Quicksand

Quicksand gives AI agents a Linux virtual machine for executing code and working with desktop apps on your own hardware. It's for developers who need a separate operating system for agent tasks, controlled through an async Python API.

• Command execution: Agents can run shell commands inside the VM and receive output, errors and exit status.
• Checkpoints and saved state: Capture the full VM state to roll back an operation, or save disk state and resume later, including on another machine.
• Desktop interaction: Agents can take screenshots, type, click and move the mouse in an Xfce4 desktop with Firefox or Chromium.
• Network and file access: Sandboxes have networking disabled by default, with optional internet access and port forwarding; host directory sharing gives agents access to selected files.

Prebuilt Ubuntu and Alpine images cover command-line and graphical tasks. Agent-focused images include Python development tools, while a computer-use image adds Playwright and browser access through noVNC. Multiple agents can have separate Linux user accounts within one VM.

Quicksand uses QEMU virtual machines with their own kernels and hypervisor-level isolation. It runs locally on macOS, Linux and Windows, supports x86_64 and ARM64, and can use hardware acceleration. It doesn't require root privileges, Docker or a cloud account. The Python project is open source under the MIT license.