As cited
Copy frozen at (site build).
ai security
Anthropic 'SneakPeek' Flaw Is a Silent Zero-Click Threat
A flaw in Claude Code Desktop's preview pane enables attackers to silently exfiltrate developer credentials, including SSH private keys, AWS credentials, and Claude OAuth tokens. The vulnerability operates as a zero-click attack requiring no user interaction. The issue affects a widely-used artificial intelligence (AI) development tool.
Why it matters: Developers using Claude Code Desktop face credential theft without visible warnings, exposing cloud infrastructure, code repositories, and AI platform access. Security teams should assess whether the tool is in use and apply patches immediately.
- Source published
- First seen by Cybersecurity Tracker