Automate Meta Description Removal

I maintain a hand-coded HTML site with hundreds of pages that still carry legacy
<meta name="description" …> tags—sometimes duplicated, sometimes outdated. I now want every single description meta tag stripped out across the entire codebase while leaving all other markup, including title and keyword tags, untouched.

Here’s what I need from you:

• A repeatable script or small utility (Python, Bash, Node, or even a portable desktop tool) that scans a folder tree of .html /.htm files and removes every description meta tag it finds, even when there are multiple per file.
• The routine must preserve file encoding, indentation where possible, and absolutely must not touch any other tag or inline content.
• Simple usage instructions so I can run it myself whenever I add new pages—ideally a one-liner such as python clean_meta.py /path/to/site.
• A quick backup option (e.g., automatic copy to a /backup folder or a dry-run flag) so I can roll back if needed.
• A short report or console log listing the files it processed and how many tags it removed.

I’ll provide a ZIP of the site tree on project start. Please test on a handful of pages first, share the before/after diff, then proceed to the full batch once I confirm everything looks good.

If you’ve tackled similar HTML cleanup or have a reliable regex/DOM approach ready, let’s get this wrapped up quickly and cleanly.