I am currently generating a substantial amount of code with Claude Code. I also pay for ChatGPT Pro. I am thinking of having ChatGPT's codex review Claude's code. Do any of you do anything like that? What are your experiences?

I do it all the time. It finds errors CC doesn’t. Especially in math formulas. I ping pong reviews between the two agents until they converge on no errors.