A critical vulnerability in the Cursor code editor allows authenticated attackers to achieve Remote Code Execution by injecting malicious workspace configurations.

This vulnerability involves improper control of code generation (CWE-94) where an attacker who compromises the chat context, such as through a malicious MCP server, can use prompt injection to modify workspace files. This action bypasses existing security controls and allows for the execution of arbitrary code within the user environment.

The ability for an attacker to execute arbitrary code on a developer machine presents a severe risk to organizational intellectual property and production environments. A successful exploit could lead to the theft of source code, the insertion of backdoors into software projects, or unauthorized access to cloud infrastructure credentials stored locally. Given the CVSS score of 7.5, this high severity flaw warrants immediate attention to prevent lateral movement within the software development lifecycle.

Immediate Action: Update the Cursor application to version 1.7 or higher immediately to apply the necessary security fixes.

Proactive Monitoring: Monitor developer workstation logs for unusual file modifications, particularly changes to .code-workspace files or unexpected outbound network connections initiated by the editor.

Compensating Controls: Exercise caution when interacting with untrusted AI agents or third-party MCP servers, as these represent the primary attack vector for this vulnerability.

Public Exploit Available: Unknown

The severity of this vulnerability necessitates a swift response from security teams and individual developers. Organizations should enforce a mandatory update policy for all instances of the Cursor editor to version 1.7 or later. By ensuring the software is patched, teams effectively neutralize the primary vector for this remote code execution flaw and protect their development environments from potential compromise.