> because LLMs produce worse code than the typical

I'm not so sure that's true anymore.

But regardless - no one uses just one coding agent. It's a series of agents - a group to plan the work - coders to implement it - and then agents to review the work. Overall it comes out better than human code.

Here - let's test it - give me your GitHub and we can compare your code to theirs.

> Then you aren't a very good programmer

Ohh.... my sweet summer child...