Repository navigation
Data Sources
cvescope downloads public feeds at run time and keeps them in its local database. It doesn't redistribute any of their data; each feed keeps its own terms.
| Feed | What it adds | Size / cadence |
|---|---|---|
| CVE List | Every CVE record, including CISA Vulnrichment (SSVC, CPEs, CVSS) | ~660 MB once, then hourly or end-of-day delta zips |
| CISA KEV | Confirmed exploitation in the wild | ~2 MB, fetched only when it changes |
| FIRST EPSS | Daily probability of exploitation in the next 30 days | ~3 MB a day |
| Metasploit module metadata | Weaponized exploits and scanners | ~11 MB, only when it changes |
Nuclei templates cves.json
|
Mass-scannable checks | ~2 MB |
| Exploit-DB index | Published exploit code | ~10 MB |
| PoC-in-GitHub | GitHub repos claiming a PoC (metadata only) | ~85 MB shallow clone, needs git |
| Your own folders | Any file that mentions a CVE ID | Local; only new files after the first scan |
| Debian security tracker (OSV export) | Fixed package version per CVE per Debian release | ~80 MB, only if you run Debian |
| Ubuntu OVAL | Fixed package version per CVE per Ubuntu release | ~2 MB per release you run |
The NVD isn't a feed here. Its analysis backlog means it often lags the CVE List itself. It does help with one thing: some CVE records carry no usable version data (old MITRE-era records, vendors that list only their own rebuilds), while NVD's analysts add CPE ranges with exact fixed versions.
With [nvd] enabled = true, cvescope asks NVD about those records after each match, but only for
matches that are still check and showing up in act now or watch, then matches again with the
ranges. It's off by default because each lookup tells NIST a CVE ID you care about.
- Without an API key NVD allows 5 requests per 30 seconds, so 40 lookups take about 4 minutes.
A key (
api_key_envorapi_key_file) raises that tenfold. A rejected key falls back to keyless for that run. -
max_per_run(default 40) caps lookups per sync. - Answers are cached. One with no ranges yet is asked again after 7 days, the rest after 60.
Fake proof-of-concept repos that drop malware on whoever runs them are a long-running trick. cvescope lists PoC repos from the index's metadata and flags the usual tells: no stars or forks, no description, a pitch to download a release or make contact, created within two days of disclosure. It never clones or opens them. Read the code first, and run it only in a throwaway VM.
cvescope sync applies whatever changed since the last run: CVE List deltas, a new KEV catalog,
today's EPSS scores, and so on. cvescope status shows when each feed last succeeded and any
error. cvescope sync --only kev,epss refreshes just those feeds, and --rebaseline re-imports
the full CVE List.