Comment by VladVladikoff

Seems like a good example of depending on LLMs too much. After the first bad output the next step should have been to open inspector, not to try three more LLMs.