A critical unrestricted file upload vulnerability in PromtEngineer localGPT exposes the system to remote exploitation through the backend server.

The application fails to properly validate file uploads within the do_POST function of backend/server.py, allowing an unauthenticated remote attacker to bypass access controls and upload malicious files.

The ability to perform unrestricted file uploads presents a severe risk of remote code execution, which could lead to a full system compromise. Given the CVSS score of 7.3, this vulnerability is classified as High severity, as it enables attackers to inject malicious scripts or backdoors into the application environment.

Immediate Action: Since no official patch is available, immediately restrict network access to the affected backend server to prevent unauthorized interaction with the vulnerable endpoint.

Proactive Monitoring: Review web server logs for irregular POST requests directed at the backend/server.py file and monitor the file system for unexpected new files.

Compensating Controls: Implement a Web Application Firewall (WAF) rule to intercept and block file upload attempts targeting the vulnerable endpoint until a formal security update is released.

Public Exploit Available: Yes, a published proof-of-concept exists as referenced in the GitHub issue tracker associated with this vulnerability.

Due to the absence of a vendor-supplied patch and the presence of a public proof-of-concept, users must prioritize isolating the vulnerable instance. Organizations should restrict access to the localGPT server to trusted internal networks only and verify that current file upload handling is effectively blocked by existing security controls to minimize the risk of unauthorized code execution.