What I'd like for my app is to get more complex configuration data, including properties like multiple names and email addresses. Currently Echo is not so great at this. To recognize a name you have to do like scorekeeper and give all possible variants of a name and for email addresses I don't know how you would do it in any practical way. Perfectly understandable, early days and all that. So the thought is to make up a URL based on the user id. That URL would be sent to the echo app and the user could retrieve it, visit the site, setup their password, and input their configuration. That configuration would go into dynamodb and then skill would retrieve on invocation. Not ideal, but it's the best I could come up with. Is there a better way to go about this? thanks
The best way for now is to have your own companion web app. Use a sync code (like Firefox) to synchronize between your echo app and your web app. Do all the one-time setup or lengthy data-entry from your web app. The only short term thing Amazon could do to make that easier is to have clickable links in the companion app.
I'm hoping to hear of some other alternatives as well. We have a service, which creates separate urls for each user to communicate with a system running in their home. They set the username/password (digest authentication) and we just pass the traffic from our hosted service to their system via ssh tunnels. This is great because it means from a security perspective we don't know the user's credentials and can't get into their system. This is a big problem for Echo since there's no way for the user to enter their URL, username, and password so the skill can connect to their system. Any other thoughts would be great.