alt Hacker NewsProgrammers are "program reviewers" now, more and more.
But the tools we have to do this suck. We don't have much beyond "here's the source code as text, look at it, and think really hard".
Another comment[0] suggested having AI generate "code walks" through the code. This seems better than our current approach of "my PR contains a diff, good luck".
I'm working on a tool in Emacs that allows me to mark certain lines, or entire projects, as needing review, and then helps me track which parts have been reviewed. I need something to help me find parts of the code I have never reviewed before.
Replies
> Programmers are "program reviewers" now, more and more.
Our job, much like a locomotive operator, is to make sure the agents don't go off the rails. We call ourselves software engineers in recognition of that.
As PRs get beefier I find myself using the LLM to ask pointed questions and do general surveys of the diff to flag architectural violations, smells, etc. before I read through the entire thing by hand. If you put all of your architectural and business decisions in the repo via spec-driven tools or otherwise then this becomes quite useful. It also helps if you have a vision in your head of what the system is supposed to look like structurally so you can ask the LLM specifically how the PR is handling things that you'd be on the lookout for anyway.