A critical OS command injection vulnerability in Roo-Code allows unauthenticated attackers to achieve remote code execution if specific user configurations are met.

The application fails to properly neutralize special characters, such as process substitution and ampersands, within its command parsing logic. This allows an unauthenticated attacker to inject and execute arbitrary OS commands when the user has enabled auto-approved execution for commands.

Successful exploitation allows an attacker to execute arbitrary commands on the host machine with the privileges of the user running the editor. Given the CVSS score of 8.1, this flaw poses a severe risk of unauthorized system access, data exfiltration, and potential compromise of the local development environment.

Immediate Action: Update Roo-Code to version 3.25.5 or later to apply the necessary command parsing sanitization.

Proactive Monitoring: Review application logs for unusual command execution patterns or unexpected process initiation linked to the Roo-Code agent.

Compensating Controls: Ensure that the auto-approved execution feature is disabled, which serves as a critical defense-in-depth measure against this injection vector.

Public Exploit Available: No (exploit_available unknown)

The severity of this command injection vulnerability necessitates immediate attention to prevent potential system compromise. Administrators and users must prioritize upgrading to version 3.25.5 and should maintain strict security policies regarding the automated execution of AI-generated commands to minimize the attack surface.