Contop vs open-computer-use

Contop is an open-source remote compute agent for controlling a desktop from a phone. It streams the screen in real time, accepts voice or text input, and can carry out tasks autonomously. The agent can run terminal commands, interact with the desktop and Chrome, and read, edit, or search files; the maker lists 33 built-in tools for commands, screen control, file management, and workflow automation. Users can create reusable skills using prompt instructions, YAML workflows, Python tools, or combinations. Listed AI providers are Google Gemini, OpenAI, Anthropic, and OpenRouter, with per-role model selection. Connections include LAN, Tailscale, and Cloudflare Tunnel, with automatic path selection and fallback. The site says phone-to-desktop traffic is end-to-end encrypted with DTLS and does not pass through third-party servers. Risky commands can be blocked, sandboxed, or held for approval, and session logs record prompts, commands, tools, and results. Android is available via Google Play; iOS is coming soon. Desktop downloads for Windows, macOS, and Linux require Python 3.12 or later.

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.