Agreed, I’ve found that LLMs favor wrapping entire functions in broad try:catch structures, making it nearly impossible to identify and debug problems. Dangerous if you don’t correct it.