Hello, I am developing a skill that simply grabs a piece of information from the web and reads it to the user. The way I'd like my skill to work is without any intents and that is how it works when testing. However, I am having trouble with the Interaction Model part of the certification process. How can I get this by without an intent schema, and as a result no sample utterances. Any help is appreciated, thank you!
Hey guys, Just to chime in on my understanding of the rational behind providing at least one intent. Users trying to invoke your skill may often times tag on additional words and phrases as a natural part of the invocation process. Using the the popular Cat Facts skill as an example, simply saying 'launch Cat Facts' will return the desired result of a one-shot skill. However, when users understand the functionality of your skill (knowing Cat Fact's will provide me a fact), they may begin to inadvertently elaborate on the launch request. After using it multiple times or seeing someone else use the skill they might choose to say 'Alexa, launch Cat Facts and give me a fact about cats', or 'Alexa, ask Cat Facts for today's awesome cat fact'. Without providing sample utterances to accommodate for these variable lengths in invocations, the system may have a harder time interpreting the nature of your request, leading to misfires. These are great points that you guys brought up, so we'll continue to reach for more clarity to avoid these one-shot skill hurdles -- I hope the above explanation may help in understanding the current process.