I wanted to ask if this idea might be considered in scope for doc8: Parsing Python source code (*.py), extracting the docstrings, and validating those as RST.
i.e. Something equivalent to the functionality my plugin https://github.com/peterjc/flake8-rst-docstrings brings to flake8. Internally this takes the AST of a Python file, and passes the docstrings to docutils for validation via restructuredtext_lint.