I want to introduce an idea for utilizing AI for huge repositories. It’s a tool-agnostic technique. The IDE you use does not matter. The size of your code base can be huge; AI will still work with this technique. tl;dr; My .cursorrules: - Firs...
Markdown supported · ⌘↵ to post
Fyodor
Why'd you (software engineers) have to go and make things (software development) so complicated...
That should work, good idea. I also have a huge monorepo where Cursor fails at doing reasonable helpful work too often. Need to try that. A bit tedious though, but it makes sense to adopt something like that progressively and iteratively.
BTW, WDYT about the new
.cursor/rulespractices? I feel a bit chaotic about that for now TBH but with this idea of yours it may actually play well…