← ClaudeAtlas

markdown-converterlisted

Convert documents and files to Markdown using markitdown. Use when converting PDF, Word (.docx), PowerPoint (.pptx), Excel (.xlsx, .xls), HTML, CSV, JSON, XML, images (with EXIF/OCR), audio (with transcription), ZIP archives, YouTube URLs, Outlook (.msg), or EPubs to Markdown format for LLM processing or text analysis. This is a ONE-OFF conversion of a file or URL you already have. For fetching an external source ONCE and keeping it as a pinned, provenance-stamped artifact that agents re-read instead of re-fetching -- reproducible, offline, versioned -- use source-snapshot instead.
kevin-burns/claude-skills · ★ 8 · Data & Documents · score 81
Install: claude install-skill kevin-burns/claude-skills
# Markdown Converter Convert files to Markdown using `uvx markitdown` — no manual install required. ## Choosing the invocation (read this first) markitdown ships format support as **optional extras**. A bare `markitdown` install only handles HTML, plain text, CSV, JSON, and XML. PDF, Office files, audio, YouTube, and Outlook each need an extra, so you must request them in the `uvx` call — otherwise the run fails with a missing-dependency error. ```bash # All formats, simplest — use this unless you have a reason not to: uvx 'markitdown[all]' input.pdf # Lighter: pull only the extra you need (faster first run, smaller cache). # With a single extra the command name no longer matches the package spec, # so name it explicitly with --from: uvx --from 'markitdown[pdf]' markitdown input.pdf uvx --from 'markitdown[docx]' markitdown report.docx -o report.md ``` Available extras: `pptx`, `docx`, `xlsx`, `xls`, `pdf`, `outlook`, `audio-transcription`, `youtube-transcription`, `az-doc-intel`, `az-content-understanding`, and `all`. ## Basic Usage ```bash # To stdout uvx 'markitdown[all]' input.pdf # To a file uvx 'markitdown[all]' input.pdf -o output.md uvx 'markitdown[all]' input.docx > output.md # From stdin — give a type hint since there's no filename to sniff cat input.pdf | uvx 'markitdown[all]' -x .pdf > output.md ``` ## Supported Formats - **Documents**: PDF, Word (.docx), PowerPoint (.pptx), Excel (.xlsx, .xls) - **Web/Data**: HTML, CSV, JSON, XML - **Media**: Images (