Someone built a site called claude-blog.md that is, quite literally, just a raw Markdown file rendered directly in the browser. No static site generator, no serverless hydration layer, no build steps, and no dynamic client-side router running three separate JS workers. Just text and a clean parser.
It brings back memories of the early web’s best impulse: fetching a raw document across a wire, reading it, and moving on with your day.
The Build Step That Wasn’t
For the last decade, personal blogging software went down a strange path. What used to be a index.html file on a university server somehow turned into a full-blown software project. You don’t just write a post anymore; you configure Next.js, manage Node dependencies, fix broken Tailwind builds, and tweak GraphQL queries just to publish three paragraphs about a movie you saw on Tuesday.
Projects like claude-blog.md strip that entire apparatus away. The source document is the page.