dearilos

Put up guardrails to enforce quality code.

We’re fixing this slop problem - engineers write rules that are enforced on PRs. Fixes the problem pretty well so far.

We’re trying to solve a similar problem, but using linters instead over at wispbit.com

We're trying to solve a similar problem at wispbit - this is an interesting way to do it!

We use ast-grep for the determinism part. I should have clarified - we don’t charge for fully deterministic runs. Only ones where the LLM is involved as a judge.

It supports fully deterministic rules, which we use LLMs to help you write.

Agreed on all of this too. This is why we built the CLI tool - to shift left the work.

We combine determinism + LLMs to catch things a human would normally have to. If the LLM finds a violation, it generates a comment.

Big agree on the CLI being open and letting you bring your own inference provider. We’re holding off on it until we get more feedback from some of our hardcore users.

<3

We do a two week trial and then it's $0.2 per file reviewed. Buying in bulk + optimizing rules gives a significant discount.

Love it :) Thank you!

This is pretty much what I’m solving. LLMs need a “linter” of sorts that can guide them to write good code.

I'm building something to solve exactly that - automating all the boring and repetitive parts of code review.

I'm building something to do exactly that - just reduce and automate the boring parts of code review like enforcing standards.

problem with the tools you're using is that they're not built for code review

im building one that lets you write and enforce your own rules so you don't get the typical slop

email in profile if you'd like to try it - i can send you a link

i'm taking an approach where we scan your codebase and keep rules up to date

you can enforce these rules in code review after CC finishes writing code

email ilya (at) wispbit.com and ill send you a link to set this up

do you use anything today for automated code reviews?

ive been building out a directory of code review rules for the last couple of months!

are you open to chatting and sharing notes on what works/doesn't work?

my email is ilya (at) wispbit.com

I'm actually building something like this.

A code review tool that can be ran by Claude Code (and other tools like cursor, kiro, windsurf) where it does the first pass.

You can also define the rules on how code should be reviewed, and we have a free library of prompts for common code review checks.

I didn't see a way to contact you in your bio so please reach out - ilya (at) wispbit.com - and I can give you access.

Do you use any tools to help with the code review part?

How has that worked out so far?

I'm building something similar and I found that code review with LLMs is really good when:

- You give it specific rules. I built a directory for these because they made the reviewer so much better [1]

- The rules you write are things your team already looks for during review (proper exception handling, ensuring documentation, proper comments, etc.)

[1] https://wispbit.com/rules

AI coding agents help you solve the problem faster

AI code review helps you catch issues you've forgotten about and eliminates the repetitive work

These tools are helping developers create quality software - not replace them

how'd you get your code review automated?

There is so much slop that you end up ignoring the important stuff

AI code review really does help but only when used correctly

I'm building the tool to fix that

This is what sucks about code review today

It's just slop and useless context

Code review should only help with the repetitive and tedious parts of review

I'm building a tool to fix that

if you do specific prompts based on your team's tribal knowledge and standards it works really well

"look at this code for bugs" doesn't end up working well, which is what most code reviewers do.

how are you dealing with the accuracy of the review comments?

in my experience "review this PR" is very generic and ends up giving slop.

There's a lot of arguments being made here that background agents aren't effective because code review is the bottleneck.

I built a tool that understands the codebase and can give the agent the first pass. So you don't need to spend mental bandwidth to review mountains of code.

If anyone is open to trying - email is in profile

I'm actually building something to fix this.

The biggest bottleneck for background agents is code review.

I'm building a tool that can give the first pass so the result of the background agent isn't garbage most of the time.

I found that early and often code reviews can offset the reduction in productivity. A good code review process can fix this.

I'm thinking about this a little differently.

You have to catch most issues at code review.

You can have an agent spit out code, but on PR open you should have another agent verify rules based on rules you define as a team.

It's what I'm building at wispbit.