I have not seen any AI review do codebase conventions consistently well. Scenarios like - the code may be valid, but it doesn't use our library for x feature, or in FE it isn't using the new component / styling, etc. Enterprise things. Once someone figures that out it's nearly solved except when requiring undocumented context IMO

See also https://news.ycombinator.com/item?id=47041268

I have no idea who that is.

This is a beautiful car.

Would need auto-fix & commit capabilities to really replace lefthook or lint-staged for me. https://github.com/j178/prek/issues/1051

Sacrifice is for the cost center, not the decision-makers

Has everyone really built their own microVMs? I don’t think so.

Saw quite bit on HN.

A quick search this popped up:

https://news.ycombinator.com/item?id=45486006

If we can spin up microVM so quickly, why bother with Docker or other containers at all?

10 seconds is actually not that impressive. we spin up Sandboxes around 50-200ms at E2B

isnt that number mostly due to firecracker?

I think a 413 commit repo took a bit of time.

That's just over one day worth of commits in a few friends' activity at this point. Thanks to Anthropic.

I'm not interested in integrating an unguided 400-commit single Ralph iteration as part of critical infrastructure at this point.

A lot of the personal projects posted there are immediately abandoned. There's obviously no guarantee that a project you advertise will be maintained, but before there was a likely chance that the submitter was going to. Now, not so much. A bit frustrating when going through new projects, so I get it.

I liked their product. They seemed to run out of funding in the last year though, and the founders lost steam.

I did not get it either. Let me know when new users can trial the product.