Sitemap & Crawlers

Three endpoints, all generated from the same in-memory page list, all rebuilt automatically when your docs change. You don't run a separate command to produce any of them.

sitemap.xml ​

curl http://localhost:5000/sitemap.xml

<?xml version="1.0" encoding="UTF-8"?>
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
  <url><loc>/</loc><priority>1.0</priority></url>
  <url><loc>/getting-started/getting-started</loc><lastmod>2026-06-20</lastmod><priority>0.8</priority></url>
  ...
</urlset>

The home page gets priority 1.0; everything else gets 0.8. <lastmod> comes straight from each Markdown file's last-write timestamp on disk: no Git history lookup, no separate metadata to keep in sync. Edit the file, the sitemap reflects it on the next rebuild.

robots.txt ​

User-agent: *
Allow: /
Sitemap: https://your-host/sitemap.xml

Because this is generated dynamically for every request rather than being served as a static file, the Sitemap: line always points to the specific host that handled the request. Whether you are working locally at http://localhost:5000 or browsing your live site in production, it points to the correct location automatically, so you never have to manage multiple versions of the file.

llms.txt ​

# Bark

- [Getting Started](https://your-host/getting-started/getting-started): Get Bark running locally in under a minute
- [Configuration](https://your-host/getting-started/configuration): appsettings.json options, docs/config.json, and theming
...

This file lists every page with its title, URL, and a description. It helps AI agents read your site. These tools often struggle to ignore your site's navigation and layout when parsing HTML, so this file gives them a clean index instead.

None of these three files require any setup. If you put a page in your docs/ folder, Bark includes it in all of them.