
Loading

Loading
We use strictly necessary cookies to run this site, and analytics cookies to understand how it's used. See our Privacy Policy for details.
A GitHub-reviewed security advisory, not necessarily under active exploitation the way our CISA KEV advisories are, but a real, disclosed vulnerability in an open-source dependency.
Affected packages
Vulnerable: >= 4.0.0, < 4.2.6
Patched in 4.2.6
Vulnerable: >= 4.3.0, < 4.3.5
Patched in 4.3.5
Vulnerable: >= 4.4.0, < 4.4.1
Patched in 4.4.1
Details
The email second factor is an mt_rand() six-digit code verified with no attempt limit or lockout
The email second factor is a six-digit code drawn from mt_rand() and verified with no attempt limit, so an attacker who already has the password can brute-force it inside its validity window.
lib/Domain/Service/MfaService.php, in generateEmailVerificationCode():
$numericCode = mt_rand(100000, 999999);
That is a 900,000-value space, and mt_rand() is a Mersenne Twister rather than a CSPRNG, so it is not appropriate for a security token in the first place. PHP provides random_int() for exactly this.
The bigger problem is that nothing limits guesses. lib/Application/Controller/MfaVerifyController.php calls:
$isValid = $this->mfaService->verifyCode($userId, $code);
and on failure the only action taken is:
$this->logger->warning('[MfaVerifyController] Verification failed for user ID: {user_id}', ['user_id' => $userId]);
I grepped that controller for any counter, lockout, throttle or rate limit and found nothing beyond log statements. So an attacker holding valid credentials submits codes until one lands. Against 900,000 values with no ceiling on attempts, the only thing standing in the way is the code's validity window and however fast requests can be issued.
That defeats the purpose of the second factor: it is meant to hold even when the password is known, and here it reduces to a race the attacker is likely to win with sustained requests.
Two separate fixes, and I would treat the second as the urgent one:
random_int(100000, 999999) so it is not predictable from RNG state.Scope: source-verified at master. I confirmed the generator, the verify call site, the log-only failure handling and the absence of any throttling construct in that controller. I did not stand up a Poweradmin instance and run a brute-force against a live code, so the rate at which the space can be covered in practice is not something I measured. If throttling is applied at a layer I did not look at, such as a web-server or reverse-proxy rule shipped with your deployment docs, then the second half of this weakens considerably and I would want to know that.
I used AI assistance while reading the code and preparing this report.
Confirmed, and thanks for the detail. I've widened the affected range: this reaches back to 4.0.0, where MFA was first added, and 4.4.0 still ships it, so the original <= 4.3.4 was too narrow. The 3.x line has no MFA at all and is unaffected. Fixes are in for 4.2.x, 4.3.x, 4.4.x and the development branch.
To answer your question about other layers, there is nothing. No middleware runs in front of that controller, so what you found missing in the controller is the whole picture.
I've adjusted the score down from 9.1 to 7.4. Walking 900,000 values inside the code's ten minute life needs sustained request rates, and MFA is off by default, so there are real preconditions beyond knowing the password. It is still a second factor that does not hold, which is the point of the report.
The mt_rand swap is in as well, though I would treat that as hardening rather than the bug. Recovering the generator state from these codes is not practical in this setup. The missing attempt limit was the actual problem.
EPSS, exploit probability
Low0.62%
estimated chance of real-world exploitation in the next 30 days, higher than 50.5% of every CVE FIRST.org scores
Refreshed 11/20/2026, via FIRST.org's EPSS model, not CVSS, this measures likelihood of exploitation, not how severe it would be.
More composer advisories
Talk to us about this
Find out whether this dependency (or others like it) is actually reachable in your codebase.
Get help patching, or building dependency-scanning into your pipeline going forward.
A secure code review or supply chain review finds this before an advisory does.