MD Format Conversion Tool

Problem context

Teams often move documents between Markdown, PDF, DOCX, HTML, and TXT. Switching tools manually, verifying dependencies, and checking output quality can easily miss encoding problems, broken heading levels, or unsafe markup. This skill consolidates those workflows into a document format conversion engine that is useful for repeatable or batch conversions.

How it works

It supports eight conversion paths: MD → PDF, MD → DOCX, MD → HTML, MD → TXT, and the reverse paths PDF → MD, DOCX → MD, HTML → MD, TXT → MD. Different paths use different parsing or rendering libraries, for example python-docx for DOCX conversion, markdownify for HTML to Markdown, and pdfminer.six for PDF extraction.

When scripts/common/security.py exists, a security module is enabled automatically: path sandboxing restricts input and output directories, a file size limit defaults to 50MB, HTML/Markdown sanitization removes dangerous markup, and error messages sanitize paths, usernames, and IP addresses. If that file is absent, checks are skipped for backward compatibility.

When the output is Markdown and scripts/common/quality.py exists, quality validation runs: it detects garbled PDF text layers, compares input and output word counts, and checks heading-level continuity. Results are saved as a same-named .quality.json file, with a score printed to the terminal.

Scope and caveats

It is best suited for documents with an intact text layer. It does not include OCR, so scanned PDFs must be recognized externally first. Chinese garbling commonly comes from non-UTF-8 source files or damaged PDF text layers. For batch failures, check file type, size, and required dependencies first.