folderdocs
Markdown documentation, ready for search engines

SEO and limitations

What the build does for search engines

Every Markdown file becomes its own complete HTML page, so crawlers see your content without running any JavaScript. For each page folderdocs generates:

  • a clean, stable URL (/guide/markdown-syntax/);
  • a <title> (Page · Site) and a meta description, from the front matter or the first paragraph;
  • a canonical link and Open Graph / Twitter tags for good link previews;
  • breadcrumb structured data (JSON-LD) matching the visible breadcrumbs;
  • hreflang links between translations (see Languages);
  • a sitemap.xml and robots.txt, and a 404.html marked noindex;
  • descriptive image alt text, heading anchors and a "skip to content" link.

The canonical links, the sitemap and og:url need your site URL, which GitHub Pages deployments detect automatically.

What you should do

  • Give every page a clear title and, for the important ones, a hand-written description:.
  • Write real alt text for images.
  • Set site.url if you use a custom domain.
  • Submit https://your-site/sitemap.xml in Google Search Console and Bing Webmaster Tools.
  • Link to your documentation from your README, your project's website and other places people visit.
!Important

Good URLs and metadata make your pages indexable; they don't decide where you rank. Ranking depends on the quality of your content, links from other sites and how long the site has existed. Expect it to take weeks for new pages to show up.

Other limitations

  • It needs a build step. Previewing and publishing use Node (npm run dev, npm run build); GitHub Pages runs it for you. Your server only receives static files.
  • Search is plain text. It runs in the browser on an index generated at build time, with no fuzzy matching or semantic ranking. It does need JavaScript; everything else works without it.
  • No documentation versioning. Languages are supported through one folder per language.
  • HTML in .md files is not filtered. Only publish content you trust.
  • The search index holds the text of every page (the first 20,000 characters of each). That is fine for hundreds of pages; for a very large site, use a dedicated search service.