Gherkin features linter written in Typescript

Usage no npm install needed!

<script type="module">
  import gherkinLintTs from '';


Travis Coverage Status npm

Gherkin lint

Uses Gherkin to parse feature files and runs linting against the default rules, and the optional rules you specified in your .gherkin-lintrc file.

The original idea came from this project:

The changes are:

  • The project has been rewritten to Typescript
  • Supports for writing your own rules in Typescript
  • Supports error level (Error/Warn)
  • Pretty terminal output
  • Error output was changed from console to process.stdout, to have ability yarn --silent lint:features > features-errors.log
  • Fixed a number of issues that came from the original lib (incl. vsiakka#254: Using Rule keyword)
  • A couple more changes that will be described as soon as the project is finally stabilized


npm install gherkin-lint-ts


To see the output for all the errors that the linter can detect run:

git clone
yarn demo

Or check this: console

Available rules

Name Functionality
no-tags-on-backgrounds * Disallows tags on Background
one-feature-per-file * Disallows multiple Feature definitions in the same file
up-to-one-background-per-file * Disallows multiple Background definition in the same file
no-multiline-steps * Disallows mutiline Steps
allowed-tags Just the listed tags are allowed
file-name Restrict feature file names to a commmon style
indentation Allows the user to specify indentation rules
max-scenarios-per-file Allows the user to specify the max number of scenarios per feature file
name-length Allows restricting length of Feature/Scenario/Step names
new-line-at-eof Disallows/enforces new line at EOF
no-background-only-scenario Disallows background when there is just one scenario
no-dupe-feature-names Disallows duplicate Feature names
no-dupe-scenario-names Disallows duplicate Scenario names
no-duplicate-tags Disallows duplicate tags on the same Feature or Scenario
no-empty-background Disallows features with backgrounds without steps
no-empty-file Disallows empty feature files
no-examples-in-scenarios Disallow the use of "Examples" in Scenarios, only allowed in Scenario Outlines
no-files-without-scenarios Disallows files with no scenarios
no-homogenous-tags Disallows tags present on every Scenario in a Feature, rather than on the Feature itself
no-multiple-empty-lines Disallows multiple empty lines
no-partially-commented-tag-lines Disallows partially commented tag lines
no-restricted-patterns A list of patterns to disallow globally, or specifically in features, backgrounds, scenarios, or scenario outlines
no-restricted-tags Disallow use of particular @tags
no-scenario-outlines-without-examples Disallows scenario outlines without examples
no-superfluous-tags Disallows tags present on a Feature and a Scenario in that Feature
no-trailing-spaces Disallows trailing spaces
no-unnamed-features Disallows empty Feature name
no-unnamed-scenarios Disallows empty Scenario name
no-unused-variables Disallows unused variables in scenario outlines
one-space-between-tags Tags on the same line must be separated by a single space
required-tags Require tags/patterns of tags on Scenarios
scenario-size Allows restricting the maximum number of steps in a scenario, scenario outline and background
use-and Disallows repeated step names requiring use of And instead
keywords-in-logical-order Requires that Given, When and Then appear in logical sequence

* These rules cannot be turned off because they detect undocumented cucumber functionality that causes the gherkin parser to crash.

Rule Configuration

The not-configurable rules are turned on by default and cannot be turned off. Configurable rules can be customized using a file.

The configurable rules are off by default. To turn them on, you will need to create a json file, where you specify the name of each rule and its desired state (which can be error, warn or off). E.g.,

  "no-unnamed-features": "error"

will turn on the no-unnamed-features rule.


allowed-tags should be configured with the list of allowed tags and patterns:

  "allowed-tags": ["error", {"tags": ["@watch", "@wip"], "patterns": ["^@todoquot;]}]

Any tag not included in this list won't be allowed.


file-name is configured with a style to enforce. The default is PascalCase:

  "file-name": ["error", {"style": "PascalCase"}]

The list of supported styles is:

  • PascalCase - first letter of each word capitalized (no spaces) e.g. "MyFancyFeature.feature"
  • Title Case - first letter of each word capitalized (with spaces) e.g. "My Fancy Feature.feature"
  • camelCase - first letter of each word capitalized, except first e.g. "myFancyFeature.feature"
  • kebab-case - all lowercase, hyphen-delimited e.g. "my-fancy-feature.feature"
  • snake_case - all lowercase, underscore-delimited e.g. "my_fancy_feature.feature"


no-restricted-patterns is a list of exact or partial patterns whose matches are dissallowed in feature name and description, and in background, scenario and scenario outline name, description and steps. All patterns are treated as case insensitive. The rule can be configured like this:

  "no-restricted-patterns": ["error", {
    "Global": [
      "^globally restricted pattern"
    "Feature": [
      "poor description",
    "Background": [
      "show last response",
      "a debugging step"
    "Scenario": [
      "show last response",
      "a debugging step"


  • Step keywords Given, When, Then and And should not be included in the patterns.
  • Description violations always get reported in the Feature/Scenario/etc definition line. This is due to the parsed gherkin tree not having information about which line the description appears.


indentation can be configured in a more granular level and uses following rules by default:

  • Expected indentation for Feature, Background, Scenario, Examples heading: 0 spaces
  • Expected indentation for Steps and each example: 2 spaces

You can override the defaults for indentation like this:

  "indentation" : [
    "error", {
      "Feature": 0,
      "Background": 0,
      "Scenario": 0,
      "Step": 2,
      "Examples": 0,
      "example": 2,
      "given": 2,
      "when": 2,
      "then": 2,
      "and": 2,
      "but": 2,
      "feature tag": 0,
      "scenario tag": 0

There is no need to override all the defaults, as is done above, instead they can be overriden only where required. Step will be used as a fallback if the keyword of the step, eg. 'given', is not specified. If feature tag is not set then Feature is used as a fallback, and if scenario tag is not set then Scenario is used as a fallback.

This feature is able to handle all localizations of the gherkin steps.


The max-scenarios-per-file supports some configuration options:

  • maxScenarios (number) the maximum scenarios per file after which the rule fails - defaults to 10
  • countOutlineExamples (boolean) whether to count every example row for a Scenario Outline, as opposed to just 1 for the whole block - defaults to true

The configuration looks like this (showing the defaults):

  "max-scenarios-per-file": ["error", {"maxScenarios": 10, "countOutlineExamples": true}]


name-length can be configured separately for Feature, Scenario and Step names. The default is 70 characters for each of these:

  "name-length" : ["error", { "Feature": 70, "Scenario": 70, "Step": 70 }]


new-line-at-eof can be configured to enforce or disallow new lines at EOF.

  • To enforce new lines at EOF:
  "new-line-at-eof": ["error", "yes"]
  • To disallow new lines at EOF:
  "new-line-at-eof": ["error", "no"]


no-dupe-scenario-names can be configured to search for duplicates in each individual feature or amongst all feature files. To enable searching for duplicates in each individual feature (same scenario name in different features won't raise an error) you need to configure the rule like this:

  "no-dupe-scenario-names": ["error", "in-feature"]

The default case is testing against all the features (same scenario name in different features will raise an error). To get that behavor use the following configuration:

  "no-dupe-scenario-names": "error"


  "no-dupe-scenario-names": ["error", "anywhere"]


no-restricted-tags should be configured with the list of restricted tags and patterns:

  "no-restricted-tags": ["error", {"tags": ["@watch", "@wip"], "patterns": ["^@todoquot;]}]


required-tags supports some configuration options:

  • tags (array) the array of tag patterns that must match at least one tag - defaults to []
  • ignoreUntagged (boolean) whether to ignore scenarios that have no tag - defaults to true
  "required-tags": ["error", {"tags": ["^@issue:[1-9]\\d*quot;], "ignoreUntagged": false}]


scenario-size lets you specify a maximum step length for scenarios and backgrounds. The Scenario configuration applies to both scenarios and scenario outlines:

  "scenario-size": ["error", { "steps-length": { "Background": 15, "Scenario": 15 }}]

Configuration File

The default name for the configuration file is .gherkin-lintrc and it's expected to be in your working directory.

The file contents must be valid JSON, though it does allow comments.

If you are using a file with a different name or a file in a different folder, you will need to specify the -c or --config option and pass in the relative path to your configuration file. Eg: gherkin-lint -c path/to/configuration/file.extention

You can find an example configuration file, that turns on all of the rules in the root of this repo (.gherkin-lintrc).

Ignoring Feature Files

There are 2 ways you can specify files that the linter should ignore:

  1. Add a .gherkin-lintignore file in your working directory and specify one glob pattern per file line
  2. Use the command line option-i or --ignore, pass in a comma separated list of glob patterns. If specified, the command line option will override the .gherkin-lintignore file.

Custom rules

You can specify one more more custom rules directories by using the -d or --dirs command line option. Rules in the given directories will be available additionally to the default rules.


gherkin-lint -d "/path/to/my/rulesdir" "from/cwd/rulesdir"

Paths can either be absolute or relative to the current working directory. Have a look at the src/rules/ directory for examples; The no-empty-file rule is a good example to start with.