11 ms·
Show HN: I canceled my AI code reviewer and wrote a free local one
- xshahneil76 2mo agosimilary i also solved the local search for all AI IDE tools i used and built a proactive memory and its all local completely and rest encrypted. myinventory.site
- mukundzzha 2mo agowhat problem does it solves?
- doc_ick 2mo agoWhat problem does your solution solve that the others don’t?
- mukundzha6 2mo ago[flagged]
- kosolam 2mo agoCan you share some highlights of what it can do? (I’m just asking the agent to review the pr usually)
- mukundzha6 2mo ago[flagged]
- doc_ick 2mo agoHow is this different from bandit, python black that have differential flags?
- mukundzha6 2mo ago[dead]
- benatkin 2mo agoSeems to be mixing up linting with code review - though this does lint only the changes which is an option (I prefer to lint the whole codebase though)
- mukundzha6 2mo agoI am about to merge this feature as well in upcoming updates.
- launchforge 2mo ago[flagged]
- StableAlkyne 2mo agoI've never seen this idiom before: try: AVOUCH_VERSION = importlib.metadata.version("avouch") except importlib.metadata.PackageNotFoundError: AVOUCH_VERSION = "0.3.2" What's the intended purpose? It seems to be protecting against a scenario where the package is not installed but the CLI is somehow running... ...But that seems like a very odd thing to be concerned about. Plus, unless I'm missing something, the program will crash with an an ImportError before that block can ever run.
- gchamonlive 2mo agoCould be for unit testing, maybe it's something that makes sense for the developer, or it's AI slop. In all cases it's a minor nitpick in my opinion.
- StableAlkyne 2mo agoI'm assuming it's vibe coded, but it's a code smell I've noticed is specific to vibe coded projects. AIs tend to have a debilitating phobia of any kind of exception. Claude especially will throw try/except and weird coersion everywhere in the code. This bites you because it hides legitimate errors, even when the program is in such an unrecoverable state that it doesn't make logical sense to continue. It becomes impossible to debug, and requires the test suite to be absolutely perfect because you end up throwing away the only way for the language to warn you that something went wrong. So while yes, it's relatively minor, like many code smells it is indicative that other things may be wrong.
- gchamonlive 2mo agoAre other things wrong? I'd be wary to shame people's work without doing legwork to getting evidence to support your claim, otherwise you risk passing as someone just seeking attention and upvotes, riding on the tides of HN anti-AI sentiment. Which is valid for vote farming just to be clear, I'd just respect you less if this was the case.
- 2mo ago
- ramon156 2mo agohow is tjis better than (OpenCode|Pi|CC|Codex) that had a Reviewer agent? the whole idea of e.g. CodeRabbit is that it allows to review PRs autonomously. making it local defeats the purpose. it doesnt help that these projects have AI generated README's. Not AI-assisted, or AI-curated, but fully generated. it's a bunch of sentences that say nothing. "the diff is whats reviewed". okay, what about per-commit reviews? does it keep context of that? just diffs isnt enough to review a PR, it just sounds efficient but youre leaving a lot of context out, so the reviewer will have trouble confidently reviewing your changes.
- mukundzha6 2mo agoI just formatted and fixed it using ai. I didnot knew its looking fully ai generated. I will fix it. Till day after tommorow readme is going to be updated. And we are not competing with ai tools. I am aiming it to make it an standard for developer before making commit. Later its gonna have ai intergartion as well. THANKS BUDDY FOR THAT SUCH GOOD FEEDBACK!.
- DemolitionMan 2mo agoI can't quite figure out what distinguishes it from other similar projects—and, above all, what problem it solves.
- mukundzha6 2mo agoI wanna make it standard for developers before making pre commit. We are local focused , no data stored of anyone and provide lightning fast analysis for now. Later we are gonna make ai integartion for better output. THANKS BUDDY FOR THAT FEEDBACK.
- doc_ick 2mo agoHowever other tools like gitlab sast, python black, and others have differential flags that only scan for changed files. Worst case on a precommit that can be wired to pass only git changed files to be reviewed. Not for criticism, but is this targeting that “air gapped” git diff scanning area?
- mukundzha6 2mo ago[flagged]
- manganate06 2mo ago[flagged]
- doc_ick 2mo agoThought this looked familiar. A repost likely by the same person. Og: https://news.ycombinator.com/item?id=49219370 https://news.ycombinator.com/item?id=49219370 Especially with a few commits just renaming the project and removing the roadmap.
- mukundzha6 2mo agobro sorry if you felt inconvenience but i did it because my previous post from different account got flagged. so i just checked it by uploading whether it got post or not.it got somehow posted and it boomed so i was unable to delete it.
- doc_ick 2mo agoThen why’d you rename the project and remove the roadmap, and does this only solve the untuned-ruff problem?
- mukundzha6 2mo agoi renamed this project because we were having a conflict name with 2 big names. Theres an scut automation also there in google. There is also an github repo by facebookincublator nmaed scrut which is cli testing tool so i changed name.i removed the roadmap because when we changed name i have to manualyy change word scrut to avouch in roadmap.md so i deleted it and gonna add new soon. It solves just finding the problems for now.later we are gonna add features like ai integrtion, fixing codebase many more
- doc_ick 2mo agoAll right, I could see the issues with the naming. Though I don’t understand solving the problems (linting) and adding ai how are different from ruff with git diff, I hope the roadmap eventually answers that.
- mukundzha6 2mo ago