Critical RCE Flaw in Claude Desktop Extensions Endangers 10,000+ Users

A critical security flaw in Claude Desktop Extensions enables zero-click remote code execution (RCE), putting over 10,000 users at severe risk. Security researchers from Foresiet and LayerX have confirmed that the vulnerability exploits the extension architecture embedded within Claude’s desktop application, allowing attackers to execute arbitrary code on a victim’s machine without any user interaction. This means simply having the extension installed and active is enough for an attacker to gain full system control.

Zero-Click Exploitation: No User Action Required

Unlike traditional exploits that rely on phishing links or malicious file downloads, this RCE flaw operates silently in the background. By abusing local API endpoints exposed by Claude’s integrated extensions — such as those connecting to Google Calendar, Outlook, or file-sharing services — attackers can inject and execute malicious payloads without the user clicking anything. LayerX’s analysis shows that the vulnerability leverages the extension’s trust in local system resources, turning routine productivity features into attack vectors.

Google Calendar and Integrated Apps Become Attack Pathways

Computing.co.uk highlights how the integration with widely used services like Google Calendar turns them into Trojan horses. A maliciously crafted calendar event or document shared via Claude can trigger the exploit during normal usage. Security experts liken this to setting your building’s door code to ‘1234’ and leaving the door unlocked — the system assumes trust, and attackers exploit that blind spot. GBHackers reports the flaw carries a CVSS score of 9.8 (Critical), enabling full system compromise, data exfiltration, lateral movement, and persistence.

As of now, no CVE has been officially assigned, but the severity demands immediate action. Users are strongly advised to disable all Claude Desktop Extensions until Anthropic releases a patched version. Organizations using Claude in enterprise environments must conduct endpoint audits and restrict extension permissions to trusted sources only. The window to mitigate this threat is narrow — and delay could lead to widespread compromise.