There's no reason Flash Briefings should be confined to ports 80 and 443, as webservices can (and do) run on any almost any port. This restriction causes serious limitations for (among others) people trying to run multiple services on the same EC2 node.

Right now Flash Briefings won't parse SSML in their preamble, error message, or text output. This means content written for a custom skill (for example) cannot be used for Flash Briefings, and any output that could be massaged to sound correct will instead come out sounding wrong.

Love it! So clean and consistant. One feedback on the left hand navigation. Make it scrollable so when we have lots of intents open we are able to see the right side of the page. Or better yet keep the right side of the page fixed and we can scroll the left hand navigation freely and open / collapse intents etc. Keep up the good work team! :-)

Please auto save! I've lost many edits on the older-version of this tool due to the lack of autosave. :)

New i/f is neat. Thanks.
It would be good if you could add the link to CREATE a Lambda endpoint again (it was on the old console).

Right now if you're building something on top of SMAPI you're not able to understand is your flash briefing popular or not without going to the UI.

Currently, if a dialog is mid progress, it overrides any displays that we may want to show relating to the dialog. Would greatly enhance user experience on multi-modal skills.

If I create a new Routine in the Amazon Alexa app on IOS Device I can currently only add the following actions:
Smart Home, Traffic, Wether, News.
I would like to have a action which stops Alexa from playing any kind of music or other sound.
Background: You create the Routine "Goodby" which should turn off all HUE lights and it should stop the Radio(tuneIn on Alexa, or Spotify music). The feature of stopping Music is not available in the app. If I say "Alexa, Stop" the music is stopped. Why is that not available in the app?

The VideoApp for the Echo Show only supports to play one video. Once this video is started the current skill looses control. Please add a queue to the VideoApp like the AudioPlayer already supports.

Allow users to reorder items within a list ie; move an item up or down and allow sub-categories.
Also, provide a way to see all items on all lists and move an item to another list

This feature would be similar to playing music with an Alarm but it would open a skill.

With many intents that share baseline handling intents (help,cancel,etc.) it would be nice to be able to define all these at a top level, and have specialized states cascade down. e.g. Define NewSession state with all the default handlers. In a new, specialized state, SpecialState, you only define those intents that are specific to that state, and any non defined events cascade down to NewSession state.

This would allow more simplified code, as well as the ability to better maintain consistency through out experience.

While developing alexa skill I have to use 2 different portals (aws console and dev portal). Would be more convenient if I just had one

At times, request processing takes more than it is allowed by Alexa timeout. I want to be able to wake Alexa up and give user feedback

When Fire TV interacts with Alexa popup is displayed on the TV with dialog conversation blocking TV screen. If my firetv app can respond to alexa command I would like to utilize TV for displaying app related information instead of that dialog

The interaction model has two view - the old preview one and the new Beta preview one. They support different features and switching in between, deletes the changes made. The only reason why my team is still using the old preview, is because we can easily upload new and edit the old training. Would be great if this is also possible in the Beta preview, as to use the other new features available, like synonyms, dialogues, etc.

Expanding the current implementation of Push would unlock a massive developer opportunity. Please allow devs to build and users to select skills that can speak without the user initiating the conversation. If a skill gets annoying they can turn it off.

You can't use the API outside of a website with the current model. Ex: would like to run Lambda function to update a skill model and then submit for cert. You can't authenticate to use the API in this context.