This vulnerability involves improper handling of input in the eval function within premsql/agents/baseline/workers/followup.py in premAI-io premsql versions 0.2.0 and 0.2.1. Manipulating the 'result' argument enables remote code injection. The attack vector is network-based with low attack complexity and no required privileges or user interaction. The vendor has not issued any advisory or patch, and the exploit is publicly accessible.

Successful exploitation could allow an attacker to execute arbitrary code remotely on affected systems running premAI-io premsql up to version 0.2.1. This could compromise system integrity and confidentiality. The CVSS score of 5.3 reflects a medium severity impact with limited scope and privileges required. No known exploits in the wild have been reported beyond the public availability of exploit code.

Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. Since the vendor has not responded or issued a fix, users should consider mitigating exposure by restricting network access to the vulnerable service and monitoring for suspicious activity related to the eval function. Avoid using affected versions 0.2.0 and 0.2.1 until a vendor fix or official guidance is available.