Deploy and Host Heyo Docs on Railway

Built for an LLM agent to deploy and use. The deploy form has no fields to fill: the site URL is read from the Railway domain at boot, so the sitemap, llms.txt, RSS feed and canonical tags point at the real address from the first request. Heyo Docs is a themeable documentation site (MDX pages, navigation, search, OpenAPI reference pages) that also publishes every page as plain Markdown and an llms.txt index for AI agents.

About Hosting Heyo Docs

This template runs one service, heyo-docs (a single Node process serving the prerendered site, the Markdown endpoints, llms.txt, sitemap.xml and rss.xml). The starter site and its content are compiled into the image, so there is no database and no volume to manage.

The URL is rewritten at boot. Heyo Docs bakes its site URL into the sitemap, llms.txt, RSS feed and canonical tags at build time, and Railway only assigns a domain after the first deploy. This template's image is built with a placeholder URL and rewrites it to https:// plus the service's Railway domain when the container starts. To serve a custom domain, add a SITE_URL variable (for example https://docs.example.com) and redeploy.

Publishing your own pages. Content is compiled at build time. Fork the wrapper repository, replace the starter content/index.mdx and add pages as .mdx files, then point the Railway service at your fork.

Agent contract

Public URL: https://RAILWAY_PUBLIC_DOMAIN (the generated domain of the heyo-docs service)
Private URL: http://heyo-docs.railway.internal:8080
Credentials: none, this is a public read-only site (no template variables to read)
Auth: none; every page is public. Markdown for any page is at /__heyo-docs/markdown/PAGE.md (/PAGE.md redirects there), and /llms.txt indexes them
Health: GET / returns 200

# needs: SITE_URL (public https URL of the heyo-docs service); also curl, sed
# write-auth: none - read-only public docs site, no credential exists and no write endpoint is usable (AI chat is unconfigured, the OpenAPI proxy refuses unconfigured specs)
curl -s -o /dev/null -w "health: %{http_code}\n" "$SITE_URL/"
curl -s "$SITE_URL/llms.txt" | head -3
curl -s "$SITE_URL/sitemap.xml" | tr '\074' '\n' | sed -n 's|^loc.https://|sitemap loc: https://|p'
curl -s -o /dev/null -w "markdown page: %{http_code} %{content_type}\n" "$SITE_URL/__heyo-docs/markdown/index.md"
curl -s "$SITE_URL/" | grep -o 'rel="canonical" href="[^"]*"' | head -1 | sed 's|rel="canonical" href="|canonical: |; s|"$||'

A fresh deploy prints health: 200, the first lines of llms.txt (# Heyo Docs), a sitemap loc: line with the service's own https:// domain, markdown page: 200 text/markdown; charset=utf-8 and a canonical: line with the same domain.

Why Deploy Heyo Docs on Railway?

Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.

By deploying Heyo Docs on Railway, you are one step closer to supporting a complete full-stack application with minimal burden. Host your servers, databases, AI agents, and more on Railway.

Agent-ready - no deployer fields, a contract block and one script that proves the site, llms.txt, sitemap, Markdown endpoint and canonical URL.
Correct URLs from the first request - sitemap, RSS and canonical tags carry your Railway domain, not a placeholder.
Runs unprivileged - the container runs as the node user and binds IPv4 and IPv6 for Railway's private network.
Pinned - the image is pinned by digest.
No database - nothing to back up, nothing that can drift between redeploys.

Common Use Cases

Product and API documentation - a themeable docs site with search and OpenAPI reference pages.
Docs that AI agents can read - every page is also served as Markdown and indexed in llms.txt.
Internal handbook or runbook site - MDX pages in a Git repository, rebuilt on every push.
Open source project site - a public site with sitemap and RSS for search engines and subscribers.

Dependencies for Heyo Docs Hosting

Nothing else: content is compiled into the image. A custom domain is optional.

Deploy Heyo Docs | (Just Updated) Documentation Site With llms.txt and Markdown for AI Agents