In short: KEV is the catalog maintained by the US cybersecurity agency CISA. It lists vulnerabilities with evidence that they have already been exploited in the wild. CISA calls it the authoritative source for such flaws and recommends that organizations use it as an input when they decide what to fix first.
Every entry carries a few fields: a name and short description, the CVE number, the related weakness (CWE), the date added and a due date. There is also the required action, a flag for whether the flaw is known from ransomware campaigns, and a flag for whether forensic triage is required. The last one means you check not only whether you are patched, but also whether someone has already been in.
The score and the catalog say different things
The CVSS number is a formula: how bad the flaw would be. A KEV listing is a fact: someone is already using it. The first can be high and still sit at the bottom of the queue. The second is the reason to pull it out of the queue today. So with a pile of patches, you look up your numbers in the catalog first.
An example from 13 November 2026: CISA added CVE-2026-114713 in UnrealIRCd, the IRC chat server. By the description, any user on a network with two or more linked servers can inject commands that run with server authority. MITRE rates it 8.6 out of 10, and NVD has not yet given its own score. The catalog records the date added as 13 November, the due date as 16 November, and forensic triage as required. We write only about this entry, because it is the only one the CISA page confirms. Three days between listing and due date is the usual rhythm: the catalog gives you no time to think it over.
One limit: for ransomware campaigns, the UnrealIRCd entry says "Unknown". That does not mean there are none. It means CISA does not know.
The visual is generated code-art, no third-party images.