Where to place robot.txt and sitemap.xml for wordpress installed in its own directory

My website is set up for wordpress to be installed in its own subdirectory, but the content is broadcast as it was at the root of the domain.
(This wordpress codex page details the configuration)

WordPress URL: https://www.example.com/wordpress
Website URL: https://www.example.com

So in this case the URLs are rewritten to NOT include / wordpress in the URL, it serves pages from (although the URLs of the image make include it: https://www.example.com/wordpress/wp-content/uploads / ...).

The URL of a page is: www.example.com/mypage/
DO NOT: www.example.com/wordpress/mypage/
but the link of the image is: www.example.com/wordpress/wp-content/uploads/2018/12/25/image1.jpg

Question:
Do I put robots.txt and sitemap.xml in the "root of the website" directory on my server (exemple.com)
or his wordpress directory (example.com/wordpress)?

I do not find this particular case in research.
Moreover, testing and errors do not help, because I get conflicting results from various sitemap validators and robots.

In addition, it seems that a plugin or part of WordPress insists on creating a virtual sitemap (sitemap.xml.gz) and robot (robots.txt), which further exacerbates my efforts to definitively determine the appropriate location to place these files for this common but non-standard installation.