I'm not committed to sticking with Jekyll at the moment. This is more of an attempt to figure out what it would take extract the most complete docsets available.
The result should be mostly usable by anything that understands Jekyll frontmatter. mkDocs is one idea I had on my list. VitePress, Docusaurus, and Starlight are also worth considering.
One strike against mkDocs is that it is Python-based and the others are JavaScript. If we're going to move outside of the Ruby ecosystem, it might make the most sense to go with the web-native language.