Threat Intelligence Database

CVE-2026-18684 is a critical command injection vulnerability in the GL.iNet GL-MT3000 device affecting versions 4.4.0 through 4.4.5. The flaw exists in the remove_profile function within the /cgi-bin/glc script of the modem.so component. This vulnerability allows remote attackers to execute arbitrary commands without authentication. The vendor has confirmed the issue, but no official patch or remediation guidance has been provided yet. Public exploit code is available, increasing the risk of exploitation.

GL.iNet GL-MT3000 devices up to version 4.4.5 contain a remote command injection vulnerability in the ovpn-client.get_recommend_config function within the ovpn-client.so plugin. This vulnerability arises from improper handling of the Hostname argument, allowing an attacker to execute arbitrary commands remotely without authentication. The vulnerability has a high severity with a CVSS score of 9.8 and has been publicly disclosed. The vendor has confirmed the issue but no patch information is currently available.

CVE-2026-18616 is a critical command injection vulnerability in the GL-iNet GL-MT3000 device up to version 4.4.5. The flaw exists in the server.set_peer function within the /cgi-bin/glc file of the wg-server.so Native Plugin, where manipulation of the public_key argument allows remote command injection. Exploitation does not require authentication or user interaction, and a public exploit is available. The vendor has confirmed the vulnerability but no official patch or remediation details are provided yet.

A vulnerability was determined in GL-iNet GL-MT3000 up to 4.4.5. The affected element is the function wg-server.generate_publickey of the file /cgi-bin/glc of the component wg-server.so Native Plugin. Executing a manipulation of the argument private_key can lead to command injection. The attack may be launched remotely. The exploit has been publicly disclosed and may be utilized. The vendor was contacted early about this disclosure and confirmed the existence of the vulnerability.

A vulnerability was found in GL-iNet GL-MT3000 up to 4.4.5. Impacted is the function s2s.enable_echo_server of the file /cgi-bin/glc of the component s2s.so Native Plugin. Performing a manipulation of the argument port results in command injection. The attack may be initiated remotely. The exploit has been made public and could be used. The vendor was contacted early about this disclosure and confirmed the existence of the vulnerability.

CVE-2026-18599 is a command injection vulnerability in the GL.iNet GL-MT3000 device up to version 4.4.5. The flaw exists in the logread.set_config function of the Logread Lua RPC Plugin, where manipulation of the record_size argument allows command injection. Exploit code has been published, but no confirmed active exploitation in the wild is reported. The vendor has confirmed the vulnerability's existence but has not provided a patch or official remediation guidance yet.

A security vulnerability has been detected in GL.iNet GL-MT3000 up to 4.4.5. Affected by this vulnerability is an unknown functionality of the file /usr/bin/one_click_upgrade of the component Online Firmware Upgrade Handler. Such manipulation leads to command injection. The attack can be launched remotely. The exploit has been disclosed publicly and may be used. Upgrading to version 4.7 addresses this issue. Upgrading the affected component is advised. The vendor was contacted early, responded in a very professional manner and quickly released a fixed version of the affected product.

A vulnerability has been found in GL.iNet GL-MT3000 up to 4.4.5. Affected is the function FUN_0042e200 of the file /cgi-bin/glc of the component SET_USER_PWD Handler. The manipulation of the argument Password leads to command injection. The attack can be initiated remotely. Upgrading to version 4.8.1 is able to address this issue. The affected component should be upgraded. The vendor explains: " The current code escapes single quotes in the password parameter and handles it inside a shell single‑quote context. The payloads in the report, which rely on $() or backticks to trigger command substitution, are not executed under the current code path. We tested on a GL‑MT3000 device running firmware 4.8.1 using similar payloads, and no command‑execution marker file was created."

A flaw has been found in GL.iNet GL-MT3000 4.4.5. This impacts the function snprintf of the file /cgi-bin/glc of the component FTP Protocol Handler. Executing a manipulation of the argument media_dir can lead to command injection. It is possible to launch the attack remotely. Upgrading to version 4.8.1 will fix this issue. You should upgrade the affected component. The vendor explains: "In version 4.8.1, before writing media_dir to the FTP configuration command, the code escapes single quotes using escape_single_quote(). The payloads in the report—which rely on closing a single quote, appending commands with a semicolon, and commenting out the tail with #—cannot escape execution under the current code path. We also verified this on a GL‑MT3000 device running firmware version 4.8.1 using similar payloads calling the /NAS_API_SET_PROTO_CONFIG interface. Although the interface returned success, the marker file intended to prove command execution was not created; the payload was written into /etc/vsftpd.conf only as ordinary configuration content and did not trigger any shell command execution. Therefore, with the current firmware version and default runtime environment, we could not reproduce the claimed “unauthorized command injection in set_proto_config”."

A vulnerability was detected in GL.iNet GL-MT3000 4.4.5. This affects the function dlopen in the library /usr/lib/oui-httpd/rpc/ of the component Path Normalization Handler. Performing a manipulation of the argument dev_name results in command injection. It is possible to initiate the attack remotely. Upgrading to version 4.7 mitigates this issue. It is advisable to upgrade the affected component. The vendor confirms: " From version 4.7 onward, we have enabled method‑level validation at the HTTP /rpc layer. nas‑web.eject_disk is no longer in the whitelist of allowed methods. Consequently, directly calling eject_disk through the default /rpc endpoint returns Invalid params, preventing entry into subsequent dangerous functions and blocking the remote exploit chain described in the report."