It turns XLSX files into structured JSON with cells, formulas, merges, tables, charts, dependency graphs, and token-counted chunks for LLM use. Each chunk includes a source URI, token count, rendered HTML and text, dependency summary, and content hash. It provides a full workbook graph with typed cells, parsed formulas, and every chunk addressable back to exact source coordinates.
It can be installed via pip and runs on Python 3.10+ on macOS, Linux, and Windows. It is released under the MIT license. It requires only three lines of code to parse an XLSX file into LLM-ready chunks.
It can be directly integrated into agent frameworks like LangChain, LangGraph, CrewAI, and OpenAI Agents. It is safe by default with no macro execution, no external links, ZIP-bomb guard, and size limits.
No comments yet.