Security Affairs reports that a malicious version of Anthropic.claude-code (2.1.313) has been identified in the Cursor Extension Marketplace. The notice describes the code as a security threat that could be deployed via signed ChatGPT and Claude binaries, potentially affecting users who install the extension. The article places the finding among other malware trends, including a virtualization-based Android spyware platform targeting messaging apps and a suspected crypto-theft implant in Trezor hardware wallets. The report does not detail a specific remediation but highlights the presence of the compromised Claude code in a legitimate-looking marketplace, underscoring the risk of malicious binaries masquerading as trusted AI tools. 1
Published by SecurityAffairs · On GETSECURE.ai ·
SECURITY AFFAIRS MALWARE NEWSLETTER ROUND 121
⚠️ Critical Alert
Why This Matters
If you use the Cursor Extension Marketplace, you should verify that any Claude-code (2.1.313) you install is authentic and not the malicious version flagged by Security Affairs.
Title card · GETSECURE.ai
References
- SecurityAffairs. (2026, November 1). SECURITY AFFAIRS MALWARE NEWSLETTER ROUND 121. SecurityAffairs. https://securityaffairs.com/202209/malware/security-affairs-malware-newsletter-round-121.html
Original Source
SecurityAffairs
Read original →