CVE-2025-61592
Published: 03 October 2025
Description
Adversaries may exploit software vulnerabilities in client applications to execute code.
Security Summary
CVE-2025-61592 is a high-severity vulnerability (CVSS 8.8, CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H) affecting Cursor, an AI-powered code editor for programming, in versions 1.7 and below. The issue stems from the Cursor CLI's automatic loading of project-specific configuration from the current working directory file .cursor/cli.json, which can override certain global configurations. This flaw, classified under CWE-829 (Inclusion of Functionality from Untrusted Control Sphere), enables remote code execution when combined with permissive configuration settings that allow shell commands and prompt injection delivered through project-specific rules in .cursor/rules/rule.mdc or other mechanisms.
The attack requires a user to run the Cursor CLI inside a malicious repository, typically after cloning it via social engineering or other user interaction (UI:R). No privileges are needed from the attacker (PR:N), who can achieve network-accessible exploitation (AV:N) with low complexity (AC:L). Successful exploitation leads to high-impact compromise of confidentiality, integrity, and availability (C:H/I:H/A:H) through arbitrary code execution on the victim's system.
The official advisory is available at https://github.com/cursor/cursor/security/advisories/GHSA-v64q-396f-7m79. Mitigation is provided via patch 2025.09.17-25b418f, released prior to the CVE publication on October 3, 2025, though no stable release version incorporates the fix as of that date.
Details
- CWE(s)
Affected Products
AI Security Analysis
- AI Category
- Enterprise AI Assistants
- Risk Domain
- LLM/Generative AI Risks
- OWASP Top 10 for LLMs 2025
- None mapped
- MITRE ATLAS Techniques
- None mapped
- Classification Reason
- Cursor is an AI-powered code editor designed for programming with AI, classifying it as an Enterprise AI Assistant due to its integration of AI for developer workflows.
MITRE ATT&CK Enterprise Techniques
Why these techniques?
The vulnerability enables remote code execution via malicious project-specific CLI configuration (.cursor/cli.json) and rules (.cursor/rules/rule.mdc) files that override globals and leverage prompt injection to execute shell commands, facilitating proxy execution via the trusted developer utility Cursor CLI and exploitation of a client application.