Up to 5,000 users
A software supply chain attack targeted developers using the Cursor IDE, a fork of VS Code. On October 29, 2026, a trojanized version of the popular 'Claude Code' extension from Anthropic was published to the Cursor marketplace. The malicious version, 2.1.313, was disguised as a legitimate update and is estimated to have been installed by up to 5,000 users who had the extension or received it via auto-update. The malware's primary function was to install a persistent SSH backdoor by adding an attacker's public key to the user's ~/.ssh/authorized_keys file, granting the attacker remote access to the compromised development environment. Cursor has since removed the malicious package and is notifying affected users.
The incident represents a classic supply chain attack targeting the software development lifecycle, where trusted tools and update mechanisms are abused to distribute malware.
T1195.002 - Compromise Software Supply Chain). The attackers managed to publish a malicious package under the legitimate publisher ID anthropic.claude-code within the Cursor marketplace.The attack was executed in a simple yet effective manner:
v2.1.313 to the Cursor marketplace. The method used to compromise the publisher account or the publishing process is not detailed.~/.ssh/authorized_keys file on the host system. To appear benign, the key was commented with the label svc-telemetry.This provides the attacker with the ability to log into the victim's machine via SSH without needing a password, effectively creating a permanent backdoor.
T1195.002 - Compromise Software Supply Chain: The attack leveraged a trusted marketplace and extension to distribute malware.T1059.007 - JavaScript: The payload was written in JavaScript, executed by the underlying Node.js engine of VS Code.T1552.004 - Private Keys: The attack modifies the authorized_keys file, which is directly related to the management and use of SSH private/public key pairs for authentication.T1547.001 - Registry Run Keys / Startup Folder: While not a direct match, the extension's auto-execution on IDE startup serves a similar purpose for achieving persistence.The potential impact is severe. Up to 5,000 developers may have had their machines compromised. A backdoor on a developer's machine provides attackers with access to source code, API keys, internal network resources, and other sensitive intellectual property. The swift removal of the package by Cursor and public notification helped limit the window of exposure. However, any user who installed the malicious version remains vulnerable until the backdoor is removed. This incident also damages trust in the security of IDE marketplaces and the extension ecosystem.
No specific file hashes, domains, or IP addresses were mentioned in the source articles. The SSH key itself was not published.
Security teams may want to hunt for the following patterns to identify compromised systems:
file_path~/.ssh/authorized_keysstring_patternsvc-telemetryprocess_namenode or Codeauthorized_keys.log_sourceIDE extension logsanthropic.claude-code extension around the incident date.~/.ssh/authorized_keys. EDR solutions can be configured to detect processes spawned by IDEs that attempt to modify system configuration files. Regularly auditing installed extensions and their permissions is also a crucial step.~/.ssh/authorized_keys file and remove any unrecognized public keys.D3-PA: Process Analysis.Use endpoint security solutions to detect and block malicious scripts or unauthorized file modifications.
Implement policies to restrict the execution of unsigned or untrusted code within development environments.
To counter threats like the trojanized 'Claude Code' extension, organizations must implement continuous System File Analysis, specifically focusing on critical configuration files on developer endpoints. This involves using a File Integrity Monitoring (FIM) tool to create a baseline of sensitive files, such as ~/.ssh/authorized_keys, and immediately alerting on any unauthorized modifications. The rule should be highly specific: trigger an alert if any process associated with the IDE (e.g., Code, node) writes to this file. The alert should capture the process, user, and the exact change made. This provides security operations with a high-fidelity signal of a potential backdoor installation, enabling rapid investigation and response before an attacker can use the newly established access.
In a development environment, a strict executable allowlisting policy can be challenging. However, a more targeted approach can be effective. Organizations should maintain a repository of approved and vetted IDE extensions. A policy should be enforced, either technically or administratively, that prevents developers from installing extensions from un-vetted sources. For marketplaces like Cursor or VS Code, security teams can curate a list of approved extensions and their specific, known-good versions. Scripts can be used to audit installed extensions against this allowlist. This moves the organization from a reactive posture to a proactive one, preventing the installation of malicious extensions in the first place and reducing the risk of supply chain attacks.
Malicious version 2.1.313 of 'Claude Code' extension becomes available in the Cursor marketplace.
The compromised package is removed from the Cursor marketplace.

Cybersecurity professional with over 10 years of specialized experience in security operations, threat intelligence, incident response, and security automation. Expertise spans SOAR/XSOAR orchestration, threat intelligence platforms, SIEM/UEBA analytics, and building cyber fusion centers. Background includes technical enablement, solution architecture for enterprise and government clients, and implementing security automation workflows across IR, TIP, and SOC use cases.
CyberNetSec.io uses automation to assist source monitoring, deduplication, observable extraction, and structured intelligence generation. Published analysis follows human-defined editorial standards and adds defensive context including MITRE ATT&CK, D3FEND, STIX, and Sigma where applicable. Read our editorial policy.
Help others stay informed about cybersecurity threats
Every tactic, technique, and sub-technique used in this threat has been identified and mapped to the MITRE ATT&CK framework for consistent, actionable threat language.
Observables and indicators of compromise (IOCs) have been extracted and cataloged. Risk has been assessed and correlated with known threat actors and historical campaigns.
Detection rules, incident response steps, and D3FEND-aligned mitigation strategies are included so your team can act on this intelligence immediately.
Structured threat data is packaged as a STIX 2.1 bundle and can be visualized as an interactive graph — relationships between actors, malware, techniques, and indicators.
Sigma detection rules are derived from the threat techniques in this article and can be converted for deployment across any major SIEM or EDR platform.