Cristal is not crawlable by web indexers

Reproduction steps

1. deploy a cristal instance on the internet (e.g., cristal.demo.xwiki.com) with several pages
2. on a search engine, search for:
   1. url:cristal.demo.xwiki.com
   2. url:cristal.demo.xwiki.com some content from the main page (e.g., site:cristal.demo.xwiki.com Cristal is a new tool to organize your knowledge)
   3. url:cristal.demo.xwiki.com some content from a sub page (e.g., site:cristal.demo.xwiki.com The sandbox is a part of your wiki that you can freely modify.)

Expected

Each search engine query returns the relevant pages

Actual

Only queries targeting the main page are returned in the results.
Meaning that other pages are not correcrly indexed.
Hypothesis:

• some metadata expected by search engine are missing
• the document tree is loaded too late, ignored by search engines, and consequently not used to find new pages to index