Changelogs

Bug Fixes

• Test runs exceeding the 60-minute execution limit are now correctly marked as terminated instead of having a system error.
• Fixed an issue where internal node:vm frames were included in the call stacks for each test step.

Bug Fixes

• Fixed a race condition where test runs could occasionally be flagged with system errors prior to dispatch
• Fixed an issue where test workspace creation failures were not properly propagated and handled.

Improvements

• YAML & YML File Support: Added support for creating, uploading, viewing, and editing .yaml and .yml files in the workspace, with syntax highlighting and linting.
• Non-Test JavaScript File Support: Added support for creating non-test .js and .cjs files.
• Visible .test.js File Extensions: .test.js extensions are now explicitly visible in the project directory tree, editor title bar, and rename/clone dialogs.

Bug Fixes

• JSON Linting in Code Editor: Fixed false linting errors in valid JSON files

Major

Global HTTP Request Headers:

• Added new commands to support setting custom headers on all outgoing browser requests:
  • TEST.setRequestHeader(name: string, value: string) - to set a single header
  • TEST.setRequestHeaders(headerMap: object) - to set multiple headers using a map
• You can get and delete the headers with these commands:
  • TEST.getRequestHeaders() - get all request headers as a map
  • TEST.getRequestHeader(name: string) - get the value of a single request header
  • TEST.clearRequestHeaders() - get all request headers as a map
  • TEST.deleteRequestHeader(name: string) - delete a single request header

Project Configuration:

• Added support for project configuration files uilconfig.* or uilicious.config.* in JSON (recommended), YAML, and JS format. These files are auto-detected and useful for setting up common logic across tests, for example:
  • The headers config allows you to define a map of global request header values that will be configured at the start of the tests and applies to all outgoing requests.

Bug Fixes

• Fixed unexpected TypeError error masking the actual webdriver error in I.seeAlert, UI.COOKIE.isSet and UI.COOKIE.isEqual command.
• Fixed 2-3 second slowdown in the execution of each command due to unnecessary retries as a result of actual webdriver errors being masked by unexpected TypeError.
• Fixed race condition / edge case where in TEST.run may attempt to run a test file before its contents are fully downloaded and written.

Major

• Node 20 Upgrade: Updated the test engine runtime to Node 20 LTS for improved performance and security.
• Applitools Deprecation: Removed support for the legacy Applitools plugin.

Improvements

• Upgraded linter used for code editor

Bug Fixes

• Fixed: Expanding or collapsing folders in the test file selector no longer unexpectedly submits the create / edit job form.

Improvements

• Added --retryWhen <mode> option supporting always (default), never, and onSystemError to configure CLI retry behavior upon errors when starting a test run.

Fixed

• Fixed error extraction on test start failure so error messages and API responses are formatted clearly without stringifying as [object Object].

Deprecated

• Deprecated --disableSystemErrorRetry flag in favor of --retryWhen onSystemError (flag is now hidden from help documentation).

Improvements

• Improved how the I.fill command replaces values in input fields that already contain text. This resolves an edge case on some websites where clearing a field would trigger the original value to be restored, causing the new value to be appended instead of fully replacing the existing value. To ensure the value is always replaced, I.fill will now send Ctrl/Cmd + A when a field contains existing text before entering the new value.
• Improved element targeting by rebalancing attribute weights to prioritize attributes with higher intentionality, resulting in more accurate element matching.

Bug Fixes

• Fix an issue with downloading the files downloaded for a test run.

We've made several updates to improve the Auto-Pilot experience in TAMI Studio.

Improvements:

• Added a text input so you can send messages to steer the AI agent while a run is in progress.
• Added controls to pause and resume a run.
• Refined the interface to reduce visual clutter by hiding less important logs and making the UI more compact.