300,000 installs on GitHub 🎉

+ PII leaking into logs. + An auth server error. + A concurrency bug in an idempotency path. CodeRabbit surfaced all three in the systems that protect 1.6M+ customers, before a human even opened the PR. Find out how we managed that (and more use cases) below 👇

VS Code CodeRabbit extension v0.21.2 is live 🚨 > In-IDE reviews now stay connected and reconnect more reliably during longer sessions. > Sign-in flow improved to help you reach the right workspace. Install or update to take advantage of the latest version ✌️

CodeRabbit can now investigate and fix CI failures on GitHub! Once you trigger the action using '@​coderabbitai fix-ci', CodeRabbit will investigate, fix, and deliver a stacked PR ✌️ Check out the docs below to get started.

Pattern-based scanners can parse your syntax but they have no idea what your app actually does. That's why the logic flaw ships clean while you burn a sprint triaging a "critical" in dead code. Reasoning and trust across the codebase is the whole game now.

CodeRabbit can now take action AFTER a PR has merged, not just before ✌️ You can configure up to 5 post-merge actions across your org and set them up via Settings > Post-merge actions! (Full docs below)

Maintainers lose whole days answering the same setup question, triaging half-formed bug reports, nudging stalled PRs. CodeRabbit Agent now does that inside your Discord server. Answers from your docs, triages bug reports, drafts fixes as PRs. All inside the thread! Free for open source projects.

We tested GPT-5.6 Sol and Terra where developers will really feel the difference: Coding Agents and Code Review. Sol followed long task better, stayed oriented in repo work, and added recall in CodeRabbit review benchmarks. Terra looks more useful as a cheaper lane for scoped work.

One codebase, three identical PRs, three different CodeRabbit review profiles. > Quiet - only the most important feedback > Chill - balanced (the default) > Assertive - most comprehensive Dial in the review depth that fits your team ✌️

"quiet" reviews now join the roster alongside "chill" and "assertive". "quiet" will surface only the most important findings inline while keeping everything else inside collapsable sections. You can adjust your review profile in repository settings!

CodeRabbit's review comments now cite their sources! Bad guideline? delete it. Irrelevant linked repo? disable it. The review starts matching how your team actually works.

🐛🪓🏃

Your PR history shouldn't be a mystery. Timeline View is live in CodeRabbit Review UI. Filter by context, see what happened and why it happened. Full explainability, so you can ship with confidence!

CodeRabbit Agent now runs in Discord. PR fixes, automations and code work, in the channels your community already uses. Scoped so your server can't be misused. Free for open source.

Two years ago everyone assumed AI would just make code review faster. Same job, less time...right? That assumption is the whole problem. ~1.7x more issues Logic bugs up 75% Security holes up around 2x. The bottleneck has shifted right, from writing code to reviewing it.

A little teaser of what we’re cooking for you. CodeRabbit Agent inside Discord. The same one that reviews your PRs now lives in your community.

Enterprise users can now programmatically access their organization’s Learnings via the new 𝙶𝙴𝚃 /𝚟𝟷/𝚕𝚎𝚊𝚛𝚗𝚒𝚗𝚐𝚜 API endpoint. Use it to export, audit, or integrate learnings into your own workflows! See the Learnings API reference below for the full parameter and response schema ✌️

Tokens are an input. Merges are the outcome. You won't win by burning more compute, you'll win by shipping good code faster without rationing which PRs get reviewed. Out with tokenmaxxing, in with mergemaxxing!

CodeRabbit now includes React Doctor v0.5.6 to make sure your React code is secure and performant! Enabled by default, but easy to switch off. 👨‍⚕️

You shouldn't have to read the diff to understand the PR. Our new Overview page gets you oriented before you touch a line. PR summary, walkthrough, actionable-blockers, comments all in one place. Live now in the Change Stack 👇

Don't let code reviews be an expensive bottleneck during your agent loops. CodeRabbit CLI can now run 1.5x faster with light mode. Update & run 𝚌𝚘𝚍𝚎𝚛𝚊𝚋𝚋𝚒𝚝 --𝚕𝚒𝚐𝚑𝚝 today.

CodeRabbit now lets admins review newly created Learnings before they are added to the knowledge base! Set 𝚔𝚗𝚘𝚠𝚕𝚎𝚍𝚐𝚎_𝚋𝚊𝚜𝚎.𝚕𝚎𝚊𝚛𝚗𝚒𝚗𝚐𝚜.𝚊𝚙𝚙𝚛𝚘𝚟𝚊𝚕_𝚍𝚎𝚕𝚊𝚢 to 𝟷 through 𝟹𝟶 days at the organization or repository level to hold new chat-sourced learnings for review!

The hard part of AI code review isn't checking what changed in PR #​4306. It's checking everything that changed because of it.

Give a dev 30 seconds to approve or reject a PR and watch the bugs sail through. We ran exactly that at App.js Conf, then asked devs there, plus JS Nation and React Summit, how you actually review code. Thanks to all three for having us! Here's what we’ve heard 👇

Best way to test a feature you built? Run it on real repos. So we forked the octokit repos into an org and used our automatic repository linking feature. CodeRabbit now walks the dependency graph on its own and flags cross-repo impact at review time.

An AI code review company telling you not to talk to humans like machines. yes, really. Humans don't have an API (yet). Read more in our latest blog post.

Most rework starts the same way. Nobody really mapped the change before the first line got written. CodeRabbit Agent for Slack plan feature turns your task into a codebase-aware plan. Right in your Slack thread.

We benchmarked Fable 5 on 105 code reviews. It nearly matched on coverage and then trailed on precision while filing 253 comments. Great at building, noisy at reviewing. Full breakdown 👇

There was a critical issue in Swiggy's codebase and their existing security tooling missed it entirely. CodeRabbit caught it. That single finding turned into a 70% drop in merge time and a rebuilt review workflow across 1,000+ devs.

CodeRabbit Review UI now available to ALL GitLab users (gitlab​.com and self-hosted) > Click "Review Change Stack" in the gitlab MR to see the new UI. > Navigate through layers, cohort summaries, semantic diff view, and more in a guided review flow. 🦊