Reading a CVE record: affected versions, the advisory, and whether you are exposed
How to read a CVE entry and its vendor advisory, find affected and fixed versions, check whether your systems run them, and decide how fast to act using KEV and EPSS.
Diagram
From a CVE ID to a decision
Step 1 of 4
Record
The CVE record names the product, the weakness (CWE), and references.
curl -s https://cveawg.mitre.org/api/cve/CVE-2024-3094 | jq '.containers.cna.affected[] | {product, versions}'The example on the open step is one illustration, or the fix you can paste into an application you own.
What this page is made of
3 pieces. Each one says what fails, who is affected, what you would notice, and what you change.
The version range is the finding
- What fails
- A CVE matters to you only if you run an affected version in an affected configuration.
- Who is affected
- Owners of hosts in the range.
- What you would notice
- A ticket that lists a CVE without the installed version.
- What you change
- Record the installed version and the fixed version in every triage note.
Severity is not priority
- What fails
- CVSS measures technical severity, not whether attackers use it or whether you are exposed.
- Who is affected
- Teams that patch by score alone.
- What you would notice
- A 9.8 on an unreachable lab host above a 7.5 on the VPN appliance.
- What you change
- Combine CVSS with CISA KEV, EPSS, and exposure.
Backports change version numbers
- What fails
- Distributions often fix a bug without changing the upstream version string.
- Who is affected
- Scanners and people who compare version strings naively.
- What you would notice
- A scanner flagging a package the distribution already patched.
- What you change
- Check the distribution’s security tracker for the fixed package release.
Parts of a record
A CVE ID is assigned by a CNA, often the vendor. The record contains a description, affected products and versions, CWE classification, references, and often a CVSS vector. The NVD adds enrichment, but vendor advisories are usually more precise about fixed versions and workarounds.
Deciding how fast
CISA’s Known Exploited Vulnerabilities catalog lists CVEs with evidence of exploitation in the wild. EPSS estimates the probability of exploitation in the next 30 days. A KEV-listed bug on an internet-facing system is an emergency; a high CVSS with low EPSS on an internal host can follow the normal cycle.
$ curl -s "https://api.first.org/data/v1/epss?cve=CVE-2023-4966" | jq '.data[0] | {cve, epss, percentile}'
{ "cve": "CVE-2023-4966", "epss": "0.9xxx", "percentile": "0.99x" }Writing it down
Each triage note should say the CVE, the affected range, what you run, whether it is exposed, the decision, and the date. That note is what auditors and the next on-call engineer need.
Check yourself
Recognition and the fix. Nothing here is a procedure.
A scanner flags a CVE on Ubuntu, but the distribution tracker says the package release you run includes a backported fix. What do you do?
Sibling techniques
Pages at the same depth, plus a lab and a glossary card for this one.
This page stands on its own. The lab and the glossary card are the next reading.