Brosh
Captures scrolling screenshots of webpages using Playwright, offering intelligent section identification, multiple output formats, and Model Context Protocol (MCP) integration.
概要
Brosh is a powerful Python-based tool leveraging Playwright's async API to capture comprehensive screenshots of webpages. It intelligently navigates and scrolls through content, identifying visible sections for descriptive filenames, and supports various output formats including animated PNGs. Beyond visual capture, it can extract HTML and Markdown text from visible elements. Designed for versatility, Brosh operates as a command-line interface, offers a Python API for programmatic control, and integrates as an MCP server, enabling AI tools like Claude to access and analyze web content, making it invaluable for documentation, QA testing, and content archival.
主な機能
- Multiple Output Formats including PNG, JPG, and animated PNG (APNG)
- MCP Server for seamless integration with AI tools
- Smart Section Detection for descriptive filenames based on visible content
- 0 GitHub stars
- HTML and Text Extraction from visible page elements
- Async Playwright Integration for fast browser automation
ユースケース
- Performing visual regression testing and bug reporting for QA teams
- Capturing long technical documentation or API references for archival or sharing
- Providing visual and textual web content context to large language models via MCP