Open Source · Documentation

SEO and indexing

OrbitPage generates search and sharing metadata from saved page data. Configure it in Page and Publish.

OrbitPage generates search and sharing metadata from saved page data. Configure it in Page and Publish.

Set the public URL

Set these runtime values through Configuration:

PUBLIC_SITE_URL=https://page.example.com
PUBLIC_SITE_NAME=My page
SEO_INDEXING=true

Include any mount path in PUBLIC_SITE_URL. Without it, the server derives the URL from the request host and protocol; proxies may supply an internal address.

In Page, set the page name, browser title, description, profile image, and social links.

Generated metadata

Output	Source
HTML title and description	Saved page/profile fields
Canonical URL	Public URL and active public path
Open Graph and Twitter cards	Page metadata and image
Schema.org JSON-LD	Saved public profile/content
Robots meta	Indexing setting and route
No-JavaScript links	Public content for crawlers
Sitemap	Enabled public destinations and content-change dates

Dashboard, API, health, and unknown routes are excluded from indexing.

The no-JavaScript fallback includes configured policy links. Local privacy and cookie pages include a heading, a home link and escaped hosted policy text; embedded providers expose safe policy links without executing their scripts.

For staging or a private instance, set SEO_INDEXING=false. It emits noindex, nofollow, noarchive and restrictive robots rules. This requests that crawlers stay away; it does not protect a page from visitors.

Sitemap

Open Publish → Sitemap and select Generate sitemap. The XML at /sitemap.xml follows current saved data; regenerate to record a new generation time.

It includes the main page, published additional pages, the enabled menu, and local policy pages when configured. Demo mode also includes About.

The workspace shows the URL, route count, generation time, and copy/open controls. Sitemap settings are included in Discovery files exports.

Text files and machine-readable pages

Publish → TXT edits:

robots.txt;
llms.txt (with /llm.txt as an alias);
humans.txt, security.txt, and ai.txt;
up to 20 custom /name.txt or /.well-known/name.txt paths.

Custom paths are normalized to lowercase and cannot collide with reserved names or traverse directories. Text files are included in Discovery files exports.

Enable Machine-readable access under Page → Online presence to expose generated LLM discovery and Markdown page representations:

Request	Enabled	Disabled
Public URL with `Accept: text/markdown`	Saved public content as Markdown, with `Vary: Accept`	`406`
Generated AI discovery files	Served	`404`
Normal public URL	HTML with JSON-LD	HTML with JSON-LD

These responses use local data and do not call an AI provider. SQLite stores daily counts by format and public path, without IP, user-agent, referrer, cookie, or page-content data. GET /api/analytics/machine-readable returns the latest 30 days to users with analytics:read.

Keep public links as real anchors and avoid blocking assets or public media in robots rules. See Publishing for sharing tools.