Sphinx could be an option here. It parses and render Databricks notebooks in the documentation. You might want to look into it to see if it fits your needs. However, it may not handle magic commands very well, and it assumes your notebooks are exported in .dbc format.
If your Databricks notebooks are heavily reliant on magic commands, this could not work well. In this case, you might need to consider creating a separate documentation process that doesn't rely on extracting docstrings from your notebooks.
