NEWS
SkillCloak research exposes AI agent skill scanner evasion risk
Fresh research shows malicious AI agent skills can evade static scanners, pushing defenders toward runtime sandboxing and stricter skill governance.
Tag archive
All visible posts matching this normalized tag.
NEWS
Fresh research shows malicious AI agent skills can evade static scanners, pushing defenders toward runtime sandboxing and stricter skill governance.
Jul 02, 2026
5 itemsSource-backed July 1 digest covering active exploitation, critical enterprise patching, AI/browser attack research, and fake-PoC supply-chain abuse.
NEWS
A June 30 report says attackers exploited SimpleHelp CVE-2026-48558 to gain RMM technician access and deliver TaskWeaver plus Djinn Stealer.
NEWS
Microsoft temporarily disabled repositories after reports of Miasma worm activity targeting developer IDE and AI coding tool workflows.
NEWS
Kaspersky and CyberPress report a SharkLoader campaign that abuses public exploits and fake installers to sideload Cobalt Strike onto government and enterprise systems.