The Python package does the reading and writing; the other tools wrap it. A second writer in JavaScript, written from the spec alone, passes the same conformance kit.
Encode, decode, validate and compare sizes, in batch or in CI.
cellprose encode book.xlsx cellprose decode book.sheet.md cellprose check book.sheet.md cellprose stats book.xlsx
The core library, for pipelines.
import cellprose
text = cellprose.dumps("book.xlsx")Agents list sheets, read workbooks as Cellprose, and write cells back to .xlsx.
pip install -e ".[mcp]" claude mcp add cellprose -- cellprose-mcp
An index of the format for agents, and the spec, skill and install steps in one file.
/llms.txt /llms-full.txt
A short SKILL.md that teaches an agent the blocks, so it needs no spec lookup.
skills/cellprose/SKILL.md
Readable diffs for .xlsx files already in a repo, and a pre-commit hook that writes a .sheet.md twin.
cellprose git-setup cellprose pre-commit
An independent writer with no dependencies. It runs this site's playground.
node js/bin/cellprose-encode.js book.xlsx
A browser extension for .sheet.md files on GitHub: sheet tabs, and click a cell to see its real formula and the cells it reads.
viewer/extension
Language-neutral cases (JSON workbooks, expected .sheet.md, decoded cells) and a checker, for writers and readers in other languages.
python conformance/check.py --encoder "your-encoder"