I would like to see a client integrate OpenAI's moderation API into a client to filter out notifications that are unpleasant.

Of course long term it would be better to host a local AI model for moderation but the only model I've seen dedicated to moderation is OpenAI's