Skip to article content

Received a vulnerability report? How to understand what needs attention

Received a report about possible security flaws? Learn how to distinguish confirmed findings from warnings that need checking, and understand what to address first.

Understand the scope before counting alerts

A PDF arrives with coloured warnings and technical names. You want to know whether your website is at risk and what to do next. An alert alone does not prove that someone has gained access to your website. To decide what to do, establish what the report detected, which part of the website it affects and its possible consequences.

Check which domains, pages or components were reviewed, when and with what access. An external assessment observes part of the system; it should not imply that every internal function has been examined.

Distinguish a tool reporting a possible exposure from a check confirming a problem under specific conditions. The warning deserves attention, but its colour alone does not explain your website’s situation.

Six questions to help you understand each security warning

From an alert to a specific check
QuestionWhat it should establish
What is claimed?The exact observation, without extending its meaning.
Which data or check supports the warning?The response, configuration or test supporting the alert.
What is affected?Component, version, page and known conditions.
What consequence is demonstrated?Confirmed impact distinguished from possible impact.
What remains open?Access or checks still needed.
Who takes the next step?An owner and a method for verifying the outcome.

Use the worksheet with your maintenance team. If the report already answers these questions, it provides a reading guide. Missing information identifies where to continue without repeating the entire assessment.

An example: an alert based on a detected version

Fictional example, without a real component or vulnerability: a tool identifies a possible library version and associates it with a security advisory.

First establish whether the identification is correct. Then check whether the advisory covers the installed version and whether its conditions apply to this system. A possible match does not establish that access has been compromised.

If identification is unsupported, correct the conclusion and document why. An alert identifying a problem that is not present in the assessed system is a false positive. An alert still awaiting verification is not enough to justify that label. If an affected component is confirmed, the team can assess exposure and suitable measures. In both cases, retain the original advisory and the checks performed.

On WordPress, that may require establishing the component’s purpose before replacing it. Our plugin removal worksheet helps collect that context.

Technical severity needs operational context

A CVE identifier identifies a published vulnerability and its references. Check the vendor’s advisory and affected versions; an identifier in the report does not by itself establish that your installation is vulnerable.

The CVSS Base score describes the technical severity of a security flaw. To decide its priority on your website, the team also needs to check whether it affects you and under which conditions it could be exploited. See FIRST’s official CVSS user guide.

Decisions also need context: the affected system, how it is accessed, the information it handles and existing safeguards. Keep confirmed findings separate from matters awaiting investigation.

An unconfirmed serious alert still needs an owner and a suitable investigation priority. A useful ordered list explains that reasoning rather than simply sorting colours.

Agree the correction and verify the change

A useful recommendation explains the proposed action, dependencies and outcome check. Updating a component or changing a configuration may require testing to preserve the website’s functions.

Do not reproduce intrusive tests merely because a report describes them. Active verification needs authorisation and agreed scope. If there are signs of an ongoing incident, involve the security lead and the incident procedure rather than treating it as routine maintenance.

After a correction, retain the follow-up evidence and remaining limitations. Resolving one finding does not establish the security of the entire system.

A report should support a decision

You should finish knowing what is established, its possible consequences and the next step. That information also helps when a client requests security evidence.

Nodo91 can assess your website’s security within agreed scope and help interpret the findings. If other needs compete for attention, this prioritisation worksheet brings them together while keeping uncertainty visible.