screenshot-mcp
A cross-platform MCP server that allows AI agents to capture screenshots of specific windows, displays, or regions for native application testing. It provides tools to list active windows and monitors, enabling precise visual verification and interaction during automated workflows.
README
screenshot-mcp
A cross-platform MCP (Model Context Protocol) server for capturing screenshots. Designed for agent-based native application testing.
Also available as a Claude Code plugin with screenshot testing guidance.
Features
- List Windows: Get all visible windows with IDs, titles, and bounds
- List Displays: Get all available monitors/displays
- Screenshot Window: Capture a specific window by ID or title
- Screenshot Screen: Capture an entire display
- Screenshot Region: Capture a specific screen region
Installation
No installation required - use directly with npx or bunx.
Usage
Configure in Claude Code
Add to your MCP settings (~/.claude/settings.json):
{
"mcpServers": {
"screenshot": {
"command": "bunx",
"args": ["screenshot-mcp"]
}
}
}
Or with npx:
{
"mcpServers": {
"screenshot": {
"command": "npx",
"args": ["-y", "screenshot-mcp"]
}
}
}
Run Standalone
# Using bunx (recommended)
bunx screenshot-mcp
# Using npx
npx -y screenshot-mcp
MCP Tools
list_windows
List all visible windows.
// Response
[
{ "id": "1234", "title": "VS Code", "app": "Code", "bounds": { "x": 0, "y": 25, "width": 1920, "height": 1055 } }
]
list_displays
List all available displays.
// Response
[
{ "id": 1, "name": "Built-in Retina Display", "primary": true, "bounds": { "x": 0, "y": 0, "width": 2560, "height": 1600 } }
]
screenshot_window
Capture a specific window.
// Parameters
{ "window_id": "1234" }
// or
{ "window_title": "VS Code" }
// optional: save to file
{ "window_id": "1234", "save_dir": "/path/to/screenshots" }
screenshot_screen
Capture the entire screen.
// Parameters (all optional)
{ "display_id": 1, "save_dir": "/path/to/screenshots" }
screenshot_region
Capture a specific region.
// Parameters
{ "x": 100, "y": 100, "width": 800, "height": 600 }
// optional: save to file
{ "x": 100, "y": 100, "width": 800, "height": 600, "save_dir": "/path/to/screenshots" }
Platform Support
| Platform | Status | Requirements |
|---|---|---|
| macOS | ✅ Supported | Screen Recording permission |
| Windows | ✅ Supported | PowerShell (built-in) |
| Linux (X11) | ✅ Supported | xdotool or wmctrl for windows, gnome-screenshot/scrot/import for capture |
| Linux (Wayland) | ⚠️ Partial | GNOME Shell only, grim for screenshots |
Configuration (Plugin Settings)
Create .claude/screenshot.local.md in your project to set a default save directory:
---
default_save_dir: /Users/yourname/Desktop/screenshots
---
When this file exists, Claude will automatically use this directory for saving screenshots.
Note: Add .claude/*.local.md to your .gitignore.
Requirements
- Runtime: Bun v1.0+ or Node.js 18+
macOS
- Screen Recording permission (System Settings > Privacy & Security > Screen Recording)
Windows
- PowerShell (built-in, no extra installation needed)
Linux (X11)
- Window listing:
xdotool(recommended) orwmctrl - Screenshots:
gnome-screenshot,scrot, orimport(ImageMagick) - Display info:
xrandr
# Ubuntu/Debian
sudo apt install xdotool scrot
# Fedora
sudo dnf install xdotool scrot
# Arch
sudo pacman -S xdotool scrot
Linux (Wayland)
- Window listing: Only GNOME Shell supported (via gdbus)
- Screenshots:
gnome-screenshotorgrim - Display info:
wlr-randror falls back to xrandr
As Claude Code Plugin
Install as a plugin to get the screenshot-testing skill:
# Install from npm
npm install -g screenshot-mcp
claude plugin add screenshot-mcp
# Or use local development
claude --plugin-dir /path/to/screenshot-mcp
The skill provides guidance on:
- Effective screenshot testing workflows
- Comparison testing strategies
- Multi-display testing
- Electron app testing examples
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.