During the pre-engagement discussions and in the SOW, it is important to include two disclaimers that outline two important points about the penetration test.
First, you should have a disclaimer that states that the penetration test is a point-in-time assessment — meaning you have tested against known vulnerabilities and exploits as of the current date. As time goes on and new software and systems are installed on the network, your assessment would not have tested those new items.
Second, you should have a disclaimer that indicates that the comprehensiveness of the penetration test is based on the types of tests authorized by the customer and the known vulnerabilities at the time. For example, if the customer requests that no denial of service (DoS) attacks are performed (which is common), your penetration test would not have tested how the company stands up against a DoS attack. This disclaimer will help protect you if the customer is hit with a DoS attack after the penetration test is performed.
Scoping the Project
During the pre-engagement activities, it is important to have an initial meeting with the customer that allows you to discuss the scope of the project and get an understanding of what the customer’s goals are for the penetration test.
When preparing for the initial meeting with the customer, you should plan out scoping questions that will help you understand the magnitude of the project. Some common questions to ask when determining the scope of the pentest are:
What is the goal of the penetration test? (Why is it being done?)
Is the penetration test going to test internal systems, external systems, or both?
What are the Internet Protocol (IP) ranges of the internal and external systems that are being tested?
What are the internal and external domain names of the systems to be tested?
Does the company own the systems using those IP addresses?
Are there any systems hosted by third-party companies such as an ISP or a cloud provider?
What applications and services will be tested?
What types of tests are to be performed? For example, are you testing physical security and/or social engineering, and are DoS attacks allowed?
If performing a black box test, which is discussed in Chapter 1, the penetration tester is typically responsible for discovering target services, and some would say the target IP addresses. The important point here to remember is that you want the customer to give you the target IP addresses and domain names so that you can be sure you have proper authorization to perform testing on those systems. If it is up to the pentester to discover the IP addresses, especially external IP addresses, the tester runs the risk of performing the penetration test on an unauthorized IP address or system owned by someone else.
Depending on the type of testing being performed, there are a number of other questions you can ask during the scoping of the project. The Penetration Testing Execution Standard (PTES) website found at www.pentest-standard.org has an extensive list of questions you can ask. The following sections list example questions for each different type of test.
General questions
What is the goal of the penetration test? (Why is it being done?)
Is the pentest being performed for compliance reasons?
What hours of the day can the penetration test be performed (business hours/non-business hours)?
What are the internal and external target IP addresses?
Are security controls in place such as firewalls and intrusion detection systems?
If a system is compromised, what actions should be taken next (for example, no action, elevate privileges, and so on)?
Web application testing questions
How many web applications/sites are being tested?
How many of those require authentication?
How many static pages are in those sites?
How many dynamic pages are in those sites?
Is the source code available for review?
Is authentication testing to be performed?
Wireless network testing questions
How many wireless networks are there?
What wireless encryption protocol(s) are being used?
What is the area covered by wireless?
Should detection of rogue devices be performed?
Should wireless attacks against clients be performed (or just focus on the access point)?
How many wireless clients are there?
Physical security testing questions
Is physical security testing part of the pentest?
How many locations are there?
Are the locations shared with other businesses? If so, what floors do you occupy?
Are lock picks and bump keys allowed to bypass a locked door?
Are video cameras being used? If so, does the customer own those devices?
