The Shell Runtime component of Nuclio accepts a custom HTTP header, X‐Nuclio‐Arguments, which is directly concatenated into shell commands without sanitization. This omission allows an attacker who can send an HTTP request to the function endpoint to inject arbitrary shell instructions, thereby achieving remote code execution and potentially escalating privileges on the host that runs the Nuclio server. The vulnerability is an instance of command injection originating from insecure handling of user input.

Nuclio, Version 1.15.20 introduced the fix. All versions of Nuclio older than 1.15.20 are affected. The flaw exists within the Nuclio Shell Runtime component that processes HTTP requests to user‑defined serverless functions. Vendors or users deploying versions older than 1.15.20 in any environment—including Kubernetes, Docker, or bare‑metal installations—are susceptible to exploitation.

The CVSS score of 8.9 indicates a high severity with potential for complete system compromise. The EPSS value of less than 1 % suggests that, although the flaw is severe, its exploitation frequency remains low, possibly due to a limited attack surface. Nuclio does not list this vulnerability in the CISA KEV catalog, but the lack of exposure does not reduce the risk for an attacker who can reach the vulnerable HTTP endpoint. Exploitation requires the ability to send crafted HTTP requests to a function; normally the X‑Nuclio‑Arguments header can be controlled by an authenticated or unauthenticated user depending on the function’s access settings, implying that privilege escalation is possible if the function runs with elevated rights.