Comments URL: https://news.ycombinator.com/item?id=47317986

Points: 4

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=46795074

Points: 2

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=45745666

Points: 1

# Comments: 0

It runs directly on GitHub PRs and automatically comments with suggestions based on your team’s own patterns. Kody learns from what you usually accept or reject in reviews, so the feedback gets more aligned with your style over time.

You can also create custom rules to enforce the practices your team already follows, without relying on traditional linting or static analysis.

And if you want to go further, you can bring in external data through MCP — for example, pulling context from tools like Jira, Notion, or Playwright to make the suggestions even more relevant.

The goal is to speed up reviews without losing quality. If you want to give it a try: (https://kodus.io)

From my experience (and from what I hear from users), AI really helps raise the bar for code reviews: it catches bugs, suggests style improvements, flags inconsistencies, and helps maintain security and quality standards. What I find most interesting is how the AI learns from the team’s own history, so the suggestions become more and more aligned with the reality of each project.

At Kodus, we built Kody, an AI agent that integrates with GitHub, GitLab, Bitbucket, and Azure DevOps. Kody provides contextual suggestions and evolves with your team’s feedback. A cool feature is that Kody can analyze your project’s review history and suggest custom rules, which makes it much easier to maintain standards and retain knowledge as your team grows.

For mid-level developers, I see AI as a powerful ally: it brings objective criteria, suggests best practices, and speeds up learning—making it easier and more comfortable to participate in reviews, even alongside more experienced developers.

If you want to learn more, Kodus is open source, easy to integrate, and very flexible. You can check it out at https://kodus.io

Bottom line: AI doesn’t replace human reviewers, but it really helps speed up the process, standardize practices, and support team growth. If you want to chat more or have questions about Kodus, feel free to reach out!

Comments URL: https://news.ycombinator.com/item?id=44200538

Points: 1

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=44191991

Points: 2

# Comments: 0

→ Don’t use new Date() directly. Inject a Clock. → This code is duplicated. → We don’t use lodash here.

Feels like we’re doing reviews on repeat in 2025.

That’s exactly why we built one of the most used features in Kodus: Kody Rules.

Team rules, your way, right inside the PR flow.

And the best part?

Kody learns from your team’s reviews.

It watches the comments you leave on PRs and starts suggesting those same things on the next ones.

No config upfront. No model training.

I recorded a quick video showing one of the rules we use: https://youtu.be/0J4TtooYDHU

→ “Avoid using new Date() directly in services. Inject the Clock.”

Simple, but it prevents annoying bugs, saves repetitive back-and-forth, and keeps standards in place without anyone having to remember them.

If you could automate just ONE comment you keep repeating in reviews, what would it be?

Comments URL: https://news.ycombinator.com/item?id=44061534

Points: 1

# Comments: 1

Teams that rely purely on LLMs like ChatGPT or Gemini run into this all the time — generic comments, no real understanding of the project, and sometimes subtle mistakes. It’s hard to trust, let alone use it as a real part of your review workflow.

That’s exactly why we built Kodus. It’s an AI-powered code review platform that runs directly on your PRs, combining generative intelligence with deterministic rules based on how your team already works. The result is less noise, more consistent feedback, and more time to focus on what actually matters.

If you want to give it a try and share feedback, we’d love to hear it. We’re improving the platform based on real-world usage https://kodus.io

We’re building Kodus, an AI code review agent that works directly on GitHub, GitLab, Bitbucket, or Azure. The idea is that Kody, our agent, leaves comments on PRs with suggestions based on your team’s existing patterns. And over time, it learns from the reviews your team has already done.

You can configure what it comments on and how it gives feedback, and everything happens right inside the PR so the team doesn’t have to leave their usual flow.

If you feel like testing it out and sharing feedback, we’d really appreciate it. We’re listening closely to the community to make Kody more useful in real-world workflows. https://kodus.io

Just wanted to share a quick update we just launched at Kodus.

For those who don’t know it yet, Kodus is a code review agent that runs directly in your team’s Git workflow (GitHub, GitLab, Bitbucket… and now Azure DevOps as well). It helps maintain code quality and consistency by analyzing each PR based on your team’s own rules and repository standards.

Support for Azure had been a common request — so if your team uses it, you can now plug Kodus right into your workflow and give it a try.

Docs: https://docs.kodus.io/how_to_use/en/overview Repo: https://github.com/kodustech/kodus-ai

We’re constantly improving the tool, so any feedback, suggestions, or criticism would be super helpful!

Comments URL: https://news.ycombinator.com/item?id=43872568

Points: 2

# Comments: 0

Here’s the repo if you want to check it out https://github.com/kodustech/kodus-ai

Feedback is more than welcome

I’m part of the team at Kodus—we automate code review, but in a way that actually makes sense for your team. You can set up your own review standards and tweak the minimum severity for suggestions—basically, tell Kody what really matters to you.

If you wanna give it a try and let us know what you think, we’d love to hear your feedback! https://kodus.io/en