Is this vulnerability remotely exploitable?

It can be, if an attacker can influence the prompts the Cursor Agent processes—either by directly interacting with the user and agent, or by placing malicious instructions in project files, documentation, or external content that the agent reads. The actual commands run on the user's machine, but the attacker does not need local system access.

What level of access does an attacker need to exploit this?

The attacker needs the ability to inject or influence natural-language instructions that the agent will consider, while the agent is in Auto-Run Mode with allowlist mode enabled and shell access configured. They do not necessarily need system credentials; the risk comes from controlling the agent's input rather than logging into the machine.

What is the safest general fix pattern?

The most robust fix pattern is to funnel all shell interactions—including built-ins—through a tightly controlled mediation layer that enforces explicit allowlists and applies special scrutiny or hard blocking to environment-altering operations. Additionally, isolating the agent in a sandboxed environment and limiting the environment variables it can read or write significantly reduces impact.

Severity:

CRITICAL RISK

(9.8)

Shell Environment Manipulation via AI Agent Auto-Run (CWE-15) in Cursor Editor

CVE-2026-22708

Published:

Jan 14, 2026

Status:

Analyzed

Summary

Affected Context

Exploit Preconditions

Why This Happens

Potential Impact

General Mitigation Guidance

How Teams Detect Issues Like This

Frequently Asked Questions

Source

NDV

Recent Vulnerabilities

CWE-77;CWE-78

HIGH RISK

HIGH

(7.3)

OS Command Injection (CWE-78) in Totolink A7100RU setGameSpeedCfg CGI Handler

Remote Code Execution on the Router

April 6, 2026

CWE-77;CWE-78

HIGH RISK

HIGH

(7.3)

OS Command Injection (CWE-78) in Totolink A7100RU setFirewallType Handler

Remote Code Execution on Router Firmware

April 6, 2026

Vulnerability Type

Category:

Improper Control of Environment / Shell Execution by AI Agent

CWE:

CWE-15;CWE-74;CWE-77;CWE-78;CWE-94;CWE-269

Impact:

Privilege-Scoped Command Manipulation and Potential Remote Code Execution

Severity:

CRITICAL RISK

(9.8)

CVSS Version:

3.1

Vulnerability Type

Category:

Improper Control of Environment / Shell Execution by AI Agent

CWE:

CWE-15;CWE-74;CWE-77;CWE-78;CWE-94;CWE-269

Impact:

Privilege-Scoped Command Manipulation and Potential Remote Code Execution

Severity:

CRITICAL RISK

(9.8)

CVSS Version:

3.1

Catch This CVE While Code Is Being Written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Get Started

Ship clean & secure code faster

START 14 DAYS FREE TRIAL

BOOK A DEMO

Made with Love in San Francisco

355 Bryant St. San Francisco, CA 94107, USA

Ask AI for summary of CodeAnt

Made with Love in San Francisco

355 Bryant St. San Francisco, CA 94107, USA

Ask AI for summary of CodeAnt

Made with Love in San Francisco

355 Bryant St. San Francisco, CA 94107, USA

Ask AI for summary of CodeAnt

Shell Environment Manipulation via AI Agent Auto-Run (CWE-15) in Cursor Editor

Summary

Affected Context

Exploit Preconditions

Why This Happens

Potential Impact

General Mitigation Guidance

How Teams Detect Issues Like This

Frequently Asked Questions

Source

Recent Vulnerabilities

HIGH RISK

HIGH

(7.3)

OS Command Injection (CWE-78) in Totolink A7100RU setGameSpeedCfg CGI Handler

HIGH RISK

HIGH

(7.3)

OS Command Injection (CWE-78) in Totolink A7100RU setFirewallType Handler

Vulnerability Type

Vulnerability Type

Vulnerability Type

Catch This CVE While Code Is Being Written

Ship clean & secure code faster

Product

Pricing

Company

Legal

Developers

Compare

Product

Pricing

Company

Legal

Developers

Compare

Product

Pricing

Company

Legal

Developers

Compare