Add spam detection on TEXT channels and send STOP command

[docwho2]

adding prompting to detect political and otherwise spam and reply STOP. Usually all these messages say send stop at the end so GPT should be able to detect and reply to stop these messages.