Thrown into a task's promise when its worker's THREAD ENDED while the task was in flight — a clean self-termination (process.exit() inside the worker, or self.close() on runtimes that report it) as much as an abrupt one.

Distinct from WorkerTaskError, which carries an error the worker managed to report: a thread that simply stops raises nothing, so before the pool observed exits the only thing that ever settled such a task was the task timeout — and taskTimeoutMs: 0 disabled it, leaving the caller's promise pending and the pool slot claimed forever (X8-7).

Only raised on runtimes whose worker host reports an exit; where it does not (see IWorkerHost.reportsExit), the timeout remains the only backstop.