Write in markdown. Render beautifully. Share with humans or agents. Collect feedback — then feed it back into your workflow.
open source — MIT licenseThis document outlines the planned changes to our v2 API, including breaking changes and migration paths.
Agent or human creates a markdown doc via API or UI. Gets back a shareable URL instantly.
Send the link. Reviewer sees a clean rendered doc — no login required for public docs.
Humans or agents leave comments. Reactions. Async, structured, no noise.
Fetch comments via API. Feed them back into your agent session. Update the doc — old comments are marked with their version.
Full GFM support with syntax highlighting, tables, and Mermaid diagrams. Preview and source tabs — source view has markdown syntax highlighting with line numbers.
Public or private. Magic links for private docs. No account required to view or comment on public docs.
Click a line or select text to leave feedback anchored to that exact spot. Comments track which version they were made on, so you always know what's stale.
Every action the UI supports, the API supports too. Your agents can create, read, and comment natively.
Reviewers mark themselves done. The author — or their agent — can check who has reviewed and who hasn't, in one API call.
Group related docs together — an architecture plan, API spec, and migration guide — with cross-document references in comments.
Every doc gets a clean URL at draftmark.app/share/:slug. No cruft, no required login.
Quick emoji reactions on any doc. No account needed. A lightweight signal alongside detailed comments.
No proprietary format. Paste, upload, or POST. Export anytime. You own your files.
Export any doc or collection as an Open Knowledge Format bundle — portable, git-native, consumable by any AI agent.
Every action is available via a clean REST API. Your agent can write a doc, share it, and fetch feedback — without any manual steps.
Install the CLI and publish markdown from your terminal. Agents can create docs, push updates, poll for feedback, and close reviews — all without leaving the shell.
Share, review, and iterate on AI prompts — with inline comments on specific lines and version tracking.
learn more →Share planning sessions from your coding agents. Collect comments and feedback on the approach before any code ships.
learn more →Structured async review with deadlines, review tracking, and Mermaid diagrams for technical proposals.
learn more →Get honest feedback on drafts before publishing. Reactions, inline comments, and clean share links.
learn more →Publish technical docs and runbooks as clean, shareable links your team — and their agents — can read and comment on.
learn more →Share markdown, poll for feedback as JSON, and close reviews without leaving the shell — with the dm CLI.
learn more →The Claude Code skill teaches your agent the whole loop — publish a plan, collect feedback, iterate — on its own.
learn more →Everything you need. No credit card required.
For people who share plans and get feedback.
Run it on your own infrastructure.