If I understand this correctly, your ground has an area2d which detects the player’s area2d child? If so, you should be using the “area_entered” signal instead of “body_entered”. Kodi
The player collides with an object at a door, so the text will load, but it should only be visible when in contact with the object, for a short time, and only loaded once. Otherwise, the instruction will always be loading, if the action is already known, it wouldn’t make sense for it to be visible.