Show HN: Cpdown – Copy any webpage/YouTube subtitle as clean Markdown(LLM-ready)
github.comTL;DR: I built a Chrome extension that copies webpage content or YouTube subtitles as clean, clutter-free Markdown with one click (or a shortcut). It even shows the token count, making it super handy for LLM prompts!
Hi, HN!
I often copy information from the web into my notes (Obsidian) or to feed context to LLMs for summaries/translations. So I built cpdown, a browser extension for this.
cpdown lets you convert any webpage/youtube subtitle into clean Markdown and copy it to your clipboard with a single click (or a keyboard shortcut).
Here are the key features:
* Intelligent Content Extraction: Uses Mozilla's Readability or Defuddle (by the Obsidian devs!) to strip away ads, sidebars, and other noise, focusing only on the main article content. * Clean Markdown Conversion: The extracted HTML is then neatly converted to Markdown using the Turndown library. * Token Count for LLMs: It calculates and displays the token count of the copied text using tiktoken. This is super handy for knowing if you're within limits before pasting into LLMs like ChatGPT or Claude. * YouTube Transcript Copying: On YouTube video pages, it can copy the full transcript in Markdown format, automatically adding the video title as an H1 header. * Customizable Options: You can choose your preferred content extractor (Readability/Defuddle), opt to wrap copied content in a code block, and more.
cpdown is completely free and open-source, built with WXT, React, and TypeScript.
You can grab it from the Chrome Web Store or check out the source code on GitHub. I'd love to hear your feedback, feature requests, or any suggestions!
* Chrome Web Store: https://chromewebstore.google.com/detail/cpdown/knnaflplggjd... * GitHub: https://github.com/ysm-dev/cpdown
Thanks for checking it out!
Seems useless and also it's littered with dependencies.