AIThe Next Webabout 17 hours ago
Malicious AI skills stole credentials at scale
1 min read
Zenity Labs found credential-stealing skills on Vercel's skills.sh registry, with one typosquatted family racking up over 1.7 million installs. The fake skills hid malicious instructions that made AI agents exfiltrate SSH keys, cloud credentials, and tokens. Vercel and GitHub removed the offending listings within 12 hours.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
