In a previous post, I discussed the importance of performing scenario-based penetration testing. Now let's look at some of the endpoint-based tests you should include in a typical assessment of your ability to fend off client-side attacks. This is by no means an exhaustive methodology, but covers some of the salient areas associated with offensive techniques utilized by many of the more advanced threats we’re seeing hitting the enterprise today.

Desktop Policies
We’re not talking something you make your employees read -- rather, local system policies, such as group policies, which are likely pushed down by a domain controller. While most people have the basics covered here, such as password complexities, disabling default users, etc., there are some pretty advanced features that you can tweak with group policy these days. These include configuring Microsoft User Account Control (UAC), which can be instrumental in stopping certain malware components in its tracks, such as those that rely on their ability to make modifications to the OS.

Antivirus And HIPS Software
After walking the show floor at your favorite security trade show junket, you can be forgiven for thinking the answer to defending against spear-phishing and other client-centric attacks is a piece of commercial software. While many of the AV and other products implementing HIPS-type technologies can provide an effective defense, the devil is always in the details. As such, it's vital they be tested in a realistic manner.

In both instances, you don't need to drop tens of thousands on 0-day to effectively put these countermeasures through their paces. A desktop that is optimally configured but installed with recent-but-vulnerable piece of client-side software (such as a PDF reader or browser plugin) can be utilized to perform a realistic assessment.

In many cases, this could very well represent the state of most desktops in your environment, anyway, without having to regress a patch level. Freely available software (such as the Metasploit Framework) is capable of creating an entry vector (such as a malicious PDF attachment) and payload (such as a reverse-connect command-and-control channel), which leverage techniques used by many "real" threat-actors, such that it provides an effective assessment.

At FusionX, we maintain our own entry and command-and-control components for this purpose, and if you have the capability, you can certainly go down that path to increase sophistication levels. However, you will probably be surprised with the results, even with publicly available tools.

In my next post, I'll take a look at some key performance indicators for the above countermeasures, and look at some network-centric components that should also be tested as a part of this type of activity.

Tom Parker is Chief Technology Officer at FusionX.