My idea is to issue a command to Alexa that'll make it ssh into a server connected to the same LAN and issue commands (e.g., "sudo reboot", "sudo shutdown now") or run scripts.

Enable Echo Show camera to be used for security. If in built echo show camera detects motion, send notification to phone / other nominated Alexa devices.

Have AMAZON.Number accept decimal points in user verbal input. Right now, it handles 'minus' or negative numbers, but not when there is a decimal point.

Allow for the Echo Show to mute audio while on a video call without also stopping the video.

I would love a widget for Alexa's shopping list for both my android and OS Wear.

APL video "loaded" event to enable showing a screen before a video has fully loaded so that the user is aware that the media is still loading. e.g., "Loading..." Right now, APL only has events for after the video has started playing. This is too late to elegantly switch from a "loading" screen to the video.

In part, Alexa's voice is too loud for me, for example at night. It would be great if Alexa automatically adjusts the volume to the volume of my voice, just like people do with each other. (Certainly hard to determine if the user is talking quietly or just far away)

The volume can also be specified as a percentage, which is great. However, only 5% steps seem to be possible. For quiet music while working, however, I would like to have about 3% volume, which is unfortunately only adjustable via the smartphone.
Therefore, the possibility to adjust the volume with an accuracy of 1% would be very helpful.

Alexa Routines are nice but it's still too tech-y to set up. If a custom skill has some actions suitable for using as routines, it would be great for a custom skill to recommend, ask permission from user, and set routines instead of users. This will reduce user's conflictions and increase skill usage.

Please allow beta tester to be able to add private skills to Routine.

It would greatly help if the Deploy button, under the Code tab, could check and warn if there were any code errors before deploying the lambda code. Right now I can deploy code with error or even an empty js file. Ideally, the Deploy operation should fail or stop with a good message.

In routines option for Music and Media action. I'd like the ability to control which speaker the action plays on. Instead of the only option being for the whole routine. For example having a routine play a spotify playlist on 1 speaker speaker while white noise skill on another speaker. All from one voice command. This would allow for more flexible routines based on bed-times and wake-ups.

Sometimes, I'd have to (more or less) "continue" an interrupted SSML response. I understand that it is not possible to find out if and where a response has been interrupted.

Therefore, it would come very handy if I could find out beforehand how long a response might take to be read. From this (plus timestamp of this response and the next intent's) I can calculate how far reading might have proceed.

So, my request is:
- enter a text string plus device reading speed
- calculate approximate expected reading duratioin from this

RIght now, the simulator always sends a response for an APL device with a display. Since we have to also support devices without displays (headless devices), it would be good to also have an option for simulating the response of a headless device like the very common Echo/Echo Dot.

The current limit is 4 minutes. Our app plays backing tracks for guitarists and this is way too short. Please increase to at least 20 mins.

Ability to force Alexa always talk in whisper mode, even if I speak normally.

The purpose is, when I am playing music on high volume, Alexa doesn't need to speak so loud to me.

Another alternative would be to have separate volume control for Alexa interations

It would be very useful, when using the APL Authoring Tool and in play mode, having a separate console that logs all the fired commands. Especially the SendEvent command could be very useful to get more details on (a list of all the listed arguments), especially when working with dynamic values that might change throughout interaction with the APL document.

If support bangla language it's help for Bangladeshi region people and easy to use it.