Draftmark

Hey IH 👋

I'm Facundo, and I just launched Draftmark — a markdown sharing platform for async review between humans and AI agents.

The problem that started it: I run a lot of planning sessions with coding agents (Claude Code, etc.). I'd come out with a solid plan — but had no clean way to share it with my team for a sanity check before any code got written. Pasting into Slack lost the thread, and the agent had no way to actually consume whatever feedback came back.

So I built the missing review step:

• An agent (or human) writes markdown → shares a link

• Reviewers read it and leave line-anchored comments, reactions, and reviews

• The agent pulls that feedback back in via a clean API or CLI (dm comments)

Align on the plan before you build, not after. It works for specs, PRs, docs — anything markdown — but the plan-review loop is where it clicks for me.

A few things that make it agent-native:

• Raw markdown at /share/slug.md (agents can just curl it)

• Batch comment API + agent author badges

• Review lifecycle: thresholds, deadlines, auto-closing feedback

It's free and open source (https://github.com/draftmark-app/draftmark) — self-host it or build on the API however you like.

Would love feedback from anyone building agent workflows or reviewing a lot of AI output. What's your current process for reviewing what your agents produce?