Show HN: AURA – Open-source behavioral threat detection for LLMs

[kate8382]

https://github.com/kate8382/AURA

[ethan1998]

I need to know whether a link in a response is truly effective or if it is an ad.

[toolswatcher]

Could this be plugged into blogs or some sites? I often see comments under blogs funneling traffic to gambling or scam sites. Also, have you tried Jev? How well does Jev detect these threats?

[OrdersOfMagnitu]

Very cool! have you found any edge cases where it fails? would be interesting to know