An official Anthropic Claude extension for the Cursor IDE was compromised for nearly 12 hours, potentially affecting up to 5,000 developers. The malicious version of the extension attempted to install a persistent SSH backdoor and steal cloud credentials and API tokens. Cursor has since removed the extension and rotated its marketplace keys, though Anthropic has not yet issued a warning to its users. AI
IMPACT Compromised AI tool extensions can lead to widespread credential theft and system breaches for developers.
RANK_REASON Security vulnerability in a third-party extension for an AI-related tool.
Read on Mastodon — mastodon.social →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →