Clarity in every diff.
Scan changes without the clutter. Keep the code and its context right where you need them.
Your open-source AI code review agent. Start reviewing online,
bring your own API keys, and make every review follow your rules.
Use the hosted app. Or self-host. Your workflow, your choice.
export async function authenticate(req) { const session = await getSession(req);− if (session !== null && session.isValid) {+ if (!session?.isValid) return null;+ return session.user; }01 / The review, reconsidered
Scan changes without the clutter. Keep the code and its context right where you need them.
Turn feedback into progress with focused threads and clear resolutions. Less back-and-forth. More building.
Read the source. Shape the roadmap. Make your review workflow your own. Good tools belong to their community.
02 / Yours, all the way down
Start in the hosted app, choose what powers your reviews, and define what good code looks like. Prefer your own infrastructure? Self-host it.
Use your Claude or Codex subscription, or connect your own API keys. You choose the intelligence behind every review.
Your provider. Your credentials. Your choice.
Turn your team’s standards into Markdown. Drop them in .openreview/ and give every review the right context.
# Code review standards- Prefer clear, focused changes.- Flag missing error handling.- Ask for tests when behavior changes.# UI review checklist- Check keyboard navigation.- Keep layouts responsive.- Use accessible color contrast.Use the hosted app to get started, or self-host OpenReview on your own infrastructure. Choose the setup that fits your team.
SELF-HOSTABLE BY DESIGNYOUR NETWORK. YOUR WORKFLOW. YOUR CONTROL.
Start online. No self-hosting required.
Bring your provider and define your review rules.
Resolve, approve, and ship.
Open the app and put AI code review to work. Your keys. Your rules.
Much easier to follow. Early returns make the happy path clear.