Specially crafted string can cause a buffer overrun in the JSON parser library leading to a crash of the Zabbix Server or a Zabbix Proxy.

JavaScript pre-processing can be used by the attacker to gain access to the file system (read-only access on behalf of user "zabbix") on the Zabbix Server or Zabbix Proxy, potentially leading to unauthorized access to sensitive data.

JavaScript preprocessing, webhooks and global scripts can cause uncontrolled CPU, memory, and disk I/O utilization. Preprocessing/webhook/global script configuration and testing are only available to Administrative roles (Admin and Superadmin). Administrative privileges should be typically granted to users who need to perform tasks that require more control over the system. The security risk is limited because not all users have this level of access.

A Firewall Rule which allows all incoming TCP connections to all programs from any source and to all ports is created in Windows Firewall after Zabbix agent installation (MSI)

Zabbix Frontend provides a feature that allows admins to maintain the installation and ensure that only certain IP addresses can access it. In this way, any user will not be able to access the Zabbix Frontend while it is being maintained and possible sensitive data will be prevented from being disclosed. An attacker can bypass this protection and access the instance using IP address not listed in the defined range.

Arbitrary file read vulnerability exists in Zabbix Web Service Report Generation, which listens on the port 10053. The service does not have proper validation for URL parameters before reading the files.

An unauthenticated user can create a link with reflected Javascript code inside the backurl parameter and send it to other authenticated users in order to create a fake account with predefined login, password and role in Zabbix Frontend.

An authenticated user can create a link with reflected Javascript code inside it for the graphs page and send it to other users. The payload can be executed only with a known CSRF token value of the victim, which is changed periodically and is difficult to predict.

An authenticated user can create a link with reflected Javascript code inside it for the discovery page and send it to other users. The payload can be executed only with a known CSRF token value of the victim, which is changed periodically and is difficult to predict.

An authenticated user can create a link with reflected Javascript code inside it for services’ page and send it to other users. The payload can be executed only with a known CSRF token value of the victim, which is changed periodically and is difficult to predict. Malicious code has access to all the same objects as the rest of the web page and can make arbitrary modifications to the contents of the page being displayed to a victim during social engineering attacks.