[moad]

Thanks, it is enabled there. I tried disabling and re enabling too.

[m0ngr31]

At the bottom of that page, it should say what the invocation name is. Double check that it is Kodi. If not, try saying kanzi.

[moad]

definitely kodi

I will try and create a skill manually to see if I can invoke it

[moad]

hmm well when testing the new skill I get an audio only response. If I try it from my alexa app I just get "i don't know that", not sure how to determine if it is actually invoking the skill or not in this instance.

If I type in the test and ask "kanzi" I also get an audio only response but if I type kodi it says can't find the skill. Invocation name is definitely kodi.

I might just start over from scratch

[moad]

same issue, everything deploys fine...invocation name is kodi...I can see it has pulled data from Kodi but Alexa won't call the skill.

Even when I type it in to the test section of the skill it can't find it. 

Is this an amazon issue?

[m0ngr31]

Make sure you are on your Amazon account. My wife is always switching the profile to hers so it always throws me off when I try to use the skill.

[moad]

Yeah my wife doesn’t have an account and I only have one account.

Thanks anyway I’ll just revisit later and play with some other stuff

[moad]

there is another add on that I got working, just had to change the region of my amazon account. I still don't know why I couldn't call the skill, would much rather use this add on tbh.

If anyone in Australia has this working please PM me

[m0ngr31]

That's strange. I'm creating the Australian locale files. Does the skill work now that you've moved your account to a different region?

[LuckyTNT]

Thanks for nice project.
Can I please have some help, I am trying to get it running with no luck.

I tried following guide procedure multiple times on 2 different computers (ubuntu/win 10).
I always got stuck on deploying skill.

I got following error:

Update slot values
  → Kodi server configuration settings are wrong. Please double-check kodi.config to make sure it is setup properly.

errorlog.txt:
Error: Error talking to Kodi.

I can't really can find reason why is that when I can access to Kodi web interface from Internet without problem over my duckdns address.
I am running kodi 17.6 (default kodi web interface Chorus2).

So any ideas what can cause this problem that I can try to fix?

I would like to finally make it docker image and host it my homelab server for future but can't even get it work on AWS Lambda.

[trubio2]

I have the skill working, but as soon as I switch on my VPN , Alexa tells me there was a problem with the requested skills response. If I turn off VPN it will work again. What do I do to fix this?

[Klojum]

VPN is not a initial requirement for Kodi. If it doesn't work without it, that's too bad.

Any technical discusssion on VPNs in general can be done in the Off-topic section of this forum, any other chats on VPN are off-limits.

[trubio2]

My invocation name is set to kanzi. I want tbchange it to Kodi. When I change it when I deploy the skill the invocation name doesn't change.

[bisca330]

Hello,

I am copying here some questions that I published wrongly on Github and I also provide what I was told (jingai user) in order to continue the exchange :

I managed to deploy Kanzi in French, and it works. I did some tests and I already noticed things that could be improved.

1) I noticed that Alexa is listening only a few seconds when we say "Alexa ask kodi ...". So I wonder if we can change the call word to the skill and make sure that Kodi stays active until we say another command so that kodi does not listen anymore?
For example :
"Alexa, activate Kodi" -> for kodi to start listening without calling Alexa at each order.
"Thank you Kodi" -> for kodi to stop listening.
Of course when we start a media Kodi will automatically stop listening.

Answer :
We do keep the session open (at least for as long as Amazon allows), but not for all Intents. Basically, any Intent that would start playback of media, the session is immediately closed; most other Intents will open the session again and Amazon will time it out after 10 seconds.


2) I use Kodi's Aeon Nox skin and set up my menu banner with items like "Movies" for all movies, "Kids" for cartoons, "Anime" for mangas that are a bit violent for children or "Shows". With the skill Kanzi we can go to the menu ("Alexa ask Kodi home") and we can walk on the menu banner ("Alexa ask Kodi right", "left" ...) but I did not find any word in the skill to enter these items. Even when I'm in my movies list I can not start reading. Could we create a simple word like "OK", "Enter", "Go" or "going inside" that would simulate the click of the mouse or the key entry of the keyboard?

Answer :
The View* Intents do this. For example:

affiche les films
cherche des films d' {MovieGenre}
cherche des films de {MovieGenre}
montre des films d' {MovieGenre}
montre des films de {MovieGenre}
montre les films
ouvre les films
recherche des films d' {MovieGenre}
recherche des films de {MovieGenre}
va aux films


3) Most of the time I do not know what movie I'm going to watch, so I go to the "Movies" item of my menu banner and I have the list of all my movies that is displayed, but I have about 900 movies. When I use my remote for Kodi I can scroll through the list very quickly. But can we do it with the skill? I have not found. Maybe with the words "scroll down/up the list" and "stop" to stop scrolling. Is it possible ?

Answer :
The PageUp and PageDown Intents handle this. For example:

suivante
montre plus


4) I have the impression that it is difficult for Alexa to recognize the name of an english movie correctly pronounced in english when the skill is installed in french. I tried for example "A history of violence" and I did not manage to launch it. But no problem for simple titles like "Taken" or french movies titles. Is there any learning to be done to Alexa to improve the experience?

Answer :
Unfortunately, what Amazon passes to the skill is what we have to work with. German users had similar issues to this in the beginning, but it got better with time.


Thank you for the elements already given and here are my new questions that follow from the answers:

1) Is there not a way to leave Alexa activated for kodi the time we want it and pronounced a sentence to close the interaction with Alexa? I know this exists in the Jarvis project (https://www.openjarvis.com/plugins/kodi). Does Amazon systematically impose a timeout or can we override and set no timeout?

2) In fact when I ask Alexa to show the movies it opens my movie list in Kodi's file manager but not on my movie library. This is certainly because I use playlists for each item of my menu. So I will try calling my playlists. Otherwise I saw that there was the word "select" which allow access to the element on which one is during navigation. I tried and it works.

3) Yes I know "Page up" and "page down". Would it be possible to make voice commands dependent on another voice command? For example:
Page down
-> again / next
Page up
-> again / next
Like that instead of repeating many times "page down" or "page up" to navigate in a list we could just say "page down" once and then shorter words like "next" or "again" (in French the sentences are often longer than in English ).

4) How did the Germans manage to improve Alexa's understanding of invoking the names of the media to play?

Sorry that this message is so long.

Regards

Micky

[jingai]

1) Is there not a way to leave Alexa activated for kodi the time we want it and pronounced a sentence to close the interaction with Alexa? I know this exists in the Jarvis project (https://www.openjarvis.com/plugins/kodi). Does Amazon systematically impose a timeout or can we override and set no timeout?

Amazon sets the timeout.

2) In fact when I ask Alexa to show the movies it opens my movie list in Kodi's file manager but not on my movie library. This is certainly because I use playlists for each item of my menu. So I will try calling my playlists. Otherwise I saw that there was the word "select" which allow access to the element on which one is during navigation. I tried and it works.

It opens the Movie library for me.

3) Yes I know "Page up" and "page down". Would it be possible to make voice commands dependent on another voice command? For example:
Page down
-> again / next
Page up
-> again / next
Like that instead of repeating many times "page down" or "page up" to navigate in a list we could just say "page down" once and then shorter words like "next" or "again" (in French the sentences are often longer than in English ).

It's not really the intention of the skill to use it this way all the time. Mostly, we intend for you to just ask for a specific media title, search for one, or for her to recommend something for you. For browsing, there really will never be a replacement for physical buttons.

4) How did the Germans manage to improve Alexa's understanding of invoking the names of the media to play?

It got better on Amazon's side with time. Nothing that we have done. The skill is dependent on what Amazon passes to it. We can assist you if you have particular titles that don't seem to be coming through. We would need to know what Alexa actually heard (in the History on your Alexa app) and what the Kanzi Card says.