BrowserLoop

Captures web page screenshots using Playwright with support for full-page, element-specific, and viewport screenshots in PNG, JPEG

mattiasw 23 ↓ 11k
Claude CodeClaude DesktopGeneric
View source ↗

This MCP server provides web screenshot capabilities using Playwright, enabling AI assistants to capture full-page screenshots, element-specific screenshots, and viewport screenshots of web pages with support for PNG, JPEG, and WebP formats. Built by Mattias Wallander with TypeScript and featuring comprehensive cookie authentication support, domain filtering for security, image format conversion with Sharp, and Docker containerization, it offers configurable quality settings, network idle waiting, timeout handling, and performance optimization through browser page pooling. The implementation includes robust error handling with categorized logging, memory cleanup for sensitive data, and NPX-friendly installation, making it valuable for web testing workflows, content verification, visual regression testing, and any use case requiring programmatic web page capture with authentication support.

Source

Repository: https://github.com/mattiasw/browserloop

Maintain BrowserLoop?

Let people know it's listed here — add the badge (live metrics, light/dark aware) or a plain link to your README or docs.

BrowserLoop on getagentictools
[![BrowserLoop on getagentictools](https://getagentictools.com/badge/mcp/mattiasw-browserloop.svg)](https://getagentictools.com/mcp/mattiasw-browserloop?ref=badge)