What problem does it solve? Fetching web content as raw HTML wastes context window tokens and produces noisy output full of markup, making it hard to read, summarize, or analyze web pages efficiently. ## Core Features & Use Cases - URL-to-Markdown Conversion: Fetches any URL and returns clean Markdown with metadata (title, method, duration) via the markdown.new service, reducing token usage by roughly 80% compared to raw HTML. - Three Conversion Methods: Choose auto (Markdown-first with fallback), ai (Cloudflare Workers AI), or browser (full rendering for JavaScript-heavy sites like React or Vue apps). - Flexible Output: Print to stdout, save to a file with --output, or retain image references with --retain-images. - Use Case: When a user shares a link to a blog post or documentation page and asks for a summary, fetch the page as Markdown first, then summarize the clean content without HTML noise. ## Quick Start Fetch the content of the URL the user provided and convert it to Markdown, using browser rendering if the site is JavaScript-heavy.
