md mdreview
A review step for your coding agent

Human-in-the-loop markdown review for AI agents

Your agent writes a markdown draft — a plan, a spec, a PR summary. You review it in the browser with threaded comments. It reads them back over HTTP or MCP, replies, resolves, and pushes a revision.

$curl -fsSL https://mdreview.space/install.sh | sh

One command. It asks how you want to run it — local or hosted.

How it works

A tight loop between your agent and you

1

Your agent drafts

Claude Code writes a markdown draft — a plan, spec, or summary — and hands it to mdreview instead of charging ahead.

2

You review in the browser

Read it, leave threaded comments, request changes, or approve. No account needed for local.

3

The agent revises

It reads your comments back over HTTP or MCP, replies, resolves threads, and pushes a new draft.

Repeat until you approve — the agent never ships past a review you haven't signed off on.

Coding was never the bottleneck. Thinking was.

Agents generate code instantly — but instant, unreviewed output falls over. mdreview puts a plan-and-review checkpoint back in the loop.

mdreview promo AI writes code fast. Why is it still broken? Coding was never the work. Thinking was. mdreview: your agent asks you before it ships dream the castle plan the pieces check it works skipped mdreview plan + review before execution you stack slow, takes days robot stacks instant, free fast, and falling over shipped, and standing try mdreview on GitHub

Two ways to run it

Your machine, or ours

Local

Open to everyone

Runs on your machine — no account, no Docker. Starts mdreview on localhost:8137 and wires Claude Code straight to it.

Hosted

Early access

Connects Claude Code to the managed instance at app.mdreview.space. Invite-only in preview — request access.