On this page (7 sections)
A critical privilege escalation flaw, tracked as CVE-2026-125294, affects the Nexter Blocks Pro plugin for WordPress in all versions up to and including 5.0.10. According to the NVD record, an unauthenticated attacker can register a new account that is given the full Administrator role.
If your site runs Nexter Blocks Pro and lets visitors register, treat this as urgent. The description says the attack path is reachable on any default install where user registration is enabled.
What happened
The NVD entry, published on 2026-11-21, describes a flaw in the plugin's tpgb_ajax_register_user() function. A visitor with no account can submit a crafted registration request and end up with an Administrator account on the site. The record also references a Wordfence threat intelligence entry and the vendor site at nexterwp.com.
Only the NVD description and these references are available so far, and only one independent publisher covers the story. The research does not name a fixed version, report in-the-wild exploitation, or include a vendor advisory. Check the vendor's channels for the current patch status before assuming any particular version is safe.
How the weakness works, in plain terms
The registration handler accepts extra profile fields. To stop abuse, the plugin checks those fields against a denylist so that nobody can set sensitive WordPress keys such as the one that stores a user's role. According to the CVE description, that check works on raw bytes and only recognizes plain ASCII patterns.
The description says that look-alike Unicode characters slip past this check. A key written with accented lookalike letters does not match the blocked pattern at the byte level. However, the database's default accent-insensitive collation (utf8mb4_unicode_ci or utf8mb4_0900_ai_ci) treats that decorated key as equal to the real role key. When the plugin then calls update_user_meta(), the database matches the real role record and overwrites the new user's role with administrator.
This is a common class of bug: one layer validates text one way, and another layer interprets the same text differently. Denylists are fragile for exactly this reason, because they must anticipate every variant.
Why it is easy to reach
The description states that the nonce the handler requires is publicly emitted on any page containing the plugin's login/register block (tp-login-register). It describes the chain as nonce retrieval, one crafted registration request, and the plugin's own autologin flow. In other words, the nonce offers no real protection here. No exploit details are given in this article.
Who is affected
- Sites running Nexter Blocks Pro 5.0.10 or earlier.
- Sites where user registration is enabled, which the description names as the default condition for easy exploitation.
- Sites that use the plugin's login/register block on a public page, since that is where the nonce is exposed.
- Agencies and hosts managing many WordPress sites, where the premium plugin may have been installed by a client or a previous developer.
The research does not say whether sites with registration disabled can still be reached through the same handler. Do not assume they are safe without confirmation from the vendor or Wordfence.
How to check your exposure
Start by finding out whether the plugin is installed and which version. This WP-CLI command lists installed plugins with their status and version. Run it from your WordPress root, and look for a Nexter Blocks Pro entry at 5.0.10 or lower.
wp plugin list --fields=name,status,versionNext, check whether public registration is on. This returns 1 if anyone can register, and 0 if not.
wp option get users_can_registerThen review your administrators. This lists every account with the Administrator role. Look for any account you do not recognize, especially one created recently.
wp user list --role=administrator --fields=ID,user_login,user_email,user_registeredWhat to do now
- Confirm whether Nexter Blocks Pro is installed and which version it is, using the commands above.
- Check the vendor site and your plugin dashboard for an update newer than 5.0.10, and apply it once it is available. The research does not confirm a fixed release.
- If no fix is available, turn off open registration in Settings, General (clear the Anyone can register box) to reduce the reachable attack surface.
- If you can, remove the login/register block from public pages until the plugin is patched, since the nonce is exposed there.
- Review all Administrator accounts and delete any you cannot account for. Review recently registered users of any role.
- If you find an unexpected administrator, treat the site as compromised: reset all admin passwords, rotate secrets in wp-config.php, and review installed plugins, themes and recently modified files.
- Take a full backup before making changes, so you can compare against it later.
For a broader hardening routine, see the WordPress security checklist and the guide on whether a WordPress plugin is vulnerable. If you suspect a takeover has already happened, follow Website hacked: what to do.
Detection and monitoring
The most telling sign of this attack is an Administrator account that nobody on your team created. Check the user list regularly, and consider alerts for new users and role changes using a logging plugin or your host's tools.
After a successful takeover, attackers with admin rights can install plugins and edit files, as the CVE description notes. Watch for new or modified plugin and theme files, unfamiliar scheduled tasks, and unexpected outbound changes to your pages. Web server logs showing unusual POST requests to your registration endpoint (WordPress admin-ajax.php requests) around the time a suspicious account was created can help date an incident, though the research does not give a specific log signature.
A website firewall can add a layer of filtering in front of WordPress, but it is not a substitute for patching, and nothing here claims a specific rule covers this flaw. You can also check your public pages with the free malware scanner, which only sees what is visible from the outside.
What we do not know yet
- A fixed version number is not stated in the sources reviewed.
- No exploitation in the wild is reported, and the CVE is not in CISA KEV at the time of writing. That can change quickly for a bug this simple to reach.
- The research has no vendor advisory, so the vendor's own wording on impact and remediation is unconfirmed.
- Only one independent publisher covers the story so far.
We will treat the details above as the baseline and recommend rechecking the NVD and Wordfence entries for updates.
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.
