convez.ai
MCP ServerAutomaçãoAvançado
wizardofweb125-lab

MCP Linux Desktop

wizardofweb125-lab/mcp-linux-desktop

Controle completo de desktop Linux para agentes IA via MCP

Ver no GitHub
InicialPluginScore · de 0 a 100

Servidor MCP que oferece controle total da área de trabalho Linux, incluindo gerenciamento de janelas, entrada de mouse/teclado, clipboard, áudio, screenshots, OCR e árvore de acessibilidade. Funciona com qualquer agente IA compatível com MCP para automação de tarefas visuais e do sistema.

Instalação

$npx convez-ai wizardofweb125-lab/mcp-linux-desktop

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
1
Forks
0
Acompanhando
0
Saúde
42%
Último commit
há 5 meses
Versão
—
Issues abertas
0
Contribuidores
1

Casos de uso

  • Automação de fluxos de trabalho na GUI de aplicações Linux
  • Captura de tela e extração de texto com OCR para análise
  • Controle de janelas, mouse e teclado em scripts AI-driven
  • Integração com agentes para tarefas que exigem interação visual

Veredito

Ferramenta robusta e bem-estruturada para automação de desktop em Linux. Cobre praticamente todo espectro de interação GUI (janelas, entrada, captura, OCR) — excelente para agentes que precisam operar ambientes gráficos reais, mas requer setup de dependências e Linux com X11/GNOME.

README

A tradução para o português deste README ainda está na fila.

MCP server for full Linux desktop control — windows, mouse, keyboard, clipboard, audio, screenshots, OCR, notifications, accessibility tree, and system management. Works with any MCP-compatible AI agent.

Setup

cd mcp-linux-desktop
npm install

Add to your MCP config:

{
  "mcpServers": {
    "linux-desktop": {
      "command": "node",
      "args": ["/path/to/mcp-linux-desktop/server.js"]
    }
  }
}

Tools

Window Management

ToolDescription
window_listList all open windows with IDs, names, geometry
window_focusFocus a window by ID or name
window_closeClose a window
window_minimizeMinimize a window
window_resizeResize and/or move a window

Mouse & Keyboard

ToolDescription
mouse_moveMove cursor to coordinates
mouse_clickClick (left/right/middle, single/double)
mouse_scrollScroll in any direction
mouse_dragDrag from one position to another
mouse_locationGet current cursor position
keyboard_typeType text
keyboard_keyPress key combo (e.g. ctrl+a, alt+Tab)

Apps & Files

ToolDescription
app_launchLaunch an application by name or command
app_listList running graphical applications
file_openOpen a file or URL with its default app

Clipboard

ToolDescription
clipboard_readRead clipboard contents
clipboard_writeWrite text to clipboard

Audio

ToolDescription
volume_getGet current volume and mute status
volume_setSet volume (0-100) or toggle mute

Screen

ToolDescription
screenshotScreenshot the screen or a region
screen_observeScreenshot + screen metadata (resolution, active window, mouse position)
ocr_screenScreenshot a region and extract text via OCR

Accessibility

ToolDescription
ui_elementsRead interactive UI elements (buttons, fields, menus) with screen coordinates via AT-SPI

System

ToolDescription
system_infoCPU, RAM, disk, uptime
process_listList processes (optional name filter)
process_killKill a process by PID or name
notifySend a desktop notification
brightness_getGet screen brightness
dbus_callCall a DBus method (GNOME shell, systemd, etc.)

System dependencies

Required:

  • Node.js 18+
  • xdotool — window/mouse/keyboard control
  • xclip — clipboard (X11)
  • Python 3 + Pillow — screenshots

Optional:

  • wl-paste / wl-copy — clipboard on Wayland
  • tesseract-ocr + tesseract-ocr-fra — OCR
  • notify-send — desktop notifications
  • wpctl or pactl — audio control
  • gdbus — DBus calls
  • PyGObject + AT-SPI — accessibility tree (ui_elements)

Install on Ubuntu/Debian:

sudo apt install xdotool xclip python3-pil tesseract-ocr tesseract-ocr-fra \
  libnotify-bin pipewire-pulse gir1.2-atspi-2.0

Environment variables

VariableDefaultDescription
DISPLAY:1X11 display to target
Ver o original no GitHub

Bom para

Relacionados