Skip to content

Tiers and Alerts

magikh0e edited this page Nov 17, 2026 · 3 revisions

Tiers and Alerts

Tiers

Every match that isn't ruled out lands in one tier:

Tier What puts it there
Act now Exploited in the wild (CISA KEV, or CISA SSVC "active"), a Metasploit exploit module, or a Nuclei template for a CVSS 7+ bug
Watch Public exploit code (Exploit-DB, a GitHub PoC, your own archive), a Metasploit auxiliary module, a Nuclei template, SSVC "poc", or EPSS at or above tiers.epss_watch_pct (default top 10%)
Info It matches, but nothing suggests anyone is using it

The page shows act now and watch by default, and the "include info" box adds the rest.

A check on a distro or vendor build (a Proxmox kernel, the FreeBSD base, a package no tracker answered) means "upstream says affected, the vendor may have patched it". Those only show up when they reach act now, meaning real exploitation. A PoC or a high EPSS alone isn't enough.

EPSS and KEV describe the internet at large, not your network. A KEV entry for something only reachable on your LAN is still worth patching, just maybe not tonight.

The Changes feed

After the first sync, cvescope records events for the CVEs that touch your inventory:

  • added to CISA KEV;
  • a new Metasploit module, Nuclei template, Exploit-DB entry, GitHub PoC, or archive file;
  • EPSS jumped by tiers.epss_jump or crossed into the watch percentile;
  • CISA SSVC raised its exploitation rating;
  • a new match (a new CVE, or a new item), or a match that went away after an upgrade.

The Changes tab and cvescope events show them. Out in the wild shows recent KEV additions and newly weaponized CVEs whether you run the product or not.

Alerts

[notify]
min_tier = "act"          # act | watch | info
keyword_matches = false   # also alert on keyword-only matches

[[notify.ntfy]]
url = "https://ntfy.sh/pick-a-long-random-topic"
# token_env = "NTFY_TOKEN"

[[notify.homeassistant]]
url = "http://homeassistant.local:8123"
token_env = "HASS_TOKEN"
service = "notify.mobile_app_your_phone"

[[notify.webhook]]
url = "https://hooks.example.org/cvescope"   # receives {"title": ..., "message": ...}

Each sync sends one message per channel covering the changes at or above min_tier, with the CVE, the item, the tier and what changed. Every pending change is marked as handled after a run, sent or not, so nothing piles up.

  • cvescope notify --test sends one test message to every channel.
  • cvescope sync --dry-run prints what it would send and leaves the changes pending.
  • With no channels configured, alerts print to the console.

Acknowledging what you've handled

Some act-now entries are things you've already dealt with: a fix Microsoft made opt-in that you've turned on, a service you've firewalled off, an upgrade that's booked. Acknowledge them and they leave act now, watch, the counts and the alerts:

cvescope ack CVE-2013-3900 --item minitop -m "applied the opt-in fix"
cvescope ack CVE-2024-23897 -m "CLI disabled, upgrading Saturday" --days 7
cvescope acks              # what's acknowledged, and whether it still holds
cvescope unack CVE-2024-23897

Or open the CVE on the page and use Acknowledge under "Your inventory". Without --item, ack covers every item the CVE shows up on.

An acknowledgement comes back by itself when its end date passes (the end date is optional), or when the evidence gets worse than it was when you acknowledged it: a higher tier, a new kind of exploitation evidence (added to KEV, a first Metasploit module, SSVC raised), or a check that turns out affected. The page and cvescope report then show it again with the reason. More of the same evidence (another GitHub PoC, EPSS moving within its tier) and a new version of the item, such as a monthly Windows build, don't bring it back. Each acknowledgement and removal is kept in the CVE's history. "Show acknowledged" on the page lists the hidden ones.

Clone this wiki locally