top of page

Agentic AI Penetration Testing: A critical PHP flaw was disclosed. Within an hour, AigentX verified the exploit and the fix

Writer: Boaz Shunami
Boaz Shunami
4 hours ago
2 min read

A critical PHP flaw was disclosed. Within an hour, AigentX verified the exploit and the fix.

AigentX helps security teams assess newly disclosed exploits, determine whether their applications are exposed, and verify that remediation blocks the tested attack.


When a critical CVE lands, the security team has three questions. Can this exploit work against our application? What should we patch first? How will we know the patch worked?


On November 19, 2026, PHP disclosed a critical request-smuggling vulnerability in PHP-FPM, CVE-2026-129289, with a CVSS score of 9.2.


In affected deployments, the flaw can let an unauthenticated attacker execute code with the privileges of the PHP-FPM worker. Whether an application is exposed depends on how the web server connects to PHP-FPM and how the application handles incoming requests. PHP security advisory


Within an hour of disclosure, AigentX confirmed command execution against PHP 8.5.12 in our controlled lab. We repeated the same test against PHP 8.5.13 under the same lab conditions. The test no longer achieved command execution. That gave us evidence that the update blocked the specific attack we had demonstrated.



Lab results. PHP 8.5.12 allowed command execution as the PHP-FPM worker. On PHP 8.5.13, the verification request returned HTTP 404.

How AigentX validates newly disclosed exploits


We built this capability for the hours after disclosure, when teams have to decide which applications need immediate attention. AigentX reviews the advisory and the vendor’s fix, validates the reported attack in a controlled environment, and identifies the conditions that make it possible.


For your applications, we can review those conditions against your deployment and validate exposure in an authorized test environment. If the attack succeeds, your team gets evidence of what was demonstrated and where to focus remediation. After you patch or change the configuration, we can repeat the test to verify that the change blocks the demonstrated attack. Each result applies to the application, configuration, and attack tested.


That is what makes the capability part of continuous penetration testing. A new disclosure can trigger a focused assessment between scheduled tests, and the resulting evidence gives your team a concrete way to prioritize the fix and verify it.


What PHP teams should do now


If you run PHP-FPM, check the versions in use and upgrade to 8.2.36, 8.3.37, 8.4.28, or 8.5.13, as appropriate, or use a vendor-maintained build that includes the fix. Review whether your front end reuses FastCGI connections when assessing which deployments may be exposed.


Want to know whether this CVE applies to your applications, or see how AigentX handles the next disclosure? Book an AigentX demo.


Comments


bottom of page