to-markdown
Convert files to Markdown via markitdown — PDFs, Word, Excel, PowerPoint, images, HTML, and other formats. Triggers on 'convert to markdown', 'extract text from', 'import this document', 'turn this PDF into markdown'.
How do I install this agent skill?
npx skills add https://github.com/steveclarke/dotfiles --skill to-markdownIs this agent skill safe to install?
- Gen Agent Trust Hubpass
This skill facilitates document conversion to Markdown using Microsoft's MarkItDown library. While it uses trusted software and performs local operations, it possesses an indirect prompt injection surface as it processes untrusted external files and incorporates their content into the agent's context without sanitization.
- Socketpass
No alerts
- Snykpass
Risk: LOW · No issues
What does this agent skill do?
Convert files to Markdown using markitdown — Microsoft's utility that extracts and structures content from many file formats.
Supported Formats
| Format | Notes |
|---|---|
| Text extracted; table structure may be approximate | |
| Word (.docx) | Clean conversion including tables |
| Excel (.xlsx) | Sheets as markdown tables |
| PowerPoint (.pptx) | Slide text and structure |
| HTML | Cleaned readable content |
| Images | OCR (requires LLM vision for best results) |
| Audio | Transcription via SpeechRecognition |
| CSV / JSON / XML | Structured text |
| YouTube URLs | Transcript extraction |
| EPub | Chapter text |
Usage
# Convert to stdout
markitdown input.pdf 2>/dev/null
# Save to file
markitdown input.pdf -o output.md 2>/dev/null
# Or redirect
markitdown input.docx 2>/dev/null > output.md
Always use 2>/dev/null to suppress noisy font/parser warnings that don't affect output quality.
Workflow
-
Check if markitdown is installed:
which markitdown || echo "not installed" -
Install if missing (with all format support):
pip install 'markitdown[all]' -
Run the conversion with stderr suppressed:
markitdown "$INPUT_FILE" 2>/dev/null -
Handle the output based on user intent:
- Saving to the knowledge base → write to appropriate
.mdfile - Quick review → show in conversation
- Multiple files → loop and convert each
- Saving to the knowledge base → write to appropriate
Output Quality Notes
- PDFs: Text is extracted faithfully but table cells may land on separate lines (PDF doesn't encode table structure). If the user needs clean tables from a PDF, note this limitation.
- Word/Excel: Usually clean output with proper table formatting.
- Complex layouts: Multi-column PDFs or heavily formatted documents may have scrambled reading order.
- Scanned PDFs: Image-only PDFs produce no text without OCR/LLM vision integration.
Common Use Cases
Import a document into knowledge base:
markitdown report.pdf 2>/dev/null > knowledge/competitive/raw/report.md
Convert a Word doc someone sent you:
markitdown meeting-notes.docx 2>/dev/null > notes.md
Batch convert a directory:
for f in docs/*.pdf; do
markitdown "$f" 2>/dev/null > "${f%.pdf}.md"
done
Check what a PDF contains before deciding what to do with it:
markitdown document.pdf 2>/dev/null | head -50
How can the creator link this skill?
Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.
<a href="https://skillzs.dev/skills/steveclarke/dotfiles/to-markdown">View to-markdown on skillZs</a>