A flaw in the notification rendering logic allows malicious code to be executed when a notification template is processed. The sandboxed evaluation function mistakenly trusts the outer code object’s "co_names" and ignores nested lambda expressions, letting an attacker inject arbitrary Python statements that bypass the whitelist. Running such code would give the attacker full control of the host system.

All installations of the monitoring tool that run Tautulli versions older than 2.17.0 are affected; the issue was addressed in the 2.17.0 release.

The severity score of 7.5 reflects a high-risk vulnerability, while the probability of exploitation is very low, with an EPSS score below 1%. The flaw is not listed in CISA’s known exploited vulnerabilities catalog. Based on the description, it is inferred that the attack requires the ability to create or modify a notification template containing malicious code, after which the code is executed during rendering, granting the attacker full system compromise.