Skip to content
SPECIFICTION
Get listed

open-computer-use

github.com

Free plan WindowsmacOSLinuxAPISelf-hosted
Visit github.com
Spec score 7.8 #6 of 24 in AI Computer Use Agents

What it is

open-computer-use is a free, MIT-licensed local MCP server that lets AI agents inspect and operate desktop applications through accessibility interfaces. Its nine core tools cover app discovery, accessibility-tree and screenshot capture, clicking, scrolling, dragging, typing, key presses, secondary actions, and control-value changes. The project says it runs entirely on the user’s machine and uses the accessibility layer for core app actions without taking over the physical mouse and keyboard. Setup documentation covers Codex CLI and App, Claude Code, Gemini CLI, and opencode, with manual configuration for other MCP clients. A ZCode plugin installs the skill and connects an MCP server. macOS 14 or later requires Accessibility and Screen Recording permissions. Windows and Linux use a signed-in desktop session, and Linux requires AT-SPI2. Additional window-level tools are currently available on Windows, while macOS and Linux support is in progress. Password managers are refused; app launching, focus changes, and global input injection require explicit environment-variable gates. The project warns that high-privilege automation should be used only on authorized machines.

Spec sheet

Purpose
Open Computer Use is a local MCP server that lets AI agents inspect and operate desktop apps through accessibility interfaces.Read 9 Oct 2026 · github.com
Platforms
The runtime supports macOS, Windows, and Linux, with nine core tools available across all three.Read 9 Oct 2026 · github.com
Integrations
The project documents setup commands for Codex CLI and App, Claude Code, Gemini CLI, and opencode, and supports manual configuration for other MCP clients.Read 9 Oct 2026 · github.com
License
The repository is distributed under the MIT license.Read 9 Oct 2026 · github.com
Desktop actions
Its nine core tools support app discovery, accessibility-tree and screenshot capture, clicking, scrolling, dragging, typing, key presses, secondary actions, and setting control values.Read 9 Oct 2026 · github.com
Local operation
The project says it runs entirely on the user's machine and uses the accessibility layer without taking over the real mouse and keyboard for its core app actions.Read 9 Oct 2026 · github.com
Permissions
On macOS 14 or later, the project requires Accessibility and Screen Recording permissions; Windows and Linux work in a signed-in desktop session, with Linux requiring AT-SPI2.Read 9 Oct 2026 · github.com
Platform limitation
The additional window-level tools are currently available on Windows, with macOS and Linux support in progress.Read 9 Oct 2026 · github.com
Safety controls
Password managers are refused, while app launching, focus changes, and global input injection require explicit environment-variable gates.Read 9 Oct 2026 · github.com
Security policy
The security policy says fixes are provided for the latest npm release and asks users to report suspected vulnerabilities through GitHub's private vulnerability reporting channel.Read 9 Oct 2026 · github.com
Security scope
The policy notes that the tool performs high-privilege accessibility and input automation in the local signed-in desktop session and should only be used on authorized machines.Read 9 Oct 2026 · github.com
Support and audience
The project provides installation and troubleshooting references and invites contributions; its security policy says fixes are delivered through the regular release process.Read 9 Oct 2026 · github.com
ZCode plugin
The repository includes a ZCode plugin that installs the skill and connects an MCP server.Read 9 Oct 2026 · github.com

Plans and prices

Open source Free MIT-licensed · runs locally · macOS, Windows, Linux Read 9 Oct 2026 · github.com

Every open-computer-use plan and price →

Strong points and limits

Strong points

  • Free MIT-licensed software
  • Runs locally on the user’s machine
  • Supports macOS, Windows, and Linux
  • Provides accessibility and screenshot tools
  • Integrates with several MCP clients

Limits

  • Additional window tools currently target Windows
  • macOS requires two system permissions
  • Linux requires AT-SPI2 and a signed-in session
  • High-privilege automation requires authorized machines

Who it suits

This tool suits users who want local AI control of desktop applications on macOS, Windows, or Linux. It is intended for authorized machines and MCP-based agent workflows.

Verdict

open-computer-use focuses on local desktop inspection and control through accessibility interfaces, with integrations for several MCP clients. Its operating-system permissions, signed-in-session requirements, and platform-specific window-tool status should be checked before deployment.

Where Specifiction places it

Alternatives to open-computer-useAll alternatives →