"Papers With Code" matching MCP servers:

Enables OpenAI Codex to inspect and operate Windows desktop applications in the background while the user keeps control of the mouse, keyboard, and foreground window, with each action verified against the application's real state.

Enables AI agents to control Windows desktop applications by wrapping Codex's Computer Use capability.

A local MCP server that lets compatible agents use the official, signed Computer Use broker bundled with the OpenAI ChatGPT macOS app.

Enables AI agents to control macOS via the ZCode CUA 0.5.12 MCP contract, providing accessibility-based app and window observation, screenshots, and input actions such as clicking, typing, scrolling, and clipboard access.

Enables control and administration of a Paper Minecraft server, including process management, in-game commands via RCON, file editing, and log monitoring.

MCP server for remote execution on Windows devices. Enables Codex to run PowerShell, manage files/processes, and capture screenshots via a relay and enrolled Windows agent.

Connects ChatGPT Web to a self-hosted local coding workbench, enabling file editing, shell commands, Git/GitHub operations, multi-workspace management, code review, terminal processes, checkpoints, and optional Docker sandboxing.

Provides tools for obtaining system CPU and memory statistics through FastAPI endpoints exposed via MCP.

Enables GUI automation by executing Python code with instrumented pyautogui, providing MCP-friendly output and inline screenshot delivery.

Installs an agent on Linux hosts to allow shell commands, file edits, and service management through Claude.ai or ChatGPT via a secure outbound WebSocket connection.

A compact facade MCP server for macOS desktop automation, lazily proxying to full-featured computer-use tools to save context tokens while enabling screenshots, window control, app launching, and AppleScript execution.

Restores OpenAI Codex Computer Use on macOS 13 by building a clean-room MCP backend that adapts to the bundled sky interface, enabling screen interaction and app control without modifying system libraries.

Exposes MCP tools that let agents inspect and control any Mac application through accessibility trees and screenshots, enabling background clicks, typing, scrolling, dragging, and app-state reads.

A Windows-only local MCP server that lets Codex/ChatGPT agents control the keyboard, operate foreground windows, and send input to applications and games running with administrator privileges.

Enables ChatGPT Web conversations to operate a Windows computer through MCP with local tools, permissions, task state, and audit evidence.

Enables MCP clients on Linux to drive the official Codex Computer Use and Browser Use runtime on a remote Windows PC over SSH, operating in the logged-in Windows desktop session.

Enables driving macOS apps from Claude Code — screenshots, accessibility trees, clicks, typing and keys — by borrowing the Computer Use eyes and hands from OpenAI's ChatGPT/Codex desktop app, with persistent JavaScript scripting, text-targeted batching, diff output and macros that collapse many actions into one call without consuming Codex quota or calling a model.

Sends native Windows desktop notifications with repeating sound alerts for Claude Code, supporting states like finished, needs_help, and pending_review.

Enables MCP agents to drive Windows applications by targeting individual windows—clicking, typing, scrolling, dragging, and reading accessibility trees or screenshots—without stealing the user's focus, mouse, or keyboard. Runs entirely locally alongside the user, with per-app approval and no cloud account or usage limits.

Enables MCP clients to use the installed desktop computer-use runtime for Windows and browser observation and interaction through its tools and skills.