>>81892436
reads like chatgpt. oh well here's your (you) anyway
>I am also suggesting obligations to strangers are simply to call for help if they are in danger, or to not be cruel towards them
>That support wouldn't be self-harming, it would be directing them to appropriate help or reassuring them.
snapping them back to reality is the least cruel, most helpful response possible short of not engaging with them and trusting them to handle their own stuff they can handle only by themselves in any case whatsoever.
if you'd ever been at least at the point where OP was, you'd know that the ultimate indifferent but rationally infallible cookie cutter "get therapy" or social signaling reassurance platitudes are volatile ropefuel both short-term and long-term