When a user held the mouse cursor over blacked-out components of a tweet, the worm in the script would quickly open up links and re-write-up alone on the reader's account.