GHSA-xw67-cg5f-4m2r CVE-2026-45578

AVideo: OS command injection in on_publish.php execAsync via unescaped m3u8 URL

High Packagist

Severity: 8.8 / 10; High · CVSS 3.1
Exploit chance (EPSS): 0.32%; Chance of exploitation in the next 30 days. Higher than 24.73% of scored vulnerabilities.
Known exploited: No; Not in CISA's catalog
Weakness: CWE-78 Improper Neutralization of Special Elements used in an OS Command ('OS Command Injection')
Fix: None yet; No fixed version published

Description

Summary

Type: Classic shell-metacharacter injection. The YPTSocket notification branch in plugin/Live/on_publish.php builds an execAsync() command line by string concatenation, single-quoting each argument but never calling escapeshellarg(). A ' in any of the three interpolated values ($users_id, $m3u8, $obj->liveTransmitionHistory_id) closes the quoted token and lets the attacker append arbitrary commands. File: plugin/Live/on_publish.php, line 267. Root cause: the developer wrapped each variable in literal single quotes ('$users_id', '$m3u8', '$obj->liveTransmitionHistory_id') believing this provides shell-quoting. PHP single-quoted-into-shell is not safe quoting; it is just two literal quote characters that the shell pairs greedily. Any embedded ' closes the outer string and resumes interpretation in the shell. The rest of the AVideo codebase already calls escapeshellarg() (137 call sites across the project) for ffmpeg invocations, so the safe primitive is well-known to the project; it was simply omitted from this branch. The endpoint is web-reachable (no .htaccess rule restricts on_publish.php, no REMOTE_ADDR check), so the trigger is a direct HTTP POST without going through nginx-rtmp.

Affected Code

File: plugin/Live/on_publish.php, lines 256-271.

if (AVideoPlugin::isEnabledByName('YPTSocket')) {
    $array = setLiveKey($lth->getKey(), $lth->getLive_servers_id());
    @ob_clean();
    _ob_start();
    $lth = new LiveTransmitionHistory($obj->liveTransmitionHistory_id);
    $m3u8 = Live::getM3U8File($lth->getKey(), false, true);          // value-carrying URL: contains the stream key verbatim
    $users_id = $obj->row['users_id'];
    $liveTransmitionHistory_id = $obj->liveTransmitionHistory_id;
    if (strtoupper(substr(PHP_OS, 0, 3)) === 'WIN') {
        include "{$global['systemRootPath']}plugin/Live/on_publish_socket_notification.php";
    } else {
        $command = get_php(). " {$global['systemRootPath']}plugin/Live/on_publish_socket_notification.php '$users_id' '$m3u8' '{$obj->liveTransmitionHistory_id}'";  // <-- BUG: literal quotes, no escapeshellarg
        $pid = execAsync($command);                                  // sink: shell exec
    }
}

Live::getM3U8File($key, false, true) (Live.php:1337-1350 -> Live.php:4845-4889) returns "{$playerServer}{$uuid}.m3u8" (or "{$playerServer}{$uuid}/index.m3u8") where $uuid = $this->getKeyWithIndex(...) is the stream key string read straight out of the live_transmitions table. There is no character normalisation between database read and command construction.

Why it's wrong: '$m3u8' is not shell quoting. PHP interpolates $m3u8 into the string between two literal ' characters. The shell then tokenises the result. If $m3u8 contains ' itself, the shell sees '…' followed by <attacker bytes> followed by another ', which forms two adjacent quoted strings concatenated with whatever the attacker put between them. Embedded ;, backticks, $(), &&, |, or \n then run as shell commands. The fix is escapeshellarg(), which AVideo already uses 137 times in ffmpeg invocations (e.g. getVideos.php:1069, videos.json.php, aVideoEncoder.json.php); this branch simply forgot it.

Exploit Chain

Attacker authenticates and arranges for one of the command variables to contain '. Under the current code the readily available primitive is a canStream user supplying a stream key via the persistence path (saveLive.php's $_REQUEST['key'] is written verbatim to live_transmitions.key). State: a row exists with key = "evilkey';id>/tmp/pwn;#".
Attacker POSTs directly to https://target/plugin/Live/on_publish.php (the file is web-served, no IP restriction) with body:
```
name=evilkey';id>/tmp/pwn;#
p=<md5(attacker_password)>
tcurl=rtmp://target/live
addr=1.2.3.4
```
on_publish.php:117 runs preg_replace("/[&=]/", '', $_POST['name']) — only &/= are stripped, so ';id>/tmp/pwn;# survives. Lines 143-163 confirm $_GET['p'] === $user->getPassword() (the attacker is themself, knows their own MD5), persist a LiveTransmitionHistory row with the poisoned key, and set $obj->error = false. State: authorisation gate passed.
Line 261 calls Live::getM3U8File($lth->getKey(), false, true), returning "https://server/live/evilkey';id>/tmp/pwn;#.m3u8". State: $m3u8 carries the injection payload.
Line 267 builds the command string by concatenation:
```
php /var/www/AVideo/plugin/Live/on_publish_socket_notification.php '7' 'https://server/live/evilkey';id>/tmp/pwn;#.m3u8' '42'
```
Shell tokenisation sees: php, …/on_publish_socket_notification.php, '7', 'https://server/live/evilkey' (the attacker's ' closed the second quote), then operator ;, then command-2 id>/tmp/pwn, then ;, then #.m3u8' '42' (everything after # is a comment). State: the shell has parsed two real commands.
Line 269 execAsync($command) spawns the shell, which runs the secondary command id>/tmp/pwn as the AVideo PHP-FPM/Apache user. State: arbitrary OS command execution with the privileges of the web-server runtime user.
Final state: the attacker reads /tmp/pwn, swaps the payload for a reverse shell, exfiltrates videos/configuration.php (database password and root URL), drops a webshell into the upload tree, or pivots to other plugin credentials (PayPal/Stripe API keys, AWS keys for the CDN plugin, OpenAI key for the AI plugin).

Security Impact

Severity: sec-high. Pre-auth-friendly remote code execution: the only prerequisite is that the attacker can place a ' into one of the three command-line variables, which on a streaming platform means a single low-privilege account. Attacker capability: with one canStream account and two HTTP requests, the attacker executes arbitrary shell commands as the AVideo runtime user. From there: read database credentials, exfiltrate user data, write a webshell into a publicly-served path, pivot to plugin credentials, persist via cron, or escalate via any local sudoers entries. Preconditions: AVideo deployment with Live and YPTSocket plugins enabled (the standard live-streaming bundle); attacker can reach /plugin/Live/on_publish.php over the network; a value containing ' is reachable into users_id, m3u8, or liveTransmitionHistory_id (the current code lets canStream users supply such a value via the stream-key persistence path). Differential: source-inspection-verified end-to-end. The shell-tokenising behaviour of '…'…'…' is reproducible offline:

$ s="php /a/b.php '7' 'https://s/live/evilkey';id>/tmp/pwn;#.m3u8' '42'"
$ rm -f /tmp/pwn; bash -c "$s" 2>/dev/null; ls -l /tmp/pwn
-rw-r--r-- 1 user user N <date> /tmp/pwn        # injected `id` ran, output captured

The patched build (with the suggested escapeshellarg() fix below applied) constructs php /a/b.php '7' 'https://s/live/evilkey'\''id>/tmp/pwn;#.m3u8' '42', which the shell parses as a single argument containing the literal characters; the second command never runs.

Suggested Fix

Use escapeshellarg() on every variable interpolated into the command string. This matches established project conventions (137 other call sites for ffmpeg invocations).

--- a/plugin/Live/on_publish.php
+++ b/plugin/Live/on_publish.php
@@ -264,7 +264,11 @@
         if (strtoupper(substr(PHP_OS, 0, 3)) === 'WIN') {
             include "{$global['systemRootPath']}plugin/Live/on_publish_socket_notification.php";
         } else {
-            $command = get_php(). " {$global['systemRootPath']}plugin/Live/on_publish_socket_notification.php '$users_id' '$m3u8' '{$obj->liveTransmitionHistory_id}'";
+            $command = get_php()
+                . ' ' . escapeshellarg($global['systemRootPath'] . 'plugin/Live/on_publish_socket_notification.php')
+                . ' ' . escapeshellarg((string) $users_id)
+                . ' ' . escapeshellarg((string) $m3u8)
+                . ' ' . escapeshellarg((string) $obj->liveTransmitionHistory_id);
             _error_log("NGINX Live::on_publish YPTSocket start  ($command)");
             $pid = execAsync($command);
         }

Defence-in-depth: on_publish.php is the nginx-rtmp webhook and should not be reachable from the public Internet. Add an .htaccess/nginx location rule restricting the file to 127.0.0.1 and any configured RTMP server IPs. That blocks the trigger path independently of the sanitisation work.

Affected versions

Packagist

Package	Vulnerable	Fixed 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 8.8

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

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

Weakness

CWE-78: Improper Neutralization of Special Elements used in an OS Command ('OS Command Injection')

The product constructs all or part of an OS command using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify the intended OS command when it is sent to a downstream component.

How to prevent it

Architecture and Design: If at all possible, use library calls rather than external processes to recreate the desired functionality.
Architecture and Design, Operation: Run the code in a "jail" or similar sandbox environment that enforces strict boundaries between the process and the operating system. This may effectively restrict which files can be accessed in a particular directory or which commands can be executed by the software. OS-level examples include the Unix chroot jail, AppArmor, and SELinux. In general, managed code may provide some protection. For example, java.io.FilePermission in the Java SecurityManager allows the software to specify restrictions on file operations. This may not be a feasible solution, and it only limits the impact to the operating system; the rest of the application may still be subject to compromise. Be careful to avoid CWE-243 and other weaknesses related to jails.
Architecture and Design: For any data that will be used to generate a command to be executed, keep as much of that data out of external control as possible. For example, in web applications, this may require storing the data locally in the session's state instead of sending it out to the client in a hidden form field.
Architecture and Design: For any security checks that are performed on the client side, ensure that these checks are duplicated on the server side, in order to avoid CWE-602. Attackers can bypass the client-side checks by modifying values after the checks have been performed, or by changing the client to remove the client-side checks entirely. Then, these modified values would be submitted to the server.
Architecture and Design: Use a vetted library or framework that does not allow this weakness to occur or provides constructs that make this weakness easier to avoid. For example, consider using the ESAPI Encoding control [REF-45] or a similar tool, library, or framework. These will help the programmer encode outputs in a manner less prone to error.
Implementation: While it is risky to use dynamically-generated query strings, code, or commands that mix control and data together, sometimes it may be unavoidable. Properly quote arguments and escape any special characters within those arguments. The most conservative approach is to escape or filter all characters that do not pass an extremely strict allowlist (such as everything that is not alphanumeric or white space). If some special characters are still needed, such as white space, wrap each argument in quotes after the escaping/filtering step. Be careful of argument injection (CWE-88).
Implementation: If the program to be executed allows arguments to be specified within an input file or from standard input, then consider using that mode to pass arguments instead of the command line.
Architecture and Design: If available, use structured mechanisms that automatically enforce the separation between data and code. These mechanisms may be able to provide the relevant quoting, encoding, and validation automatically, instead of relying on the developer to provide this capability at every point where output is generated. Some languages offer multiple functions that can be used to invoke commands. Where possible, identify any function that invokes a command shell using a single string, and replace it with a function that requires individual arguments. These functions typically perform appropriate quoting and filtering of arguments. For example, in C, the system() function accepts a string that contains the entire command to be executed, whereas execl(), execve(), and others require an array of strings, one for each argument. In Windows, CreateProcess() only accepts one command at a time. In Perl, if system() is provided with an array of arguments, then it will quote each of the arguments.
Implementation: Assume all input is malicious. Use an "accept known good" input validation strategy, i.e., use a list of acceptable inputs that strictly conform to specifications. Reject any input that does not strictly conform to specifications, or transform it into something that does. When performing input validation, consider all potentially relevant properties, including length, type of input, the full range of acceptable values, missing or extra inputs, syntax, consistency across related fields, and conformance to business rules. As an example of business rule logic, "boat" may be syntactically valid because it only contains alphanumeric characters, but it is not valid if the input is only expected to contain colors such as "red" or "blue." Do not rely exclusively on looking for malicious or malformed inputs. This is likely to miss at least one undesirable input, especially if the code's environment changes. This can give attackers enough room to bypass the intended validation. However, denylists can be useful for detecting potential attacks or determining which inputs are so malformed that they should be rejected outright. When constructing OS command strings, use stringent allowlists that limit the character set based on the expected value of the parameter in the request. This will indirectly limit the scope of an attack, but this technique is less important than proper output encoding and escaping. Note that proper output encoding, escaping, and quoting is the most effective solution for preventing OS command injection, although input validation may provide some defense-in-depth. This is because it effectively limits what will appear in output. Input validation will not always prevent OS command injection, especially if you are required to support free-form text fields that could contain arbitrary characters. For example, when invoking a mail program, you might need to allow the subject field to contain otherwise-dangerous inputs like ";" and ">" characters, which would ne
Architecture and Design: When the set of acceptable objects, such as filenames or URLs, is limited or known, create a mapping from a set of fixed input values (such as numeric IDs) to the actual filenames or URLs, and reject all other inputs.
Operation: Run the code in an environment that performs automatic taint propagation and prevents any command execution that uses tainted variables, such as Perl's "-T" switch. This will force the program to perform validation steps that remove the taint, although you must be careful to correctly validate your inputs so that you do not accidentally mark dangerous inputs as untainted (see CWE-183 and CWE-184).
Operation: Run the code in an environment that performs automatic taint propagation and prevents any command execution that uses tainted variables, such as Perl's "-T" switch. This will force the program to perform validation steps that remove the taint, although you must be careful to correctly validate your inputs so that you do not accidentally mark dangerous inputs as untainted (see CWE-183 and CWE-184).
Implementation: Ensure that error messages only contain minimal details that are useful to the intended audience and no one else. The messages need to strike the balance between being too cryptic (which can confuse users) or being too detailed (which may reveal more than intended). The messages should not reveal the methods that were used to determine the error. Attackers can use detailed information to refine or optimize their original attack, thereby increasing their chances of success. If errors must be captured in some detail, record them in log messages, but consider what could occur if the log messages can be viewed by attackers. Highly sensitive information such as passwords should never be saved to log files. Avoid inconsistent messaging that might accidentally tip off an attacker about internal state, such as whether a user account exists or not. In the context of OS Command Injection, error information passed back to the user might reveal whether an OS command is being executed and possibly which command is being used.
Operation: Use runtime policy enforcement to create an allowlist of allowable commands, then prevent use of any command that does not appear in the allowlist. Technologies such as AppArmor are available to do this.
Operation: Use an application firewall that can detect attacks against this weakness. It can be beneficial in cases in which the code cannot be fixed (because it is controlled by a third party), as an emergency prevention measure while more comprehensive software assurance measures are applied, or to provide defense in depth [REF-1481].
Architecture and Design, Operation: Run your code using the lowest privileges that are required to accomplish the necessary tasks [REF-76]. If possible, create isolated accounts with limited privileges that are only used for a single task. That way, a successful attack will not immediately give the attacker access to the rest of the software or its environment. For example, database applications rarely need to run as the database administrator, especially in day-to-day operations.
Operation, Implementation: When using PHP, configure the application so that it does not use register_globals. During implementation, develop the application so that it does not rely on this feature, but be wary of implementing a register_globals emulation that is subject to weaknesses such as CWE-95, CWE-621, and similar issues.

References

Advisories

nvd.nist.gov/vuln/detail/CVE-2026-45578

Packages

github.com/WWBN/AVideo

Timeline

May 15, 2026 Published
Sep 10, 2026 Last updated by the source
Nov 1, 2026 Latest EPSS score

Advisory	Severity	Why 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