Skip to main content

UI Automation Setup Guide — Complete Reference

Sample test cases, glossary, and best practices to ensure your UI automation runs successfully every time.

Jump to: Sample Test CasesGlossaryBest Practices

Sample Test Cases

Copy these examples directly into the Create Suite editor. Replace URLs, field names, and values with your application's specifics.

A standard end-to-end login test. Uses fill, click, and assert actions.

Feature: User Login Scenario: Successful login with valid credentials Given I navigate to "https://your-app.com/login" When I enter "user@example.com" in the "Email" field And I enter "SecurePass123" in the "Password" field And I click the "Sign In" button Then I should see "Welcome back" And I should see the dashboard
TIPUse the exact visible label of the field (e.g. Email, not email_input). The AI matches by visible text and placeholder, not by HTML attribute names.

Glossary

Key terms used throughout the UI Automation system.

BDDBehavior-Driven Development. Tests written in natural language (Given / When / Then) so both technical and non-technical stakeholders can read them.

GherkinThe language used for BDD test cases. Keywords: Feature, Scenario, Given, When, Then, And, But.

TDD FormatTest-Driven Development style input — numbered steps with an Expected Result section. Detected automatically from Steps: or Test Case ID: headers.

LocatorA selector that uniquely identifies a UI element on the page. Can be an XPath expression or CSS selector. The AI resolves these automatically.

XPathXML Path Language — an expression used to locate elements in the page DOM. Example: //button[text()='Sign In'].

CSS SelectorA pattern that matches elements using CSS syntax. Example: button.btn-primary. Faster than XPath in most browsers.

data-testidA stable HTML attribute added by developers specifically for test automation. Example: data-testid="login-button". The AI prioritises these.

Stage D (DOM-first)The first stage of locator resolution — it searches the live DOM semantically (label, placeholder, aria, visible text) before using AI vision. Free, fast, and always runs first.

JIT LocatorJust-In-Time locator resolution. The AI resolves each element's locator immediately before that step executes, ensuring it is resolved against the correct live page.

Self-HealingWhen a step fails because the locator is stale, Agent 6 automatically re-resolves the element and retries the step. Healed locators are cached for future runs.

Locator CacheA persistent store of previously resolved (and validated) locators. Cache hits skip AI calls, making re-runs faster and cheaper.

Compound LoginA special action that fills username, password, and clicks submit in a single step. Triggered by phrases like "log in with email X and password Y".

Assert / VerifyA step that checks a condition without interacting with the page. Example: "I should see 'Welcome'". Does NOT click or type anything.

Base URLThe root URL of the application under test (e.g. https://staging.your-app.com). Set in Environment Settings before running suites.

Test SuiteA named collection of test cases grouped for a specific feature or flow. Suites are run against an environment and produce a pass/fail report.

Smoke TestA minimal set of critical-path tests that verify the app is basically functional. Run after every deployment to catch major regressions fast.

Best Practices

Follow these guidelines to maximise run stability and minimise failures.

Writing Steps

Each step should perform exactly one browser action. Combining actions in a single step confuses the AI classifier and produces unreliable results.

AVOIDWhen I enter my email, password and click login
PREFERWhen I enter "user@test.com" in the "Email" field
And I enter "pass123" in the "Password" field
And I click the "Login" button

Environment Setup

App-Side Improvements

Debugging Failures

Need more help? Use the bug report form or email support@qaeverest.com. Include your base URL and a copy of the failing test case for the fastest response.

Explore Now!