Looking at how to export a notebook to HTML with numbered headers, which was possible until nbconvert 5.x using the toc2 extension, I see now many unanswered(*) Github, Discourse and StackOverflow questions. See four examples below. In the meantime, did some alternate path emerge? Just to get some background: is it even possible (from a Jupyter components architecture point of view) to get nbconvert to interact with the JupyterLab Table of Contents feature that is built-in since JupterLab 3.0?
(*) I also include in the “unanswered” categories all the answers which suggest downgrading nbconvert to before nbconvert 6.0 release, which seems now irrelevant 4 years after.