I'm not sure if this is a front end issue, core issue, or chatgpt integration issue. I have been playing with 'personality' assistants. I have an angry british woman, a sassy overbearing grandma, a grumpy grandpa, a depressed robot, etc...
This worked fine until I hit 5 or 6 assistants, but just now I noticed my grumpy grandpa prompt had been replaced with my cheshire cat prompt.
I 'fixed' the grumpy grandpa prompt, only to find that all six of my assistants had now been overwritten with my grumpy grandpa prompt.
I can't find any bugs that look like this, does anyone have a suggestion whether this issue should be HA Core, HA front end, or something else? Or is it maybe a known issue I haven't found yet?