Rules

Quick Commitlint supports 14 rules. A rule tuple contains severity 0, 1, or 2, an always or never condition when enabled, and an optional value determined by the rule.

always requires the described predicate; never inverts it. [0] disables any rule. See Configuration for tuple validation and Presets for the complete side-by-side defaults.

Rule summary

Body rules

body-leading-blank

Requires a blank line between the header and body when configured with always.

"body-leading-blank": [1, "always"]

Both presets enable it as a warning. A message without a body passes.

# passes
feat: add parser

Explain the implementation.

# warns
feat: add parser
Explain the implementation.

body-max-line-length

Checks every parsed body line against a nonnegative Unicode code-point limit.

"body-max-line-length": [2, "always", 100]

The Conventional preset sets 100 as an error; Angular disables the rule. With always, every line must be at most the limit. With never, the predicate is inverted, so the rule reports unless at least one line exceeds the supplied limit. URLs receive no special exemption.

Footer rules

footer-leading-blank

Requires a blank line before a recognized final footer paragraph with always.

"footer-leading-blank": [1, "always"]

Both presets enable it as a warning. A message without a recognized footer passes.

# passes
fix: close descriptor

Refs: #42

# warns
fix: close descriptor
Refs: #42

Only a recognized footer trailer activates this rule.

footer-max-line-length

Checks every line of a recognized footer against a Unicode code-point limit.

"footer-max-line-length": [2, "always", 100]

The Conventional preset sets 100 as an error; Angular disables the rule. The same always and inverted never semantics as body-max-line-length apply.

Header rules

header-max-length

Checks the complete first line, including type, scope, breaking marker, separator, and subject.

"header-max-length": [2, "always", 100]

Conventional limits the header to 100 code points; Angular limits it to 72. With never, a header at or below the value is rejected instead.

header-trim

Checks whether the header equals the same header after removing leading and trailing ASCII spaces and tabs.

"header-trim": [2, "always"]

Conventional enables this as an error; Angular disables it. never reverses the check and therefore requires leading or trailing whitespace.

# passes with always
feat: add parser

# fails with always
 feat: add parser

Scope rule

scope-case

Checks the optional text inside type(scope): subject against one supported case.

"scope-case": [2, "always", "lower-case"]

Conventional disables this rule. Angular requires lowercase scope. An absent or empty scope is skipped; this rule does not require a scope.

# passes Angular
feat(core): add parser

# fails Angular
feat(Core): add parser

never forbids the selected case rather than requiring another specific case.

Subject rules

subject-case

Checks the parsed subject against one case or any case in a nonempty array.

"subject-case": [
  2,
  "never",
  ["sentence-case", "start-case", "pascal-case", "upper-case"]
]

Both presets use the tuple above, forbidding a subject that matches any listed case. Lowercase natural-language subjects normally pass.

# passes
feat: add native parser

# fails
feat: Add native parser
feat: Add Native Parser
feat: AddNativeParser
feat: ADD NATIVE PARSER

An empty subject is skipped here and handled by subject-empty.

subject-empty

Tests whether the parsed subject is empty.

"subject-empty": [2, "never"]

Both presets use never, requiring a nonempty subject.

# passes
fix: close descriptor

# fails
fix:

subject-exclamation-mark

Tests whether the parsed header has ! immediately before the colon.

"subject-exclamation-mark": [2, "never"]

Angular uses never and rejects feat!: and feat(scope)!:. Conventional disables the rule and allows both. With always, the rule requires the marker.

subject-full-stop

Tests whether the subject ends with the configured nonempty string.

"subject-full-stop": [2, "never", "."]

Both presets forbid a trailing period. The value is not restricted to one character; for example, [2, "always", "!"] requires an exclamation suffix.

# passes the preset
docs: explain configuration

# fails the preset
docs: explain configuration.

Type rules

type-case

Checks the parsed type against one supported case.

"type-case": [2, "always", "lower-case"]

Both presets require lowercase type. An empty type is skipped here and handled by type-empty.

# passes
feat: add parser

# fails
FEAT: add parser

type-empty

Tests whether the parsed type is empty.

"type-empty": [2, "never"]

Both presets use never, requiring a parsed type. A missing : separator or invalid header prefix can make the parsed type empty.

type-enum

Tests exact, case-sensitive membership in a nonempty list of strings.

"type-enum": [2, "always", ["feat", "fix", "docs"]]

With always, only listed types pass. With never, listed types are forbidden. An empty type is skipped and left to type-empty.

Conventional permits:

build, chore, ci, docs, feat, fix, perf, refactor, revert, style, test

Angular permits the same list except chore.

Case semantics

Case checks inspect ASCII letters and separators:

These are focused native predicates, not locale-aware text transformations. Empty fields are skipped by case and enum rules, so pair them with subject-empty or type-empty when presence is required.