MD Tune-Up · free and open source, MIT

Keep the notes your AI reads clean and consistent.

Your AI is only as consistent as the context you feed it, and most of the drift is mess on your side, not the model's. MD Tune-Up surveys your markdown notes, flags the inconsistencies, and fixes them with your okay. Think of it as an occasional tune-up for the files your AI reads.

What it checks

• Frontmatter. YAML that won't parse, and a key most of your files have that one file is missing.
• Names. A name field that no longer matches its filename.
• Links. Wiki-links and relative links that point at a file that moved or was never written.
• Descriptions. Empty, "TODO", or one that only repeats the filename.

It checks only the conventions your folder uses. A folder of plain markdown gets "nothing to tune up."

How a run goes

• It scans the folder and tells you what it found.
• It walks you through each item, one at a time: approve, skip, or quit.
• It backs up your files, then makes only the changes you approved.

Start with a dry run. It reports and changes nothing:

python3 md_tuneup.py ~/my-notes --dry-run

Install steps are in the README on GitHub.

• Free, one Python file, runs on your machine.
• No account, no network calls, no AI inside it.