEM-level operations.

vs. Building It Yourself

You can build OS-level control from scratch. Here’s what that involves:
  • Windows service with session management — running as SYSTEM, attaching to interactive sessions, handling session 0 isolation.
  • Secure Attention Sequence driver — a signed kernel driver to send CTRL+ALT+DEL.
  • RDP session transfer — detecting disconnects and keeping the desktop alive for screenshot capture.
  • Logon screen interaction — injecting input on the secure desktop.
  • Cross-version compatibility — handling differences across Windows 10, 11, Server 2019, 2022.
This is months of kernel and systems-level engineering before you write a single line of agent logic.

Comparison

CapabilityAgentOSOpen SourceDIY
OS service modeYesNoMonths of work
RDP resilienceYesNoKernel-level effort
Logon screen controlYesNoRequires SAS driver
Send CTRL+ALT+DELYesNoRequires signed kernel driver
CI/CD headlessYesNoCustom service wrapper
SYSTEM privilegesYesNoManual service setup
External device controlYesNoCustom hardware integration
Industry protocolsPlannedNoCustom integration per protocol
Maintained and versionedYesVariesOn you
Time to first automationMinutesHoursMonths

What’s Next?

Quickstart

Get AgentOS running in under 5 minutes.

Control Modes

Host Mode vs Companion Mode — how AgentOS connects to targets.

Deployment Scenarios

Find the right setup for your target.