On this page (8 sections)

A critical flaw, tracked as CVE-2026-124002, affects XWEB EVO versions 5.2.3 through before 5.7.0. According to the GitHub Advisory Database entry, an unauthenticated remote attacker can run operating system commands by injecting shell metacharacters into the username field of three scripts, before any password check takes place.

If you run, host or manage an XWEB EVO device, or any network where one is reachable, this one needs attention. The advisory also states that the product line reached its end of service date in March 2026, which affects how much you can count on a vendor fix.

What happened

On November 12, 2026, the flaw was published by the National Vulnerability Database and mirrored in the GitHub Advisory Database, which marks its own entry as unreviewed. The description says the injection happens in the username field of login.lua, cleanout.lua or api.lua, and that it occurs before the password is checked. In plain terms, an attacker does not need an account, a stolen password or any user interaction.

The advisory adds a second ingredient: the setuid-root wlua interpreter. A setuid-root program runs with the privileges of the root user regardless of who starts it. According to the description, combining the injection with that interpreter lets an attacker run commands as root, which it describes as full system compromise.

Only one independent publisher covers this story so far. The NVD record repeats the same text as the GitHub entry, so treat them as one source, not two confirmations. The advisory also points to a Copeland software update page and a VulnCheck advisory, which are the places to look for vendor-side details.

Why this class of bug is dangerous

OS command injection happens when a program builds a shell command out of user-supplied text without neutralizing special characters such as semicolons, pipes or backticks. The shell then treats part of the input as instructions. It is a well-understood weakness (CWE-78), and it is most serious when it sits in front of authentication, which is why CWE-306 is also listed.

The CVSS v4 metrics in the GitHub entry match that picture: network attack vector, low attack complexity, no attack requirements, no privileges required, no user interaction, and high impact on confidentiality, integrity and availability of the vulnerable system. A device like this, once taken over as root, could be used to alter its configuration, read stored data, or serve as a foothold for reaching other systems on the same network. The research does not describe any real-world attack, so those outcomes are general consequences of root access, not reported incidents.

Who is exposed

  • Anyone running XWEB EVO at version 5.2.3 up to, but not including, 5.7.0.
  • Devices whose web interface can be reached from the internet, a shared corporate network, a vendor VPN or a flat network where many hosts can talk to it.
  • Teams that inherited a device from an installer or integrator and are unsure which version it runs or where it is exposed.
  • Hosting and network administrators who may have such a device sitting in a rack or a remote site and not in an asset inventory.
The research does not say whether devices on version 5.7.0 or later are fixed, because the GitHub entry lists patched versions as unknown. The affected range simply ends before 5.7.0. Confirm with the vendor before assuming a given release is safe.

How to check your exposure

Start with inventory. You need to know which devices exist, which version each runs and who can reach them over the network. Check the version in the device's own administration or system information page, or in your installer's documentation, and compare it to the range above.

Next, confirm reachability from outside. From a machine on an untrusted network, such as a mobile connection, test whether the device's web port answers. The following command only attempts a connection and prints whether it succeeded. Replace the placeholder with your device's address and port; a successful connection means the interface is reachable from where you ran it.

Check whether a device's web port is reachable from your current network (Linux/macOS, uses netcat)Shell
nc -vz -w 5 DEVICE_ADDRESS 443

If you run a firewall or reverse proxy in front of the device, review its access logs for unusual requests to login.lua, cleanout.lua or api.lua, especially with odd characters in the username. This is a read-only search. Adjust the log path to your own setup; hits are worth investigating, though normal logins will also match the file names.

Search a web or proxy access log for requests to the affected scripts (adjust the log path)Shell
grep -E 'login\.lua|cleanout\.lua|api\.lua' /var/log/nginx/access.log | tail -n 200

What to do now

  1. List every XWEB EVO device and record its version. Treat anything from 5.2.3 up to before 5.7.0 as vulnerable.
  2. Remove direct internet exposure. Put the device behind a firewall or VPN and allow only the specific management addresses that need it.
  3. Check the vendor's software update page and the VulnCheck advisory for a fixed release or official guidance, and ask the vendor directly what applies to end-of-service units.
  4. If a newer release is offered, back up the configuration, then upgrade during a maintenance window.
  5. Segment the device from business systems so that a compromise cannot easily reach other hosts.
  6. Review access logs and device configuration for unexpected changes, new accounts or unfamiliar outbound connections.
  7. If you cannot patch and cannot restrict access, plan a replacement with a supported product, since the line is end of service.
Caution. Upgrading firmware or software on an operational device can interrupt the equipment it monitors or controls. Back up the configuration and schedule the work for a window when a short outage is acceptable.

End of service changes the risk

The advisory states the product line reached its end of service date in March 2026. Whether the vendor will ship further fixes for it is not stated in the research, so do not plan around a patch that may not come. Where updates are unavailable, network controls are your main defense: restrict who can reach the device, keep it off the public internet, and monitor it closely until it can be retired.

Detection and monitoring afterwards

  • Keep access logs for the device or the proxy in front of it, and alert on requests to the three affected scripts from unfamiliar addresses.
  • Watch for new outbound connections from the device, which a normally quiet appliance should rarely make.
  • Re-check exposure after any network change, such as a new port forward or VPN rule.
  • Monitor the certificate and DNS records for any hostname pointing at the device. Webforta Monitor can alert on certificate and DNS changes, and the free SSL checker shows what certificate a public endpoint presents.

If you suspect a device was already compromised, treat it as untrusted: isolate it from the network, preserve logs, and follow our guide on what to do when a website is hacked, adapting the steps to your environment.

What is still unknown

The research does not confirm active exploitation, does not list a fixed version, and does not name a CISA KEV entry. The GitHub entry is unreviewed and based on the NVD text. Details may change as the vendor or VulnCheck publish more, so recheck the linked pages before acting on any assumption about patched releases.

Sources

This article is based on the public sources below. Details can change as vendors and researchers publish more; check the original advisories before acting on version numbers.

  1. CVE-2026-124002 - GitHub Advisory Databasegithub.com
  2. CVE-2026-124002: XWEB EVO versions 5.2.3 through before 5.7.0 contains a pre-authentication OS command injection vulnerability that allows unauthenticated remote attackers to…nvd.nist.gov ·
  • XWEB EVO
  • CVE-2026-124002
  • command injection
  • OT security
  • end of service