We’re excited to announce that the Zenable MCP server now supports OAuth natively.✅ Your existing config will keep working✅ Most IDEs will no longer rely on the old node dependency✅ You can now setup a new IDE or refer a friend just by just logging in - no API keys requiredUpgrading is easy:uvx zenable installThat one command will detect your IDEs and either install or upgrade your MCP configs automatically. No breaking changes — just a cleaner, more modern setup.If you’re happy with your current setup, no rush. But if you want to future-proof your workflow, and clean up any API keys from config files, now’s the time to run the upgrade.⚠️ Note: If you were using Kiro, we’ve had to drop support for now. Kiro doesn’t currently support secure remote MCP servers, which we require for obvious reasons. We’ll be monitoring the situation closely and will re-enable support as soon as they do.

One of the things that stands out at Zenable is how extreme we take "ease of use." That's why we just released a new package which installs the Zenable MCP server and hooks for all of your projects, with just ONE command. Just run uvx zenable-mcp install and you're off to the races. Add --recursive if you're really feeling it, and check out the project on pypi to learn moreStay posted for some new features... 😁

That's a great question, we also wanted to know so today we're releasing our public status page. It looks pretty good to us 🤩 Check it out yourself at zenable.io/status

Ever wish it was just that easy to prevent AI generated mistakes? Well now it's even easier, with our graph-based validation steps, now our reviews not only have world-class context engineering and deterministic validation, now they also have built-in double checks from their peers (🤖).

I know, it's hard to believe, but the Zenable PR reviews just got even smarter! Now every single review gets all of the context it needs, repo-wide. Our automation identifies the right context to pull into the review, and pulls it in so we've got the full picture.But don't worry, we don't save or train on any of your code, that's a promise. We pull in just what we need, only while we need it, then clean everything up to keep your code nice and safe (and 100% owned by you). Learn more by going to our FAQ and Security pages

Are you a fan of the Gemini CLI? Yeah, us too. That's why we added first class support to our MCP server, including a managed installation process and full documentation. And get started in no time by running uvx zenable-mcp install gemini!

There are certain types of files that just aren't worth spending the time doing a line-by-line review of. Compiled code, generated files, or lock files are great examples in many cases. But, not everyone has the same files that they'd like to skip. That's why we now support both inclusion and exclusion criteria in our configuration file, so you can make the PR review bot even faster and more focused by telling it what to skip, and what to always pay attention to. Learn more in our docs.

When it comes to AI coding, nobody's better as of July 2025 than Anthropic (in our humble opinion). That's why we upgraded our MCP sever to use Anthropic's Opus and Sonnet models throughout! Now, not everything that our MCP server does uses AI (I know, shocking, right?), but when it does, now you're getting only the best - Anthropic's Opus and Sonnet 4 models.

We've increased our generated conformance tests (also known as Policy as Code) limit to 10KB per policy. Now you can make sure that even the trickiest rules can be enforced deterministically in code, without the toil of writing it by hand, or even needing to sift through your contracts or security policies to find it. Just upload whatever files you expect have requirements, and have us do all the work - we'll find the requirements, refine the details, generate and test the conformance tests, and then executing them throughout your entire Software Development Lifecycle.

Having multiple MCP tools can be confusing for LLMs; they don't know which one to pick. So, we made it easy. Now, all of our features are now available behind a single tool call. Don't worry, you still get 100% of the capabilities you're used to, and we've implemented a token system so even when you run out of premium tool calls, you'll still get unlimited calls to our deterministic conformance tests. Not familiar with conformance tests? Well, let me tell you, they're pretty 🔥. Specifications are the best way to guide AI tools to generate code that meets our expectations, and conformance tests are how we verify that's what's happening. If you're interested to learn more, just reach out! We're always happy to help.

Looking for a little more "oomph" from your PR Reviews? Now you've got it; we cut over to Claude 4 and have seen a dramatic improvement in contextual, valuable code reviews. Not only does it find security issues, but it flags functional and business logic issues that linters miss by thoroughly understanding the intent of your code using the PR title, description, commit messages, and corresponding files/comments. Check it out for free with two clicks by going to zenable.io/github and installing our app 🚀

Protect your repositories from AI Slop with our GitHub App; get expert-level reviews of Pull Requests in seconds, not days. Set it up for free today at zenable.io/github

Looking for a feature that you want but not seeing it? Let us know! We've just shipped our public roadmap - take a look, submit something new, and vote for features you'd like to see at zenable.io/feedback

Our GitHub App will now automatically review your PR title and description for accuracy based on the actual PR contents. When it finds that either don't align well with the actual direction of the PR, it will suggest an alternative! No more needing to painstakingly write a title and description, just drop open a PR, get the feedback, and go from there!

Now you can get all of the power of Zenable directly into over 50 IDEs by using our MCP server! To learn more, see docs.zenable.io

We're now hosting our documentation at docs.zenable.io; go check it out for our quick start, step-by-step setup guides, walkthrough videos, and more!

We now support integration directly into Windsurf! To set it up, see our documentation at docs.zenable.io

We now support integration directly into Cursor! To set it up, see our documentation at docs.zenable.io

You can now configure the GitHub App via a configuration file; clear out some of the noise by skipping reviews of any specialized lock files, binaries, or compressed files, and don't re-review release branches or Dependabot PRs, and more! See how to configure it at docs.zenable.io