TrustTunnel is an open‑source VPN protocol that allows a server‑side request forgery when a numeric IP is supplied. The SSRF protection is only applied to hostname destinations; numeric IP destinations bypass the check for private or loopback addresses, which lets an attacker reach internal services even when the server is configured to deny private network connections. This flaw is a CWE‑918 vulnerability with a CVSS score of 7.1, indicating that an attacker could achieve confidentiality or integrity violations on the vulnerable host or on internal network assets.

All installations of TrustTunnel by AdGuard running a version earlier than 0.9.114 are affected. The affected product is the TrustTunnel VPN server, which can be hosted on any platform where TrustTunnel is deployed.

The CVSS base score of 7.1 denotes a high severity issue, while the EPSS score of less than 1 % indicates a very low current exploitation probability. The flaw is not listed in the CISA KEV catalog, and no known public exploitation has been reported. The likely attack vector involves an attacker crafting a connection request to the TrustTunnel server that includes a numeric IP destination; because the SSRF validation is omitted for such destinations, the server will connect to the supplied private or loopback address, potentially exposing internal services or facilitating lateral movement. Given the default configuration of allow_private_network_connections = false, any client with access to the server’s forwarding interface can exploit this weakness.