SafeCurl

SafeCurl intends to be a drop-in replacement for the curl_exec function in PHP. SafeCurl validates each part of the URL against a white or black list, to help protect against Server-Side Request Forgery attacks.

For more infomation about the project see the blog post 'SafeCurl: SSRF Protection, and a "Capture the Bitcoins"'.

Protections

Each part of the URL is broken down and validated against a white or black list. This includes resolve a domain name to it's IP addresses.

If you chose to enable "FOLLOWLOCATION", then any redirects are caught, and re-validated.

Installation

SafeCurl can be included in any PHP project using Composer. Include the following in your composer.json file under require.

"require": {
    "j0k3r\safecurl": "~3.0"
}

Then update Composer.

composer update

Usage

It's as easy as replacing curl_exec and wrapping it in a try {} catch {} block.

use fin1te\SafeCurl\SafeCurl;
use fin1te\SafeCurl\Exception;

try {
    $url = 'http://www.google.com';

    $curlHandle = curl_init();

    //Your usual cURL options
    curl_setopt($curlHandle, CURLOPT_USERAGENT, 'Mozilla/5.0 (SafeCurl)');

    //Execute using SafeCurl
    $safeCurl = new SafeCurl($curlHandle);
    $response = $safeCurl->execute($url);
} catch (Exception $e) {
    //URL wasn't safe
}

Options

The default options are to not allow access to any private IP addresses, and to only allow HTTP(S) connections.

If you wish to add your own options (such as to blacklist any requests to domains your control), simply get a new SimpleCurl\Options object, add to the white or black lists, and pass it along with the method calls.

Domains are express using regex syntax, whilst IPs, scheme and ports are standard strings (IPs can be specified in CIDR notation).

use fin1te\SafeCurl\SafeCurl;
use fin1te\SafeCurl\Options;

$options = new Options();
$options->addToList('blacklist', 'domain', '(.*)\.fin1te\.net');
$options->addToList('whitelist', 'scheme', 'ftp');

$curlHandle = curl_init();

//This will now throw an InvalidDomainException
$safeCurl = new SafeCurl($curlHandle, $options);
$response = $safeCurl->execute('http://safecurl.fin1te.net');

//Whilst this will be allowed, and return the response
$safeCurl = new SafeCurl($curlHandle, $options);
$response = $safeCurl->execute('ftp://fin1te.net');

Since we can't get access to any already set cURL options (see Caveats section), to enable CURL_FOLLOWREDIRECTS you must call the enableFollowRedirects() method. If you wish to specify a redirect limit, you will need to call setMaxRedirects(). Passing in 0 will allow infinite redirects.

$options = new Options();
$options->enableFollowLocation();

//Abort after 10 redirects
$options->setFollowLocationLimit(10);

URL Checking

The URL checking methods are also public, meaning that you can validate a URL before using it elsewhere in your application, although you'd want to try and catch any redirects.

use fin1te\SafeCurl\Url;
use fin1te\SafeCurl\Exception;

try {
    $url = 'http://www.google.com';

    $validatedUrl = Url::validateUrl($url);
    $fullUrl = $validatedUrl['url'];
} catch (Exception $e) {
    // URL wasn't safe
}

Optional Protections

In addition to the standard checks, two more are available.

The first is to prevent DNS Rebinding attacks. This can be enabled by calling the enablePinDns method on an Options object. There is one major issue with this - the SSL certificate can't be validated. This is due to the real hostname being sent in the Host header, and the URL using the IP address.

$options = new Options();
$options->enablePinDns();

The second disables the use of credentials in a URL, since PHP's parse_url returns values which differ from ones cURL uses. This is a temporary fix.

use fin1te\SafeCurl\SafeCurl;
use fin1te\SafeCurl\Exception;
use fin1te\SafeCurl\Options;

$options = new Options();
$options->disableSendCredentials();

$curlHandle = curl_init();

//This will throw an InvalidURLException
$safeCurl = new SafeCurl($curlHandle, $options);
$response = $safeCurl->execute('http://user:pass@google.com');

Cavets

Since SafeCurl uses gethostbynamel to resolve domain names, which isn't IPv6 compatible, the class will only work with IPv4 at the moment. See Issue #1.

As mentioned above, we can't fetch the value of any cURL options set against the provided cURL handle. Because SafeCurl handles redirects itself, it will turn off CURLOPT_FOLLOWLOCATION and use the value from the Options object. This is also true of CURLOPT_MAXREDIRECTS.