Preview

How to Build

You can preview the intents, utterances, voice responses, and card assets in this spreadsheet: https://docs.google.com/spreadsheets/d/1W9EsfAOC9XOc57SOsgl0TGdwcINa7IwK4w4SpK_zFbo/edit#gid=0

There are four main parts for a Skill prototype like this: intent enumeration, utterance variation design, batch response design, and card design. Designing for voice occurs in the "VUI" tab of a Spreadsheet whereas the card assets will need to be uploaded to a public storage and referenced in the "Cards" tab.

This design has 3 intents around dog matching. DogBreedsIntent, BestBreedIntent, HowManyBreeds have their respective utterance variations. Each of these intents have 1 type of voice response and 1 type of card asset response.

We tend to use Dropbox to host our cards. After you upload your card assets based on the Makeskill Templates then you can make a public shareable URL for each. Then you'll need to reference that as a raw asset: https://www.dropbox.com/s/li0qm8uu1zxjd06/list_1.png?raw=1

Notice the ?raw=1 at the end of the public URL. This is a way of specifying that we want to grab a raw asset. This is crucial so that the Makeskill CLI can transform that URL into a background image as a response for a 1-turn interaction.