CVE-2025-32018 Arbitrary file write from Cursor Agent through a prompt injection from malicious @Docs
An attacker can trick the AI agent into writing files outside the workspace via prompt injection, though user review mitigates risk.
- Exploited in the wild
- Not known
- Login needed
- Any signed-in user
- Attackable
- Over the internet
- Victim action needed
- A user has to click or open something
- Fix
- None listed
What is CVE-2025-32018?
CVE-2025-32018 is a high-severity path traversal vulnerability in getcursor cursor, published on 8 Apr 2025. Any signed-in user can exploit it over the network, and only if a user opens a file or follows a link.
Official description
Cursor is a code editor built for programming with AI. In versions 0.45.0 through 0.48.6, the Cursor app introduced a regression affecting the set of file paths the Cursor Agent is permitted to modify automatically. Under specific conditions, the agent could be prompted, either directly by the user or via maliciously crafted context, to automatically write to files outside of the opened workspace. This behavior required deliberate prompting, making successful exploitation highly impractical in real-world scenarios. Furthermore, the edited file was still displayed in the UI as usual for user review, making it unlikely for the edit to go unnoticed by the user. This vulnerability is fixed in 0.48.7.
Weakness: CWE-22
Which versions are affected by CVE-2025-32018?
How do I fix CVE-2025-32018?
No fix has been published for CVE-2025-32018 yet. Limit exposure of the affected software until one is available.
Upgrade, and resolve every user-supplied path against an allowed base folder before opening it. More about Path traversal.
Is CVE-2025-32018 being exploited?
There is no known exploitation of CVE-2025-32018 in the wild and no public exploit code.