ScreenshotsMCP Documentation
Give your AI coding assistant eyes. Screenshot any URL, automate browsers, audit performance and SEO — all via MCP tools.
What is ScreenshotsMCP?
ScreenshotsMCP is a Model Context Protocol (MCP) server that gives AI coding assistants the ability to see and interact with websites. It works with Claude, Cursor, Windsurf, VS Code, and any MCP-compatible client.
46+ Tools in One Server
| Category | What it does |
|---|---|
| Screenshots | Capture any URL — desktop, mobile, tablet, dark mode, full page, PDF |
| Browser Sessions | Navigate, click, fill forms, hover, scroll, press keys — like a real user |
| Inspection | Extract text, HTML, accessibility trees, run JavaScript |
| Performance | Core Web Vitals (LCP, FCP, CLS, TTFB), full network waterfall |
| SEO | Meta tags, Open Graph, Twitter Cards, headings, structured data, alt text audit |
| Debugging | Console logs, network errors, cookies, localStorage/sessionStorage |
Quick Example
Once installed, just ask your AI assistant:
"Screenshot my site at https://myapp.com and tell me if the hero looks right"Your AI will take a screenshot, see your UI, and give you actionable feedback — no setup, no code.
Get Started
Quickstart
Install in your IDE in 60 seconds
Installation
Configure for Cursor, Windsurf, Claude Desktop, and VS Code
Authentication
Create and manage your API keys
REST API
Use the HTTP API directly from any language
For LLMs
Download the complete documentation as a single text file optimized for LLM consumption:
llms.txt — All 46+ tools, parameters, REST API, CLI, and installation instructions in one file.