Threat
Anthropic 'SneakPeek' Flaw Is a Silent Zero-Click Threat
First reported darkreading.com
Page published
Earliest dated coverage: 14 Oct 2026 · First observed: 14 Oct 2026 · Latest dated coverage: 14 Oct 2026
Coverage timeline
Single-source incident — one report is available.
Why it matters
SneakPeek shows how a trusted AI coding assistant's preview feature can be weaponized for zero-click credential theft, turning routine repository previews into a silent data-exfiltration channel on developer machines.
The 'SneakPeek' vulnerability in Anthropic's Claude Code Desktop lets adversaries silently exfiltrate sensitive developer files — including SSH private keys, AWS credentials, and Claude OAuth tokens — via the automated HTML preview pane with no prompt or click. Noma Labs chained three flaws (missing content security policy, unrestricted origin access, and inactive request filtering) to bypass all sandbox protections; an attacker could plant a public repo with an HTML file padded past 256KB containing a malicious script. Anthropic's initial fix was bypassed by Noma, prompting deeper architectural fixes.