G github.com Can open-source prompt-injection detectors catch realistic AI agent attacks? 8 points northbridgedev 1 month ago 4 comments
