Your agents write Markdown.Your team reviews in Google Docs.
Local AI agents work on plain .md files — no token tax. Reviewers keep the real Google Doc, link, and comments. md2gd keeps both in sync so nobody copy-pastes.
Link it once. Never sync by hand again.
Runs from a lightweight local app. Your Google access is brokered through md2gd — document content flows straight from your machine to Google.
Link a file
Pick a local .md, then create or link a Google Doc.
$ md2gd add spec.md
Work anywhere
Agents edit the Markdown. People edit and comment in the Doc.
# both sides, live
Always synced
Edits flow both ways, block by block. No copy-paste, ever.
.md ⇄ 📄
Built for agents and humans at once.
Each side works in its native tool — local Markdown for agents, Google Docs for people.
Agents stay local & cheap
Agents read and write plain .md on disk. No Google MCP round-trips, no token tax, no latency.
True bidirectional sync
Edit in either place; changes merge block by block. Non-overlapping edits just work.
Keep the real Google Doc
Reviewers use the link, comments, and mobile they already know. No new tool to adopt.
Reviewable history
Every change is tracked and recoverable — agent vs. human, for good.
We never store your docs
Content flows daemon ↔ Google directly. Your Google access stays encrypted and scoped.
Zero bookkeeping
No “which version is current?” The service owns reconciliation so you don’t.
Free while we’re in beta.
A paid plan is coming — join the wishlist and tell us what you’d want from it.
Free
- Two-way Markdown ↔ Google Docs sync
- Local sync daemon
- Provenance & version history
Pro
Paid features are still being shaped. Join the wishlist and tell us what you’d pay for.
Join the wishlist