Yavchn – Yet Another Vibe-Coded Hacker News Wrapper
Three-pane layout kills tab-hopping between HN, article, and comments.

Paginated reader mode with EPUB export when Firefox Reader only scrolls.
Heavy article readers who prefer paginated reading over infinite scroll
Mercury Reader · Firefox Reader View · Pocket
This has been my first time developing a browser extension.
In a perfect world with semantic clean markup everywhere the process of reformatting a website into a reader interface would be quite easy, but real world HTML can be really messy. Simply passing the DOM to @mozilla/readability (the engine Firefox uses for its reader view) was not sufficient. While it does a good job at identifying the main article, normalizing HTML and removing junk like widgets within it needs additional heuristics. So I built a custom pipeline of pre- and postprocessing steps around it and tested on as many websites as possible.
An interesting challenge has been how to get around CORS and CSP limitations by the host page. Some of the solutions I came up with are outlined on my blog post linked below.
Features I'm considering to add in a future release:
- one-click send to device for Kindle/Kobo etc. - High quality Text-to-Speech - Firefox/Edge/Safari release
github: https://github.com/fachkamera/booklike blog post: https://mayermarkus.net/building-booklike
Three-pane layout kills tab-hopping between HN, article, and comments.
Nice newspaper aesthetic, but HN readers are a crowded category with no technical edge.
Finally fixes the 'read later' backlog by forcing content onto actual e-ink devices.
Split-pane HN reader works in browser, but Nitter and dozens of HN clients exist.
Yet another HN reader in a space with dozens of established options.
Reader view plus AI summaries when Mercury and JinaAI already exist.