If you'd ever tried to get an LLM to review its own code, you'd know.
if you get the same session that wrote the code to review it the poor results are entirely deserved.
and if you get a different instance to review the code then you would know that it works rather well.
if you get the same session that wrote the code to review it the poor results are entirely deserved.
and if you get a different instance to review the code then you would know that it works rather well.