Vulnerability severity levels

Tier: Ultimate Offering: GitLab.com, Self-managed, GitLab Dedicated

GitLab vulnerability analyzers attempt to return vulnerability severity level values whenever possible. The following is a list of available GitLab vulnerability severity levels, ranked from most to least severe:

  • Critical
  • High
  • Medium
  • Low
  • Info
  • Unknown

GitLab analyzers make an effort to fit the severity descriptions below, but they may not always be correct. Analyzers and scanners provided by third-party vendors may not follow the same classification.

Critical severity

Vulnerabilities identified at the Critical Severity level should be investigated immediately. Vulnerabilities at this level assume exploitation of the flaw could lead to full system or data compromise. Examples of critical severity flaws are Command/Code Injection and SQL Injection. Typically these flaws are rated with CVSS 3.1 between 9.0-10.0.

High severity

High severity vulnerabilities can be characterized as flaws that may lead to an attacker accessing application resources or unintended exposure of data. Examples of high severity flaws are External XML Entity Injection (XXE), Server Side Request Forgery (SSRF), Local File Include/Path Traversal and certain forms of Cross-Site Scripting (XSS). Typically these flaws are rated with CVSS 3.1 between 7.0-8.9.

Medium severity

Medium severity vulnerabilities usually arise from misconfiguration of systems or lack of security controls. Exploitation of these vulnerabilities may lead to accessing a restricted amount of data or could be used in conjunction with other flaws to gain unintended access to systems or resources. Examples of medium severity flaws are reflected XSS, incorrect HTTP session handling, and missing security controls. Typically these flaws are rated with CVSS 3.1 between 4.0-6.9.

Low severity

Low severity vulnerabilities contain flaws that may not be directly exploitable but introduce unnecessary weakness to an application or system. These flaws are usually due to missing security controls, or unnecessary disclose information about the application environment. Examples of low severity vulnerabilities are missing cookie security directives, verbose error or exception messages. Typically these flaws are rated with CVSS 3.1 between 0.1-3.9.

Info severity

Info level severity vulnerabilities contain information that may have value, but are not necessarily associated to a particular flaw or weakness. Typically these issues do not have a CVSS rating.

Unknown severity

Issues identified at this level do not have enough context to clearly demonstrate severity.

GitLab vulnerability analyzers include popular open source scanning tools. Each open source scanning tool provides their own native vulnerability severity level value. These values can be one of the following:

Native vulnerability severity level type Examples
String WARNING, ERROR, Critical, Negligible
Integer 1, 2, 5
CVSS v2.0 Rating (AV:N/AC:L/Au:S/C:P/I:P/A:N)
CVSS v3.1 Qualitative Severity Rating CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H

To provide consistent vulnerability severity level values, the GitLab vulnerability analyzers convert from the above values to a standardized GitLab vulnerability severity level, as outlined in the following tables:

Container Scanning

GitLab analyzer Outputs severity levels? Native severity level type Native severity level example
container-scanning Yes String Unknown, Low, Medium, High, Critical

When available, the vendor severity level takes precedence and is used by the analyzer. If that is not available then it falls back on the CVSS v3.1 rating. If that is also not available, then the CVSS v2.0 rating is used instead. Details on this implementation are available on the issue for trivy.

DAST

GitLab analyzer Outputs severity levels? Native severity level type Native severity level example
Browser-based DAST Yes String HIGH, MEDIUM, LOW, INFO

API security testing

GitLab analyzer Outputs severity levels? Native severity level type Native severity level example
API security testing Yes String HIGH, MEDIUM, LOW

Dependency Scanning

GitLab analyzer Outputs severity levels? Native severity level type Native severity level example
gemnasium Yes CVSS v2.0 Rating and CVSS v3.1 Qualitative Severity Rating 1 (AV:N/AC:L/Au:S/C:P/I:P/A:N), CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H

The CVSS v3.1 rating is used to calculate the severity level. If it’s not available, the CVSS v2.0 rating is used instead.

Fuzz Testing

All fuzz testing results are reported as Unknown. They should be reviewed and triaged manually to find exploitable faults to prioritize for fixing.

SAST

GitLab analyzer Outputs severity levels? Native severity level type Native severity level example
sobelow Yes Not applicable Hardcodes all severity levels to Unknown
SpotBugs Yes Integer 1, 2, 3, 11, 12, 18
pmd-apex Yes Integer 1, 2, 3, 4, 5
kubesec Yes String CriticalSeverity, InfoSeverity
semgrep Yes String error, warning, note, none

IaC Scanning

GitLab analyzer Outputs severity levels? Native severity level type Native severity level example
kics Yes String error, warning, note, none (gets mapped to info in analyzer version 3.7.0 and later)

Secret Detection

The GitLab secrets analyzer hardcodes all severity levels to Critical. Epic 10320 proposes to adopt more granular severity ratings.