The vulnerability landscape
is not slowing down.
Every day, dozens of new Common Vulnerabilities and Exposures are catalogued, scored, and tracked. Some will quietly expire. Others will be weaponized within hours. This is a look at what the data actually tells us.
This analysis is just a snapshot. View the live dashboard →
You might also want to see my
repo →
Volume has become the new threat vector.
The NVD has been cataloguing vulnerabilities since 1999. For the first decade, volume grew steadily but manageably. Then came the software explosion — cloud, mobile, open source — and with it, an exponential surge in reported CVEs.
Severity: most are serious. A few are catastrophic.
CVSS scores divide CVEs into four buckets. Critical and High together account for roughly half of all reported vulnerabilities — and represent the realistic attack surface for any organization.
But CVSS alone is a static snapshot. It measures theoretical severity — not the probability that a specific CVE will actually be exploited in the wild.
EPSS asks a different question: will this actually be exploited?
The Exploit Prediction Scoring System (EPSS) produces a 0–1 probability score updated daily, drawing on threat intelligence, CVE metadata, and observed exploitation patterns. The gap between CVSS and EPSS reveals which "critical" bugs are real emergencies — and which can wait.
CISA KEV: the list that matters most.
The Known Exploited Vulnerabilities catalog is CISA's curated list of CVEs confirmed to be actively exploited. Every entry has a mandatory remediation deadline for federal agencies — and represents a real threat any organization should take seriously.
Vendors most represented in the KEV.
When a vendor appears repeatedly in the KEV, it signals structural exposure — either code quality issues, widespread deployment, or attacker interest in that ecosystem. The top five tell a familiar story.
OpenClaw's dominance reflects its recent hype. Google's share is dictated by the number of products and the range of clients they have rather than being a warning signal.
What the data demands of us.
A CVE dashboard is not a status display — it's a decision engine. Three principles emerge from the data every time.
- ▶EPSS over CVSS for triage. A CVSS 9.8 with EPSS 0.02 can wait. A CVSS 7.5 with EPSS 0.91 cannot.
- ▶KEV is your SLA. Any CVE on the Known Exploited Vulnerabilities list should be treated as a production incident — not a scheduled ticket.
- ▶Vendor concentration creates exposure. If your stack leans heavily on one vendor with recurring KEV entries, that's an architectural risk, not just an ops problem.
- ▶Speed beats completeness. Patching 70% of criticals within 48 hours is more effective than patching 100% in three weeks.