Hi all,
Saw this git project to run aws services fully local…
Maybe we can use this to run Alexa voice messages etc with this to be fully local without the cloud usage of Amazon?
LocalStack supports a growing number of AWS services, like AWS Lambda, S3, DynamoDB, Kinesis, SQS, SNS, and many more!
Same issue. So I installed jellyfin and set up their alexa custom skill. Works perfectly.
Hi,
Is there a way to manually force ‘sensor.last.called’ to False for other devices based on the latest timestamp being true on one device from within a template?
When working with 2 accounts, despite being able to fix the sensor.last.alexa template to report back the right results in Developer Tools it seems nothing will affect the state of ‘last.called’ itself.
No, you will need to use a script or a notify group to set up a service you can call that uses your last called template to establish the target.
Hi guys, is there a way to use this integration in Echo Auto?
it is recognized as a device but it is not possible to interact. it seems it doesn’t even update the last call status
hello,
I’m in the same situation as you, so far I have not found any way to make it work.
Hi there,
I have a problem that I had several months ago for sometime and now again for some weeks (maybe introduced with an update?):
Normally if an alexa media player is playing something, the state of the player should change to something like “playing” and I can check that in an automation like this:
condition: not
conditions:
- condition: device
device_id: ec4a42e0f69cb98074f2be2d45d13147
domain: media_player
entity_id: media_player.echo_bad
type: is_playing
Problem: state does not change when I play something (e.g. a TUNEIN Radio station) it stays in e.g. “paused” no matter what happens on the echo device (attributes and state are not updated for more than 24 hours, maybe since last server restart).
I use an automation to start alexa play radio in the bathrom, when somebody enters or the light goes on, but as the player does not change its state, motion triggers the routine again and again when I move and the radio will restart with the notorious “TUNEIN” announcement. :
alias: Badezimmer Radio an wenn Licht an
description: >
Wenn im Badezimmer das Licht angeht, startet zwischen 7 und 22 Uhr SWR 3 auf
dem Echo Show Badezimmer
trigger:
- platform: device
type: turned_on
device_id: 3428092c344e1269e46daa0411a21ce3
entity_id: light.badlicht_1
domain: light
- type: motion
platform: device
device_id: 8b729bad721b7c2ac4631f9f79b20321
entity_id: binary_sensor.hue_motion_dusche_motion
domain: binary_sensor
condition:
- condition: time
after: "07:00:00"
before: "22:00:00"
- condition: not
conditions:
- condition: device
device_id: ec4a42e0f69cb98074f2be2d45d13147
domain: media_player
entity_id: media_player.echo_bad
type: is_playing
- condition: not
conditions:
- condition: state
entity_id: media_player.party_time
state: playing
action:
- service: media_player.play_media
target:
entity_id: media_player.echo_bad
data:
media_content_id: SWR 3
media_content_type: TUNEIN
metadata: {}
mode: single
Does anybody have the same symptoms and maybe a solution?
Thanks in advance
Greets
Yes, I’ve seen similar issues… but I don’t know why and it seems kind of random.
You may be able to use the last_called_summary attribute… make sure to double check what the attribute returns when you call the media_player.play_media on your instance. The following is what it would look like on mine:
- condition: not
conditions:
- condition: state
entity_id: media_player.echo_bad
attribute: last_called_summary
state: play swr 3 on tunein
Non-Amazon stations and skills do not always report a change in player state. You can see this with the official Alexa mobile or webapp. Since we mimic those apps, we can’t do anything about this until Amazon actually changes this.
Also, since you also appear to be in the EU (due to the German), Amazon servers have consistent 500 errors when querying player state for EU users. This is a technical error from Amazon’s servers which means we can’t get the player info even if it was an Amazon station or skill. This happens sporadically but you can see people noting it a lot more in the EU.
Installed latest update today, working again… I remember to have this happen some more times in the past… working… not working…
I’m installing this new for the first time today and running into problems that I’m not seeing mentioned on this thread. I have set up the 2FA in my Amazon account as instructed and am able to configure the AMP integration with my user/pass/2FA token. That takes me to am Amazon page to login with the user/pass and then confirm the 2FA, which I do. Then it returns me to HA/AMP where it says “Loading next step for Alexa Media Player” for maybe 30 seconds, then just says “Error” with an OK button. I am not sure what the error is. I tried enabling debug in HACS via the HA Integration HACS tile and repeating this process, but nothing new appears in the home-assistant.log . I tried enabling yet more logging via the steps indicated at Logs | HACS and also checked the box for “advanced debugging” when I was setting up AMP, but still nothing in the home-assistant.log file to even indicate it’s trying to set up AMP, much less the reason why it’s failing. I’m out of ideas to try. Any tips appreciated.
Can you please share some screenshots ?
I am having the following error. Not sure how to fix it. The error started showing with the latest updates in the past weeks. Thanks in advance for your help.
Logger: homeassistant.setup
Source: setup.py:298
First occurred: 9:16:54 AM (1 occurrences)
Last logged: 9:16:54 AM
Unable to prepare setup for platform alexa_media.sensor: Platform not found (cannot import name ‘UnitOfTemperature’ from ‘homeassistant.const’ (/usr/src/homeassistant/homeassistant/const.py)).
This is the sensor.py file the error is referring to.
I just noticed this myself today.
I could have sworn it had worked the past. At least I know I’m not alone in my delusion if it never did. 
I’m not even getting an update to the last_called_summary either.
The only thing that seems to work reliably is the last_called_timestamp.
So this is why my automation has stopped working!
I have speakers turn on or off depending of the state - in my case Standby. When I did looked the state was Paused but because historically the state has been unreliable, I never thought to look further. Doh!
From time to time, around every two weeks, I notice as well, that the media_player entities loose their connet to the server and therefore do not report the current state correctly. Here I need that in order to start the receiver, when the connected echo starts play music. I can only solve that by reloading the component. I’m even happy about this because before, I always had to restart HA. It’s a little annoying but I arranged with it. As soon as I have time I will write an automation to reload the component when the echo device doesn’t report “playing” after 2 Minutes or something similar…
You’re using an unsupported version of HA. That constant has been available since 2022.11.0. You either update HA or you can downgrade to a version of AMP < 4.5.1.
I just tried to reload the integration and the echo still doesn’t update the state to playing.
Sometimes only restarting HA helps. I had this recently.
