Yes, Code Review Is Still Necessary

Agents are not average tools. They are smart and resourceful, so they may bypass instructions if they see more efficient options, or they may simply make mistakes, which happens with intelligence. Therefore, in my experience, setting boundaries tends to work well. They may make mistakes, which are recoverable if failures are taken into consideration and guardrails are set, but they learn the lessons much more quickly. Their agency, which is annoying sometimes, can also keep design honest and efficient.

Not exactly on topic, but I do like this disclaimer in the bottom of your post:

  AI Disclaimer: This article is original content produced 100% by me (a human).
  This post was composed by me dictating my thoughts to an LLM and asked it to
  produce a cleaned up version of it. I reviewed and edited the output to make
  sure it matched my original thoughts.

Seems to be an example of a healthy usage of an LLM!

I have the same experience trying to set up an 11ty based static site with the help of Gemini (Back in February). I had to specifically instruct to get proper templating set up.

It also used two templating engines (IIRC) and got itself into a situation it could not find it's way out and asked for help.

I was just trying out LLMs, so it was interesting.