Send input actions over WebDriver BiDi targeting a specific browsing context, as a lower-level alternative to element-targeted Actions

Steps

1. Enable BiDi on the session in the browser Options for your language binding, since it is not on by default
2. Construct the BiDi input module against the driver, and obtain a browsing-context ID (e.g. the window handle) to target
3. Build one or more action sequences with the normal Actions builder, retrieve its underlying sequences, and pass them to the BiDi input module's perform call together with the context ID, targeting a context rather than a single WebElement
4. Call the module's release function afterward to reset key/pointer state for that context
5. Reach for this path when input needs to target a browsing context directly rather than being tied to one element, as with the classic element-scoped Actions API