Crux parses Web pages to identify the crux of an article, minus all the fluff. - View it on GitHub