Trail of Bits has demonstrated that skill scanners, a new class of tools designed to detect malicious AI agent skills before installation, can be bypassed using simple, well-known techniques. In a coordinated disclosure, the team successfully circumvented ClawHub’s malicious skill detector, Cisco’s open-source skill-scanner, and all three scanners integrated into Vercel’s skills.sh marketplace in under […]
The post Malicious Skill Detector Bypassed on ClawHub, Cisco, and Vercel appeared first on Cyber Security News.