This is probably the most frustrating (repeated) interaction I've had with LLMs ever since ChatGPT. I truly hope Anthropic still scan for "ffs, F**KING" to work on these as a priority. I've never thought I'd say that, but, in this case, I hate "honesty", it seems. I run the