Element detached after actionability check in webServer configuration

Playwright's auto-waiting mechanism performs actionability checks to ensure an element is visible and stable before performing actions like clicks. However, in highly dynamic single-page applications (SPAs), a component may re-render or be replaced immediately between the successful visibility check and the actual execution of the action.

When the element is detached from the DOM during this narrow window, the test often fails because the locator no longer points to a valid node. This behavior is particularly relevant when using the webServer configuration to manage local application instances during test suites.

• How does Playwright handle cases where an element is detached specifically between the actionability check and the action?
• Is there a specific configuration or locator strategy to mitigate these race conditions without resorting to manual retry loops?