Ask HN: What tools are you using for AI-assisted code review? | Hacker News Reader

Ask HN: What tools are you using for AI-assisted code review?

26 pointsagos31 comments

My team (around 40 people who write code) is evaluating tools for AI-assisted code review. The market appears to be rife with options, so before starting a series of free trials, I’d like to ask a knowledgeable crowd. What tools or services are you using? Do you use them just for code review, or for other purposes as well, such as incident response or branch management? Why did you choose them, and what do you like or dislike about them?

View on HN

Co-creator of Mira [1] here.

This is exactly what we built Mira for. It's self-hosted, bring-your-own-model/BYOK, and most importantly, open source.

You point it at your own API keys (e.g. OpenRouter or a local model) so nothing leaves your infra, and it runs as a code reviewer on PRs.

It's also ridiculously quick at reviewing (benchmarks at ~77s) because your PRs aren't sitting in a queue on a cloud somewhere (alternatives are > 5 minutes)

We're working really closely with our users to build the best possible code reviewer. Feedback and contributions are highly encouraged.

[1] https://github.com/miracodeai/mira

Besides local review via codex and Claude code, we are using GitHub Copilot with custom instructions. We just assign it as a reviewer in GitHub and a couple minutes later, the review is done. It raises a lot of issues which are valid and which I never had found. https://docs.github.com/en/copilot/tutorials/customize-code-...

what custom instructions did you give it? standard stuff about your practices?

Exactly

Using dupehound for identifying duplicated code.

What I use for: I use for identifying duplicated code. It is deterministic, doesn't use AI, offline, runs from CLI and is super fast (and free).

What I dislike: I won't say it I dislike, but it is not a tool that does all the jobs of a code review. For instance, it doesn't flag security issues. It is superfocused on code duplication (it performs better than Sonar for this use case) and is specifically useful for large codebases. Disclaimer: I am one of the collaborators, so take it with a grain of salt https://github.com/Rafaelpta/dupehound

I use my own tool (released as open source "StrangeDaysTech/straymark"), which takes blocks of related implementation tasks, builds a comprehensive prompt with auditing instructions (comparing intent against implementation), then I run agents from three different models to perform the audits and generate reports. These reports are then analyzed by the main agent to rule out hallucinations or misinterpretations of intent, and finally, a remediation plan is created. This tool can perform this task fragmentation because it's part of its cognitive governance function. Its operation is somewhat complex, but these auditable work units have a coherent structure within the overall project thanks to a knowledge graph that is built from the early design phases through to implementation.

Opencode, mainly because I appreciate how one of the founders treats the UX as a first class concern. Its a great tool to learn since it can help us pivot from the potential impending provider crisis where teams may start having to consider things outside of the large labs.

As my daily driver at home, I use Pi though because it doesn't get in your way and forces you to understand how the sauce is made.

Surprised to not see it mentioned yet, I've been using Kilo and I'm pretty happy about it. Local development with Claude, review using Kilo : https://kilo.ai/docs/automate/code-reviews/github.

Built my own using Claude Code; inside a gitlab job we call Claude Code headless. This works well. There is a tiny mcp server exposed to Claude so it can post inline comments. All existing comments are fed into the reviewer to avoid double posting. The quality of feedback is high. Most complexity is in the SHA management. For example after a rebase. Luckily LLMs understand git very well otherwise it would have been impossible for me.

Claude-Code and Codex in combination, combined with an IDE such as Google Antigravity or VisualStudio-Code are very powerful tools, if your company can invest in hardware the new Mac Studio and MacBook Pro allow optimized local inference through open-source tools such as: https://github.com/antirez/ds4

I don't use these tools, but wouldn't it be better to use them only after you do a manual review to see if they find anything you missed? Otherwise I could see reviewers getting false confidence and doing a less thorough review. This happens with seeing that unit tests pass.

that's a good point and surely something to test out and see what works and what not

i've hoped on to many tools but based on working with all of those tools , o mostly liked this raptor mini(preview) model form vs code , which is very good from my perspective . i gave the same prompts to claude , raptor , kimi and a bunch of others , based on the comparison in speed , quility raptor takes the first place , you can even see the thinking of it not like claude (you cant see what it's thinking ) .

Do you already use static code analysis tools?

For Claude Code, I think the standard is Codex + Gemini. Why these two? Because it “covers” the blind spots the others would miss by themselves.

I mostly use CodeRabbit via GitHub PR

https://www.coderabbit.ai/

does it work well for you? did you do any fine tuning worth mentioning?

No I did nothing. I use it as a first check before I review the code myself on my Open Source projects.

Github copilot is a little too opinionated, but we still use it to catch obvious stuff.

Codex on top of that with specific rules and syntax requirements.

I just use codex / claude code to do reviews. It does a good job, and it's easy to explain and navigate.

/review in claude code - the skill pulls the PR from remote and review it. Can post comments also.

Honestly I get awful results using this skill. The output is way better when I simply ask it to review the changes on my branch compared to origin/main

/review flags the stupidest things

Rolled our own with OpenCode, seems to work quite well and meets the goal of being vendor agnostic :)

I usually use gpt5.5 in Codex It feels like magic

claude code and github copilot

I built my own for a specific review use case: improving code modularity and minimizing slop. Forcing the agent to use it in a loop to clean up the diff means that I get something cleaner to review manually

https://github.com/thempatel/mdlr

Rovo/Bitbucket

Frankly, coding with Claude Code and having Copilot read through the PR is complementary and helps to catch some things that slipped through

CodeRabbit

grep

Reply on news.ycombinator.com