Settings

BlinqIO gives you fine-grained control over how the AI Recorder captures interactions and how the execution engine runs your tests. The Settings page is where you configure browser behavior, timeouts, popup handling, and advanced Playwright options: all without writing code.

AI Settings

The AI Settings tab provides a JSON editor for customizing the AI Recorder and Test Execution engine. Every configuration key maps to a specific behavior, and the editor includes autocomplete so you can discover available options as you type.

How to Modify AI Settings

1. Navigate to Settings in the left sidebar and select the AI Settings tab.
2. Click Edit Settings.
3. Add or edit configuration keys in the JSON editor.

TIP

Start typing a property name (e.g., "view...") to see autocomplete suggestions. Selecting a key from the list provides a sample value you can modify.

4. Click Save.

The editor validates your input against a predefined schema. If there are syntax errors or invalid data types, the save will fail with a message indicating the issue.

Configuration Reference

1. Viewport and Browser Appearance

Key	Type	Description
viewport	Object	Sets the browser window width and height (e.g., { "width": 1200, "height": 800 }).
noViewport	Boolean	If true, the host OS decides the window size. Can lead to inconsistent results across machines.
fullPageScreenshotOnFailure	Boolean	When true, captures a screenshot of the entire scrollable page on test failure, not just the visible area.
load_all_lazy	Boolean	If true, the runner scrolls through the page to trigger lazy-loaded elements before proceeding.

2. Browser Engine Selection

By default, tests run on a bundled Chromium browser. Use these keys to override:

Key	Type	Description
useGoogleChrome	Boolean	Use the Google Chrome installation on the host machine.
useMicrosoftEdge	Boolean	Use Microsoft Edge.
overrideUserAgent	String	Spoof the browser's User Agent string for device-specific testing.

3. Execution and Performance

Key	Type	Description
fastMode	Boolean	Speeds up execution. Use with caution: steps may fail if the application needs time to render.
page_timeout	Number	Maximum time in milliseconds to wait for a page to load. Default: 60000.
find_element_timeout	Number	Maximum time in milliseconds the AI searches for an element before timing out.
stableLocatorStrategy	String	Preferred element identification strategy at runtime (e.g., prioritize data-testid over CSS selectors).

4. Dialogs and Popups

Key	Type	Description
acceptDialog	Boolean	Automatically clicks "OK" or "Accept" on native browser alerts and dialogs.
closePopups	Boolean	Automatically attempts to close unexpected popups during execution.
popupHandlers	Array	Specific locator and action pairs for handling application popups. See Handling Dynamic Popups.

5. Network and Synchronization

These settings control when the execution engine considers a page "ready" and proceeds to the next action:

Key	Type	Description
networkidle	Boolean	Wait until there are no active network requests for at least 500ms.
load	Boolean	Wait for the standard browser load event.
domcontentloaded	Boolean	Wait for the HTML to be fully loaded and parsed.

Advanced: Context Options

The contextOptions object maps directly to the Playwright Browser Context API. Use it to simulate specific user environments:

"contextOptions": {
  "ignoreHTTPSErrors": true,
  "locale": "en-GB",
  "timezoneId": "Europe/London",
  "geolocation": {
    "latitude": 51.5074,
    "longitude": -0.1278
  },
  "permissions": ["geolocation", "microphone", "camera"]
}

Common use cases: testing locale-specific formatting, verifying geolocation features, allowing browser permissions that your application requires.

Browser Options: Tag-Based Overrides

The browserOptions object lets you define named configuration sets that activate only when a scenario has a matching Gherkin tag. Tag your scenario with @browserOptions__<key> to apply the override.

Example: Run a specific scenario in a Japanese locale:

"browserOptions": {
  "languageAutodetect": {
    "locale": "ja-JP"
  }
}

Then tag your Gherkin scenario:

@browserOptions__languageAutodetect
Scenario: Verify Japanese locale formatting
  Given I navigate to the settings page
  ...

This keeps your default configuration clean while allowing per-scenario overrides for locale, viewport, or any other context option.

Handling Dynamic Popups

Applications often display popups at unpredictable times: cookie consent banners, promotional overlays, chat widgets. Instead of failing when these appear, configure popupHandlers to dismiss them automatically.

Each handler needs two pieces of information:

1. Popup locator: How to identify the popup element (CSS selector or text).
2. Action button locator: Which button to click to dismiss it.

Add the handlers in AI Settings:

"popupHandlers": [
  {
    "popup": "text=Accept Cookies",
    "action": "text=Accept"
  },
  {
    "popup": ".promo-overlay",
    "action": ".promo-overlay .close-btn"
  }
]

The execution engine checks for these popups before each step and dismisses them if found, keeping your test flow uninterrupted.

Application Map

The Application Map is a visual representation of your test coverage. It shows:

• The navigation flow between pages in your application.
• ARIA snapshots of each visited page.
• Which steps have been executed on each page.

The map builds automatically as you record and run tests, giving you a living picture of how thoroughly your application is covered.

Application Memory

As the Application Map grows, BlinqIO builds Application Memory in the background. This memory captures patterns in your application's structure and behavior, which the AI uses to:

• Generate more accurate scenarios from natural language descriptions.
• Suggest more reliable element locators.
• Predict likely user flows for new features.

Application Memory improves over time as more tests are recorded and executed.

Additional Options

Skip run reports: Set UPLOADREPORTS=false in AI Settings to suppress report uploads after test runs. You can also pass it per execution as an environment variable.

Full page screenshot on failure: Set fullPageScreenshotOnFailure to true to capture the entire scrollable page when a test fails, not just the visible viewport. This is useful for debugging failures that involve off-screen content.