I'd like each helper to be able to produce a list of brain ai snippets they are using to generate my responses and allow me to audit it. I mentioned sustainability and period properties and now that's all they suggest ideas about despite me having asked them to stop.