Blue rather than yellow colored squiggles, but ReSharper (and I expect JetBrains IDEs in general) kind of does this.
It can point out things like unreachable code, redundant if predicates, suspicious casts and countless other things through realtime semantic analysis of code.
Of course there are infinitely more kinds of logic errors that simple static analysis like this can’t pick up, but an LLM “analysis” might.
I actually implemented that! I have a private IntelliJ plugin that feeds markdown documents through an LLM and then highlights internal contradictions with the usual yellow background used for warnings. I use it when writing specs.
Blue rather than yellow colored squiggles, but ReSharper (and I expect JetBrains IDEs in general) kind of does this.
It can point out things like unreachable code, redundant if predicates, suspicious casts and countless other things through realtime semantic analysis of code.
Of course there are infinitely more kinds of logic errors that simple static analysis like this can’t pick up, but an LLM “analysis” might.
I actually implemented that! I have a private IntelliJ plugin that feeds markdown documents through an LLM and then highlights internal contradictions with the usual yellow background used for warnings. I use it when writing specs.