The now-patched vulnerabilities in the rapidly growing AI agent framework allow attackers to steal credentials, escalate privileges, and maintain persistence.

Lire l’article original →