Next-Gen App & Browser Testing Cloud
Trusted by 2 Mn+ QAs & Devs to accelerate their release cycles

Turn Cucumber feature files into KaneAI tests written as natural language steps. Conversion runs in your browser.
A Gherkin to KaneAI converter is a tool that turns Cucumber feature files, written in the Gherkin syntax, into KaneAI tests. Each Gherkin scenario becomes a numbered list of natural language steps that KaneAI runs in a real browser, with no step definitions or glue code to maintain.
In Cucumber, every Gherkin step is bound to a step definition: code that someone writes and keeps in sync with the application. A KaneAI test needs no step code, because the agent reads the instruction and acts on the page. TestMu AI built this converter for teams moving an existing Cucumber suite to KaneAI: it does the mechanical part of the move, so you start from your own scenarios instead of a blank file. To weigh the options first, see Cucumber alternatives, which shows one test written in Gherkin, in Playwright and in KaneAI.
The converter reads every keyword in the Gherkin reference and writes each scenario as a KaneAI test: a title followed by numbered steps, the same shape as the KaneAI test on Cucumber alternatives.
| Gherkin | KaneAI test |
|---|---|
| Feature name | The first part of each test title |
| Scenario or Example | One KaneAI test |
| Background | Steps added at the start of every test |
| Scenario Outline with Examples | One test per Examples row, with placeholders filled in |
| Given, When, And, But | A numbered step written as an instruction (Open, Click, Enter, Select) |
| Then | A numbered step that starts with Verify |
| Data table | Values listed in the step, such as Name = Asha, City = Pune |
| Doc string | Text added to the end of the step |
| Quoted text and URLs | Kept exactly as written |
KaneAI finds the elements on the page as it runs, so each step only has to say what to do and which text to look for. That is why the converter keeps quoted labels and values exactly as they appear in the feature file.
This feature tests the login page of the TestMu AI eCommerce Playground, a demo store built for practicing test automation. Input:
@login
Feature: Account login
Background:
Given I am on "https://ecommerce-playground.lambdatest.io/index.php?route=account/login"
Scenario: Wrong password shows a warning
When I enter "shopper@example.com" into the E-Mail Address field
And I enter "wrong-pass" into the Password field
And I click the "Login" button
Then I should see "Warning: No match for E-Mail Address and/or Password."Output, as a KaneAI test:
Account login: Wrong password shows a warning
1. Open https://ecommerce-playground.lambdatest.io/index.php?route=account/login
2. Enter "shopper@example.com" into the E-Mail Address field
3. Enter "wrong-pass" into the Password field
4. Click the "Login" button
5. Verify "Warning: No match for E-Mail Address and/or Password." is visibleThe Background step became step 1, each When step became a direct instruction, and the Then step became a Verify check. The converter also notes that the password is in plain text, so you can swap in a test account before adding it to KaneAI.
A feature file usually holds several scenarios, and each one becomes its own KaneAI test. The output choice only changes how you receive them:
Either way, the output panel shows each test on its own tab in Multiple files mode, so you can copy one test at a time.
The Gherkin reference advises that implementation details belong in step definitions, so many feature files describe behavior (When they complete checkout) rather than clicks. KaneAI has no step definitions to fill that gap, so the detail has to live in the step. The converter flags steps that look too vague; these habits make them run reliably:
Need scenarios to start from? The User Story Generator and Test Case Generator draft them from a feature description. If you would rather keep running your feature files unchanged, HyperExecute runs Cucumber suites as they are (see Cucumber on HyperExecute).
No. The Gherkin to KaneAI converter runs entirely in your browser and writes the KaneAI steps on the same page, so scenario text, test data and credentials never leave your machine. Nothing is sent to a server, so you can convert feature files that contain internal URLs or test accounts without exposing them.
The Gherkin to KaneAI converter expands a Scenario Outline into one test per Examples row and fills every placeholder with that row's values. Each row becomes a separate KaneAI test with its own title, so a three-row table gives three tests. In Multiple files mode each of them is downloaded as its own text file.
Gherkin runs Background steps before every scenario, so the Gherkin to KaneAI converter copies them to the top of each converted test. That keeps every KaneAI test self-contained: a test made from one scenario opens the same start page and sets up the same context the Background described, without depending on another file.
Cucumber advises hiding implementation details in step definitions, so a step such as When they complete checkout describes behavior, not clicks. The Gherkin to KaneAI converter keeps such steps but flags any with no quoted text, URL, data table or recognized action. KaneAI still attempts them; spelling out the clicks and the exact text makes the run more predictable.
Yes. The Gherkin to KaneAI converter reads Gherkin in more than 70 spoken languages, declared with a # language: comment on the first line. It maps the structure of any of them: scenarios, Background, Outline rows and tags. Step rewording is tuned for English, so steps in other languages are kept as written.
Create a test in KaneAI and paste the converted steps as its instructions, one numbered step per line. KaneAI finds the elements on the page as it runs, so the steps need no step definitions or selectors. If you are new to KaneAI, start with the KaneAI getting started guide.
Did you find this page helpful?
TestMu AI forEnterprise
Get access to solutions built on Enterprise
grade security, privacy, & compliance