Data source
The pipeline reads the local cvelistV5 mirror, parses datePublished fields with schema fallbacks, and normalizes CNA metadata into a shared summary file.
Methodology
How CVETime measures publication velocity
The dashboard measures the time between distinct CVE publication events, then rolls those intervals into global and per-CNA views so the signals stay comparable across the ecosystem.
Interval logic
Same-second publication events are collapsed to a logical single event before computing rolling windows, which keeps bursty releases from being counted twice.
Rolling windows
The homepage defaults to a 24-month window, while 2017+ and all-history views remain available for broader context. CNA pages use the same range model so charts stay consistent.
Interpretation
Lower intervals mean faster publication cadence. The trend line shows the monthly mean interval, while the shaded band shows the P25 to P75 spread around that monthly behavior.
Notes and limitations
- CNAs with too few events are excluded from the leaderboard and per-CNA shards.
- Metrics are optimized for broad comparison, not for forensic timing analysis of individual CVEs.
- All charts are regenerated from the same summary artifact so the dashboard stays internally consistent.