Glossary
Agent : Software that runs on a host or a Kubernetes node and sends its software inventory to the server. Each agent has a key.
Asset : A host, or a container image, with an inventory. An image that runs in several containers is one asset. Shown on the Assets page.
Cluster : A Kubernetes cluster, or the name a host agent is grouped under. Its name is set when its registration token is created.
Cluster token
: The vopk_ token that registers the agents of one cluster. Shown once, and revocable.
Container : One container of a pod, reported by the Kubernetes agent with its cluster, namespace, pod, container and image.
CPE : Common Platform Enumeration. The vendor and product names that vulnerability data uses. See CPE mappings.
CVE, GHSA : Identifiers of vulnerabilities. CVE numbers come from the CVE program; GHSA identifiers come from GitHub's advisory database.
CVSS : A score from 0 to 10 that rates how severe a vulnerability is.
Dependency : A package found in an inventory, identified by its name, version, type and Package URL.
EPSS
: The Exploit Prediction Scoring System of FIRST. A score from 0 to 1 for the chance that a CVE is exploited in the next 30 days. Shown on each finding, with the daily scores of the exploit-intel-update task.
Finding : One vulnerability in one installed package on one asset.
Fixable : A finding for which a fixed version of the package is known.
Grype : The vulnerability scanner the server uses. The server keeps a local copy of its database.
Inventory : The complete list of software an agent reports for an asset, as an SBOM.
KEV : The CISA Known Exploited Vulnerabilities catalog. A finding in it is known to be exploited in the wild, and is marked Exploited (KEV).
MCP : Model Context Protocol. The protocol of the server's endpoint for AI assistants. See MCP endpoint.
Node : The host name of the agent that reported a row. The node agent and the host agent of one Kubernetes node are one node.
OpenVEX : A JSON format that records the status of a vulnerability in a product, such as not affected. Export it from the Findings page as OpenVEX.
Placement : Where an asset was seen: a host's node, or one container of a pod.
Package URL (purl)
: A standard identifier of a package, such as pkg:npm/lodash@4.17.15.
Permission
: A right such as inventory.view. Roles are sets of permissions. See Permissions.
Role : A named set of permissions. Users hold one or more roles.
SARIF : Static Analysis Results Interchange Format. A JSON format that CI security tabs, such as the code scanning of GitHub, read. Export the findings as SARIF.
SBOM : Software Bill of Materials. The document an agent sends to the server, listing the packages of an asset.
Scan : The matching of the packages of pending assets against the vulnerability database.
Scope : The clusters and nodes a user may see. Applies to every page and every API call except the rules and settings.
Scan pending : The state of an asset whose new inventory has not been scanned yet.
SSO : Single sign-on. Signing in through an identity provider, with OpenID Connect or SAML.
Task : A recurring job of the server, such as the database update, with a cron schedule.
Triage decision : What a person decided about one finding: not affected, affected, fixed or under investigation, with an owner, a ticket link and a note. A not affected finding is hidden from the lists, exports and alerts.
Personal token
: The vopt_ token a user creates for scripts and AI assistants. It has the roles and scope of its user.