MarkItDown Skill

OpenClaw agent skill for converting documents to Markdown. Documentation and utilities for Microsoft's MarkItDown library. Supports PDF, Word, PowerPoint, Excel, images (OCR), audio (transcription), HTML, YouTube.

This skill appears to do what it says: documentation + a small batch conversion helper that relies on the markitdown pip package. Before installing: 1) Verify you're installing the official 'markitdown' package on PyPI (typosquatting is a general pip risk). 2) Install in a Python virtual environment to avoid affecting system packages (python -m venv .venv; source .venv/bin/activate). 3) Only install the extras you need (avoid 'markitdown[all]' if you only need PDFs) to reduce dependency surface. 4) If you enable LLM features (--llm-model) or Azure DocInt, know that those will make network calls and require credentials (OPENAI API key, AZURE_DOCUMENT_INTELLIGENCE_KEY); provide those only to trusted code and review usage. 5) If you need higher assurance, inspect the upstream markitdown PyPI package and its GitHub repo to confirm maintainership and release history before installing.

Type: OpenClaw Skill Name: markitdown-skill Version: 1.0.1 The OpenClaw AgentSkills skill bundle 'markitdown-skill' is classified as benign. Its primary purpose is to convert various document types to Markdown using the legitimate 'markitdown' library. The `scripts/batch_convert.py` script and all documentation (SKILL.md, POST_INSTALL.md, USAGE-GUIDE.md, reference.md) consistently support this functionality, detailing file processing, optional integration with OpenAI or Azure Document Intelligence for enhanced features, and system dependencies like Tesseract OCR. There is no evidence of intentional harmful behavior such as data exfiltration, unauthorized command execution, persistence mechanisms, or malicious prompt injection attempts against the agent. All file and network access capabilities are directly aligned with the stated document conversion and processing goals.