[FEATURE]input moderation report

[delinecnlin]

Is it possible to have a feature, when the moderation level is reached or with certain key words like hateful and violent, the conversation can be explicitly saved or reported with session ID.