Skip to main content
Glama
servo-mcp
by d11r

servo-mcp

A desktop MCP server that gives AI agents the ability to see and control your desktop.

servo-mcp.com - Installation instructions and documentation

What is Servo?

Servo is an MCP (Model Context Protocol) server that enables Claude Code and other AI agents to:

  • Take screenshots of your screen

  • Click buttons and interact with UI elements

  • Type text and press keyboard shortcuts

  • Scroll and navigate applications

  • Verify that changes actually work

Key features:

  • 100% local - no cloud, no telemetry, no data sharing

  • Works on macOS and Windows

  • Pure Node.js with native platform APIs (no Electron, no external dependencies)

  • Simple npm install

Related MCP server: open-computer-use

Installation

Install via npm (macOS and Windows):

npm install -g servo-mcp
npx servo-mcp --setup

macOS Permissions

Grant permissions to your terminal app (Terminal, iTerm, VS Code, Cursor, etc.) in System Settings > Privacy & Security:

  • Accessibility - for mouse clicks and keyboard input

  • Screen Recording - for taking screenshots

Child processes like servo-mcp inherit permissions from the parent terminal app.

Windows

No special permissions required. Windows may prompt for access when servo-mcp runs.

Setup with Claude Code

After running npx servo-mcp --setup, your ~/.claude.json will be configured:

{
  "mcpServers": {
    "servo": {
      "command": "npx",
      "args": ["servo-mcp"]
    }
  }
}

Restart Claude Code and test with: "take a screenshot"

Available Tools

Tool

Description

screenshot

Capture screen

click

Click at x,y coordinates

type_text

Type text at cursor

key_press

Press key combo (e.g., Cmd+S)

scroll

Scroll up/down/left/right

move_mouse

Move cursor to x,y

get_mouse_position

Get cursor position

focus_app

Bring app to foreground

open_app

Launch application

list_windows

List open windows

wait

Wait milliseconds

request_permissions

Open System Preferences

Development

pnpm install
pnpm dev:mcp          # Run MCP server in dev mode
pnpm dev:web          # Run website

# Build
pnpm build:mcp        # Bundle with esbuild

Architecture

Servo is a pure Node.js MCP server that uses:

  • Native platform APIs for automation (no Electron)

  • macOS: screencapture, AppleScript, Python/Quartz

  • Windows: PowerShell, .NET, user32.dll

License

MIT License

Author

Created by Dragos Strugar

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      A
      maintenance
      A local autonomous AI agent that watches your screen, understands the visual layout, and executes native OS commands (clicking, typing) without cloud APIs.
      39
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Lets AI agents see and control desktop applications through the accessibility layer, enabling clicking, typing, scrolling, dragging, and window/app management across macOS, Windows, and Linux entirely on the local machine.
      20
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables AI agents to observe and control the user's live desktop by listing apps, reading accessibility trees, and performing clicks, typing, scrolling, dragging, and other input actions with per-action approvals and local audit archives.
      1
      MIT