You got a security test report full of CVE, CVSS and red labels. What the acronyms mean, what to fix first and what to ask the auditor.
How the report is built
Reports differ in form, but they usually have the same parts.
- Executive summary: a few paragraphs on where things stand and what to fix first. If you read only one page, read this one.
- Scope and method: what was tested, when, from which account, and what was not checked. This matters, because a clean report on a narrow scope does not mean the whole company is safe.
- Findings: each with a title, a severity rating, a description, evidence and a fix recommendation.
- Appendices: technical details for the person who will do the fixing.
CVE, or the number of a known vulnerability
A CVE is an identifier for a single, publicly known vulnerability in specific software, for example CVE-2021-41773 in one version of the Apache server. If the report contains a CVE number, the flaw is publicly described and a fix usually exists: most often it is enough to update to a newer version.
Next to it you may see a CWE number: the name of the type of mistake, for example improper access control. It helps you notice which kind of problem keeps repeating in your application.
CVSS, or a number from 0 to 10
CVSS rates how dangerous a flaw is in itself. The ranges are fixed:
- 0.1 to 3.9: low,
- 4.0 to 6.9: medium,
- 7.0 to 8.9: high,
- 9.0 to 10.0: critical.
What the number does not tell you
Next to the score you will often see a string like AV:N/AC:L/PR:N. This is the vector: it describes whether the flaw can be exploited over the internet, whether you need to be logged in, and whether the user has to help. It is worth reading, because it says more than the number alone.
CVSS does not know your company. A critical flaw in a system sitting in a closed network with no outside access may be less urgent than a medium one in a form that a thousand customers use every day. A good auditor adds this to the risk description. If they did not, ask.
What order to fix things in
- First, flaws that criminals are actively exploiting. The US agency CISA keeps a public list of such vulnerabilities (KEV). If the CVE number from your report is on it, that flaw is being used in attacks right now.
- Then critical and high findings that are reachable from the internet without logging in.
- In parallel, quick fixes: updates, configuration, headers. They often remove several findings at once at a small cost.
- Then medium findings, starting with the places that hold personal data and payments.
- Plan for the low and informational ones, don’t ignore them. Several small problems can add up to a serious one.
What to ask the auditor
A report is the start of a conversation, not the end of it. It is worth asking:
- Which findings can actually be exploited in our environment, and which are theoretical?
- Does every finding have a fix recommendation that can be handed to a developer?
- After the fixes, will you run a retest to confirm the flaws are gone?
- What did you not check, and is it worth doing next time?
Key takeaway
Start with the summary and the scope, fix first what is actively exploited and reachable from the internet, and read the CVSS number together with the vector and the risk description. At the end, ask for a retest.