browser-auto-download_skill

Overview

This skill automates downloads from dynamic, client-rendered web pages by driving a headless or visible browser. It auto-detects the user platform, captures downloads triggered on page load, and falls back to navigation and button-click strategies for multi-step flows. It is optimized for cases where curl/wget fail due to JavaScript, redirects, or lazy-loaded content. Golang and Python integration points are supported for scripting and automation.

How this skill works

The skill launches a Playwright browser, opens the target URL, and listens for download events during page load. If no automatic download occurs, it searches for platform-specific links (e.g., desktop or Windows pages) and navigates there. As a final fallback it tries multiple CSS selectors to click download buttons, with optional scrolling and extended wait times to handle lazy content. Results are returned as JSON with file path, filename, size, and detected platform.

When to use it

• Sites that trigger downloads automatically when a page loads
• Multi-step flows that require navigating from homepage to a platform-specific download page
• Downloads generated or revealed by client-side JavaScript
• Interactive download flows that require clicking buttons or opening modals
• When curl/wget fail due to rendering, redirects, or dynamic content

Best practices

• Run with visible browser during initial troubleshooting to observe navigation and interactions
• Enable platform auto-detection for correct binary selection; disable if site misidentifies your system
• Use explicit selectors when auto-detection fails to ensure correct button clicks
• Allow extended wait times for large files or slow servers (up to a few minutes)
• Check write permissions and disk space in the configured download directory

Example use cases

• Downloading installer for a desktop app that only appears after clicking a "Desktop" link
• Capturing an automatic installer delivery triggered on first page load
• Automating large or slow downloads where the site uses JavaScript to generate the file URL
• Batch processing multiple product pages that each require different navigation steps
• Integrating into a CI script to fetch platform-specific tooling before builds