Thanks Keith for sharing a mechanism to get the raw text of the sound in the request to the skill app. It has saved me a lot of time to figure it out myself.