I am experimenting with a test site in Crafter CMS, and I used the Headless_blog blueprint. When I use a REST request to query the content for a Post component, the HTML tags for the body RTF field are all stripped out.
So I added a new RTF field and configured it exactly the same way and the HTML tags are not stripped out of that field.
What is it that is causing the HTML tags to be stripped out of the body? I don't see anything obvious. The RTF field is not very useful if it is converted to plain text. I would just use a plain text field if that is what I wanted.
It's a bug in the Headless blueprints and we have filed a ticket to fix it: https://github.com/craftercms/craftercms/issues/1350.
The issue is caused because the blueprints are returning data directly from the Solr index where the HTML tags are stripped for text processing, when we should have been loading the item after finding it and returning that, and the HTML would have been intact.
The new field you added is working because the name doesn't end in _html so Solr will not strip the HTML tags. You should have the field end with _html so that when you display search results they can be properly highlighted without issues from markup (hence why we strip it).
You can find out more about the naming convention for fields in the official documentation: http://docs.craftercms.org/en/3.0/developers/content-modeling.html?highlight=_html#form-control-variable-names