Anthropic says malicious Claude Code extension in Cursor marketplace was not its build
A malicious copy of the Claude Code extension in Cursor's marketplace reached at most 5,000 users, and Anthropic says it did not build or publish it.
Cursor says a malicious copy of the Claude Code extension in its marketplace affected at most 5,000 users. According to a post from @cursor_ai, Cursor pulled the marketplace, neutralized the malware and reinstated it, and told users to update the extension to v2.1.314. Anthropic says its own systems were not compromised and that it did not build or publish this extension (@felixrieseberg).
What changed
Users are told to update the Claude Code extension in Cursor's marketplace to v2.1.314 after a malicious copy was pulled and the marketplace cleaned up.
Type: Security · Applies to: Claude Code extension, Cursor marketplace
Why it matters
Teams that install AI coding extensions from third-party marketplaces face supply-chain risk even when the vendor's own systems are safe.
What you can do
Engineering leads should audit installed AI coding extensions, confirm they come from verified publishers, and update to the patched version.
Original source
We have no link to the original source for this story yet.
Where we saw it
Latent Space
AI wrote this summary from the sources listed, and nobody reviewed it individually before publication. Check the original source before you rely on it. How it works