django‑pyas2 through 1.2.3 allows an authenticated administrator to inject arbitrary operating system commands via the cmd_receive and cmd_send fields in the Partner model. The values of these fields are written straight into os.system() without sanitization, giving the attacker the ability to execute any shell command on the host that runs the application. This flaw is a classic operating‑system command injection vulnerability, classified as CWE‑78, and can lead to full system compromise, including data theft, tampering, or denial of service.

The vulnerable component is the open‑source django‑pyas2 library, versions up to and including 1.2.3. The project is hosted on GitHub and is typically integrated into custom AS2‑enabled services that rely on Django.

The CVSS score of 9.1 marks this flaw as critical. The EPSS score of approximately 1.23% indicates that, at the time of assessment, the likelihood of exploitation in the wild was low, but the flaw remains uncatalogued in CISA KEV. Exploitation requires an authenticated administrator—an attacker with such credentials can trigger command injection whenever an AS2 message is received or sent. Due to the severe impact and the need for authentication, the attack vector is internal but potentially extendable if admin privileges are obtained through phishing or credential compromise.