
anydoc
Convert office documents to Markdown

What you get
Point it at a Word, PowerPoint, Excel, OpenDocument, RTF, EPUB, CSV or PDF file and get clean GitHub-Flavored Markdown in single-digit milliseconds. Rust core with Node, Python and browser bindings, and it ships as an Agent Skill so your agent can read whatever document it runs into.
Why it exists
Every format needs its own parser, each one produces differently shaped output, and feeding any of it to a model means writing the normalisation layer yourself. The formats are old, fiddly and all slightly wrong.
How it works
One library, one consistent Markdown output, no matter which format goes in. The WebAssembly build runs in the browser, so files convert locally and never leave the machine. Firecrawl built it for their own parsing product and open-sourced the core.
For
Developers and content teams processing mixed office documents and needing LLM-ready Markdown output.
- Use cases
- / Convert documentation from Word to Markdown
- / Batch convert spreadsheets into tables
- / Extract scanned PDF content with OCR
Useful? Star it. That is the only thanks most maintainers get.