Hey everyone! If you code with AI tools, this one is worth two minutes of your time. On Nov 13, Cursor published the results of its two-week investigation into the security incident from Oct 29, when a malicious copy of Anthropic's Claude Code extension showed up in Cursor's extension marketplace. What happened, according to Cursor: - It started at Fireworks, an AI vendor. An access key Cursor had issued to Fireworks was exposed, and the attacker used it to copy old coding agent traces (mostly from 2025, only from sessions not in Privacy Mode). - A few of those traces had unredacted credentials in them. The attacker used those to get into some of Cursor's SaaS tools and then planted the fake extension (versions 2.1.312 and 2.1.313). - The fake build was live for about 11 hours and reached at most 5,000 users. It tried to install a hidden SSH backdoor. - Anthropic didn't build or publish it, and it wasn't a flaw in Anthropic's product. Cursor says the incident is contained, the last attacker activity was Nov 1, and it has contacted every affected customer directly. - Cursor moved all inference off Fireworks, rotated its credentials, and rebuilt its extension marketplace. Why it matters: AI coding tools see your code, your terminal and often your API keys. One leaked key at a vendor turned into malware on developers' machines. If you installed either of those versions, Cursor says uninstalling alone may not be enough: follow its security notice and rotate your keys. Question: Do you ever paste API keys or passwords into an AI chat or coding agent? Has this changed how you'll handle them? Sources: - Cursor (Nov 13 update): https://cursor.com/blog/security-incident-update - Cursor security notice: https://cursor.com/blog/claude-code-extension-security-notice - BleepingComputer: https://www.bleepingcomputer.com/news/security/cursor-confirms-security-breach-led-to-malicious-claude-code-extension/