HaiAI123

Curated Global AI Tools Directory

Industry update
↓ 32%

Cursor reviews the compromised Claude Code extension incident, urges reinstall and credential rotation

11/14/2026 — 11/16, 09:17·4 sources (all)·5 reports (all)

Key takeaways

  • Cursor has released a retrospective on the October 29 incident involving a tampered Claude Code extension, saying attackers used keys stolen…

Background & analysis

On October 29, 2026, Cursor announced that a malicious and unauthorized copy of Anthropic's Claude Code extension had appeared in its VS Code extension marketplace. Tampered builds of Anthropic.claude-code 2.1.312 and 2.1.313, for both Windows and macOS, were served from the marketplace for roughly 11 hours. The extension displayed Anthropic as its author, which led some developers to assume the company's account had been compromised; Cursor later clarified that attackers had published the fake extension by impersonating Cursor.

On November 13, 2026, after a two-week investigation, Cursor disclosed the findings, stating that the incident affected at most 5,000 users and that the vast majority of customers were untouched. Users running in privacy mode saw no exposure of prompts or code. Among users who did not use privacy mode, fewer than 0.05% had programming-agent trace data leaked, and Cursor said it contacted every affected customer directly with the relevant details. On November 14, Cursor published an update on X summarizing the incident and the remediation steps taken, though the post did not detail the blast radius, the timeline, or the root cause.

According to Cursor's retrospective and reporting by BleepingComputer, the attackers first used access keys stolen from Fireworks AI to copy data from a shared read-only storage bucket. That bucket held a small number of Cursor coding-agent records, mostly sessions from 2025 that ran without privacy mode and were permitted to be used for model training, and it contained unmasked credentials affecting fewer than 0.01% of customers. With those internal credentials, the attackers reached some Cursor SaaS tooling, obtained the keys controlling the extension marketplace system, and pushed the tampered Claude Code extension into the marketplace cache. Cursor's security notice lists indicators of compromise and cleanup scripts, and recommends that affected users reinstall their systems and rotate all credentials. Cursor added that it is drawing lessons from the incident to strengthen its security.

AI-generated from 5 reports · updated 1 day ago

Latest turnCursor published a postmortem of the October 29 malicious Claude Code extension incident, saying attackers first obtained an access key it had shared with Fireworks AI and then used internal credentials to push a tampered extension into its marketplace cache. The malicious version was live for 11 hours and was downloaded by roughly 5,000 developers; Cursor advises affected users to reinstall their systems and rotate all credentials.

Related AI tools

updream
24-hour heatHeat index 46 · peak 167 · 23h ago
24 hours agonow

Reports on this story headlines open the original

Yesterday
  1. Cursor published a postmortem of the October 29 malicious Claude Code extension incident, saying attackers first obtained an access key it had shared with Fireworks AI and then used internal credentials to push a tampered extension into its marketplace cache. The malicious version was live for 11 hours and was downloaded by roughly 5,000 developers; Cursor advises affected users to reinstall their systems and rotate all credentials.

    蓝点网AI score 74
Nov 15
  1. According to BleepingComputer, Cursor has confirmed a breach in which attackers used credentials stolen from Fireworks AI to publish and distribute a malicious Claude Code extension through its marketplace. Public details remain limited, but the incident highlights how fragile trust is across developer tooling and the AI supply chain.

    DEV Community · AIAI score 64
  2. Tampered copies of Anthropic's Claude Code extension (versions 2.1.312 and 2.1.313 for Windows and macOS) were served from Cursor's VS Code extension marketplace for about 11 hours on October 29. Drawing on Cursor's security notice and two independent analyses, the post walks through who was exposed, how to check a machine, and which extension controls teams should add.

    DEV Community · AIAI score 66
Nov 14
  1. Cursor has released the findings of its investigation into the malicious Claude Code extension listed in its VS Code marketplace on October 29, saying up to 5,000 users were affected while the vast majority were not. Prompts and code were unaffected for users in privacy mode, and under 0.05% of non-privacy-mode users had agent trace data leaked; those users have been notified individually.

    IT之家AI score 58
  2. Cursor said on X that it has published an update on its recent security incident, including a detailed overview of what happened and the remediation steps taken. The post itself did not disclose the scope, timeline, or cause, pointing readers to the full write-up.

    X:Cursor (@cursor_ai)AI score 52

Other stories people are talking about

How is heat calculated?About the method

Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.

This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.

Surge
Discussion rising fast
New
First report within 6 hours
Rising
Still gathering discussion

Back to AI industry updates →