What problem does it solve? Researchers and engineers often need to turn heterogeneous files—PDFs, Office documents, HTML, CSV, EPUB, images, and audio—into clean Markdown for text analysis, search, and LLM/RAG pipelines, without manually handling each format or risking unsafe network and plugin behavior. ## Core Features & Use Cases - Local and stream conversion: Use convert_local(), convert_stream(), and convert_response() to safely convert trusted files, uploaded bytes, or application-fetched HTTP responses into structure-preserving Markdown. - Batch and literature workflows: Run bundled scripts to batch-convert directories with manifests, or convert PDF collections with YAML provenance front matter, SHA-256 hashes, and year-organized indexes. - OCR, Azure, and MCP integration: Route scanned documents through the official markitdown-ocr plugin, Azure Document Intelligence, or Azure Content Understanding, and expose conversion to agents via the official markitdown-mcp server. - Use Case: You have a folder of 200 research PDFs named Author_Year_Title.pdf. Run the literature script to produce provenance-rich Markdown files plus an INDEX.md and catalog.json ready for a RAG pipeline. ## Quick Start Ask the assistant to convert a trusted local file such as report.pdf into Markdown using MarkItDown and save the result as report.md.
