PyMuPDF PDF Parser Clawdbot Skill

**Quick judgment:** A fast, lightweight local PDF parser built on PyMuPDF, best suited for mostly text-based PDFs and scenarios where speed and simplicity matter more than perfect structure or robustness. **What it does well** - Converts PDFs to Markdown and/or JSON quickly using local PyMuPDF. - Optionally extracts images and a rough, line-based table representation. - Organizes outputs into per-document directories, which is convenient for pipelines. - Works well as a fallback when heavier OCR/structured parsers are unavailable. **Key risks / limitations** - PyMuPDF is "less robust" for complex PDFs (heavy layout, multi-column, scanned/OCR-needed documents), so content/structure may be incomplete or misaligned. - Table extraction is explicitly rough and line-based; not suitable where accurate tabular reconstruction is critical. - Requires PyMuPDF and compatible system libraries; import/libstdc++ issues are called out in the docs and can block usage if not resolved. **Recommended scenarios** - Use for fast, local extraction from standard, mostly text-based PDFs where minor structural inaccuracies are acceptable. - Use as a backup parser in a multi-tool pipeline when more sophisticated parsers fail or are not installed. - Avoid relying on it as the sole source of truth for complex layouts, high-stakes table data, or scanned PDFs needing OCR; in those cases, prefer a heavier OCR/structured parser and treat this skill as a secondary option.