How it works

A quick tour of what happens between your repository and the website your readers see.

Reading the repo

Your repository must be public so we can read it (contact us if you need a private one). When someone opens your project, we call the Github API to fetch:

  • the repo metadata (name, description, homepage, default branch),
  • the full file tree of your default branch (master or main),
  • the specific markdown, config, logo and blog files we need.

From those we detect everything described in Conventions, apply any Configuration, and render your markdown to HTML.

Caching

To stay fast, respect Github's rate limits, and serve many projects at once, we cache each project after building it. This means a change you push can take up to the plan's refresh window to appear:

  • Free: up to 24h.
  • Paid: faster refresh (down to near-immediate on higher tiers). See pricing.

The sidebar navigation is generated from the headings of your rendered documentation. Your ## and ### headings become the sections and sub-sections, each linked by an anchor, so readers can jump around and share deep links. Multi-file docs are merged first, then parsed as one.

Full-text search

Every page and heading is indexed so readers can search your documentation from the sidebar (press / to focus the box). Results link straight to the matching section.

Found a typo? Edit this page on Github