0din-ai ai-scanner Remote Code Execution Vulnerability via JavaScript Injection in BrowserAutomation::PlaywrightService

A remote code execution vulnerability has been identified in 0din-ai's ai-scanner, versions 1.0.0 prior to 1.4.1. The issue arises in the BrowserAutomation::PlaywrightService component, where user-controlled values are injected into Node.js scripts without proper escaping. This allows attackers to execute arbitrary commands in the application's environment. The vulnerability is accessible to any authenticated user and has been patched in version 1.4.1.

Impact

Exploitation of this vulnerability allows authenticated users to execute arbitrary code in the Node.js runtime, with the executed commands having access to sensitive environment variables and database credentials. This could lead to unauthorized data access or modification, and the creation of persistent backdoors in the application.

Reproduction

To reproduce this vulnerability, upload a payload that exploits the JavaScript injection flaw into the 'url' parameter of the '/targets/auto_detect_selectors' endpoint. The payload must be crafted to close the JavaScript string literal and append Node.js commands, such as writing to a file or executing a command. After sending the payload, verify the execution by checking the written file's contents.

Remediation

Users are advised to update ai-scanner to version 1.4.1 or later, where this vulnerability has been patched.