Contents
- About
- Installation
- Skill Deployment
- Skill Setup
- Additional Validation of Requests
- Extra Settings for More Functionality
- Controlling More Than One Instance of Kodi
- Caching Responses
- Developer Discussion
This is a skill for Amazon Alexa that allows you to control one or more instances of Kodi with your voice.
The process of setting up the skill may seem daunting at first, but the reward -- we feel -- is well worth the effort. If you carefully follow the directions to the tee, you might find it is not as complicated as it seems.
Unfortunately, as of this moment, we cannot simply ship this skill normally as other skills on Amazon's skill marketplace. The main technical hurdle is that some features we would need are currently only supported in the US region. Beyond that, there is the consideration of cost for hosting the skill and the associated database backend. Do try to keep in mind that this is a hobby project for the developers -- we do not get paid in any way.
However, we have made every effort to here to provide clear and concise documentation to allow you to make use of this skill now.
Most everything you can do with a remote or keyboard is supported in the skill, and more:
- Basic navigation (Up/Down, Left/Right, Page Up/Down, Select, Back, Menu, Zoom, Rotate, Move)
- Open library views (Movies, Shows, Music, Artists, Albums, Music Videos, Playlists)
- Open library views by genre (Movies, Shows, Music, Music Videos)
- Open recently added playlists (Movies, Episodes, Albums, Music Videos)
- Playback control (Play/Pause, Skip, Previous, Stop, Step/Jump)
- Adjust volume
- Shuffle all music by an artist
- Play/Shuffle specific album
- Play/Shuffle the latest album by an artist
- Play a specific song
- Play/Shuffle audio and video playlists
- "Party mode" for music (shuffle all)
- Play/Shuffle music videos
- Play/Shuffle music videos from a specific genre
- Play/Shuffle music videos by a specific artist
- Shuffle all episodes of a TV show
- Play random TV show
- Play random TV show from a specific genre
- Play random episode of a specific TV show
- Play specific episode of a TV show ('Play season 4 episode 10 of The Office')
- Play random movie
- Play random movie from a specific genre
- Play specific movie
- Play trailer for a movie in your library (requires plugin.video.youtube addon)
- Play random music video
- Play random music video from a specific genre
- Continue watching next episode of last show that was watched
- Play next episode of a show
- Play newest episode of a show
- Recommend media to watch/listen to
- List/Play recently added media
- List available albums by an artist
- Clean/Update video and audio sources
- "What's playing?" functionality for music, movies, and shows
- Report time remaining on current media and when it will end
- Cycle through audio and subtitle streams
- Search for something in your library (requires script.globalsearch addon)
- Execute addons
- Shutdown/reboot/sleep/hibernate system
- Toggle fullscreen
- Eject media
Instead of providing the exact verbiage here for each command, we strive to make the experience as natural as we can. Simply try asking for what you want in a way that feels right to you. If a particular phrase doesn't work and you think it should, see Getting Help to notify us and we will see what we can do to accommodate the phrase you prefer.
If you need help getting a server going or configuring the Skill, please visit the support thread on the Kodi forum.
If you run into an actual issue with the code, please open an Issue here on Github; however, most issues you might run into will be a result of the complexity of the installation, so we urge you to first search the support thread for your issue. If you cannot find a resolution for your issue with a search, post there and someone will help you determine if your problem lies within the skill code or your particular configuration.
Upgrading the skill from a previous version includes a lot of the initial setup described here, with a few exceptions and a few additional considerations. Therefore, it is very important that you consult the additional documentation in UPGRADING.md before you attempt to update the skill.
Unless you are going to host the skill on Heroku, there are a few things you will need to install before you can get started:
- Python 2.7.x (Python 3 is not supported)
- Git
Before a command from Alexa can be sent to Kodi, you need to enable the following settings found under Settings -> System -> Services -> Web server in Kodi:
- Allow remote control via HTTP
- Allow remote control from applications on this system
- Allow remote control from applications on other systems
Note that wording might change depending on the version of Kodi you have installed. This example is for Kodi 17:
Next, supply a Username and Password that you don't use anywhere else. You can share these credentials with all of your Kodi installations, but you should not use the same credentials you might use elsewhere on the web.
Repeat this for every installation of Kodi you have that you wish to control via the skill. All of this information -- including the port number -- can be the same for every installation of Kodi if you wish. If you are unsure, set up every instance of Kodi the same way to avoid confusion.
Unless you have a very specific reason to do so, there is no reason to change the port number from the default. Your router will control access from the outside world.
Make note of the port, username, and password you are using for later steps.
If you have more than one instance of Kodi you would like to control with this skill, make note of the private IP address for each machine as well. Private IP addresses are typically addresses like 192.168.1.9 or 192.168.0.23
If you are using MySQL as a database backend for Kodi, please note that there are known issues with the optimizer in MySQL 5.7.6+ that will cause any commands that involve queueing items in bulk to be tremendously slow. There is nothing in the skill we can do to fix this, as it is technically a MySQL bug/limitation; however, you can either stay on MySQL < 5.7.6 or you can easily migrate to another that doesn't have this problem, such as MariaDB.
As far as we are aware, SQLite (the default) and Emby do not share this issue.
This skill is hosted in the "cloud," unless you opt to host it yourself locally. As such, it will need to know your internet address to contact Kodi.
While you can use your public IP address, with most residential internet providers this address can change over time, which would necessitate setting up the skill again.
To avoid this, before you continue, we suggest you set up what's called a Dynamic Domain Name with a service provider such as Dynu. HowToGeek.com provides a guide for setting this up.
Whether you choose to use your public IP address or obtain a dynamic domain name, make note of it before continuing.
Unless you are hosting the skill locally, it is required that you have the Kodi web server(s) opened up to the internet via port forwarding on your router.
For each Kodi instance, you will need a port forwarding rule on your router. For each private IP address (i.e., the local address of each machine on which Kodi is installed), you need to forward a unique external port to that private IP address on your router.
If you followed our suggestion and set up all of your Kodi instances with the default port, you would forward them all like so:
- Kodi1: external port 8080, internal -> 192.168.1.10:8080
- Kodi2: external port 8081, internal -> 192.168.1.11:8080
- Kodi3: external port 8082, internal -> 192.168.1.12:8080
And so on for each instance of Kodi that you wish to control via the skill. The key here is that the external port number needs to be unique for each instance.
For more information on port forwarding, see this HowToGeek guide.
When you ask an Amazon Alexa skill to do anything, it ultimately contacts another web server which executes code to process your request. This skill isn't any different, and needs such a server as well.
You have a few choices, ordered by difficulty:
Do note that while Heroku is the easiest to setup, it does lack some features, such as support for multiple installations of Kodi. If that is unimportant to you, we definitely recommend that you choose that deployment method.
Heroku is a great way to get a server running for free, but there is a small limitation with the free tier on Heroku where the 'dyno' will go to sleep after 30 minutes of in-activity. This might cause some commands to timeout, but so far it seems to be the best option for getting up and running as quickly as possibly. To get around this, you can either pay for a "Hobby" server which is only $7/month. If you really don't want to pay, there is a work-a-round where you get enough free hours per month to leave the server running 24/7 if you add your Credit Card to your account. Then you can use something like Kaffeine to keep it from spinning down.
Currently we do not support addressing multiple instances of Kodi with Heroku. If you wish to control multiple instances of Kodi, you will need to set up multiple copies of the skill to do so.
After you have set up an Heroku account, click on this button below to provision a new server. Select a unique name to make upgrades easy.
Make note of the endpoint
Now skip ahead to Skill Setup.
Coming soon.
Lambda is a great service which lets the skill run "serverless." AWS provides credits for new accounts and should allow you to run everything the skill needs for free for 12 months. Once you are being billed for it, it will be less than $0.20/month. Very reasonable for what it offers.
Getting going on Lambda is pretty straightforward. First, you'll need to create an Amazon developer account if you don't have one already. After that, browse to the IAM Management Console where you will create a new user:
Next, run these commands to configure your computer for AWS service access:
pip install awscli and then aws configure. Just follow the prompts, and copy paste the keys when it asks for them. When it asks for location, if you are in the US, enter: us-east-1, and if you are in Europe: eu-west-1.
After you've done that, run pip install virtualenv. This is required for a later step.
Now, clone my repo: git clone https://github.com/m0ngr31/kodi-alexa.git and cd kodi-alexa. Once you are inside the project directory, you're going to create a new "Virtual environement" and then activate it:
virtualenv venv and source venv/bin/activate (if you are on Windows, that's venv\Scripts\activate.bat or venv\Scripts\activate.ps1 for Powershell).
Next you need to create the file kodi.config from the kodi.config.example template and enter the correct information for: address, port, username, and password. I'll go over the other variables in another section below.
After you've created your config file, run pip install -r requirements.txt and pip install packaging zappa lambda-packages.
Before you can send any code to Lambda, you'll need to set up Zappa. Just run zappa init and accept the defaults for everything. If it doesn't automatically detect that this is a Flask app, tell it that the application function is alexa.app.
To make an initial deployment to Lambda, just run the following command: zappa deploy dev. It'll take a few minutes, and at the end it will give you a URL that you will need to copy. It will look like this:
You are now running on Lambda! To update after there is a change here, or you updated your env variables, see the instructions in UPGRADING.md.
Now skip ahead to Skill Setup.
If you are savvy enough to run your own web server, this is just a bog-standard WSGI app, and you're free to host it on any web server, provided it supports HTTPS. You will have to generate your own SSL certificate, be it self-signed or via a Certificate Authority (CA), such as Let's Encrypt. This is an Amazon requirement.
You will need to create the file kodi.config from the kodi.config.example template. The template file contains comments to describe the options.
Install the modules in requirements.txt.
Now skip ahead to Skill Setup.
Once you've set up your server, you'll need to configure an Interaction Model for your skill.
If you don't yet have an Amazon Developer account, you'll need to create one by visiting here and logging in with your usual Amazon credentials. Fill in the required fields, agree to the terms, and answer, "No," when asked if you're going to monetize your apps.
Then, head over to the Skills list on Amazon's developer page and hit the Add new skill button.
For Skill Type, choose Custom Interaction Model.
For Language, choose your native language.
For Name, enter Kodi.
For Invocation Name, you can choose whatever you like. kodi works well for us, but it can be whatever you'd like for the phrase:
alexa, tell invocation_name to ...
Leave everything else at their default values.
This is the model that Alexa will use to pass information to the skill that we've deployed earlier.
The Model consists of Slots, an Intent Schema, and Sample Utterances.
You need to first create the Slots that contain the library items in Kodi, broken up into categories for the skill. It is not necessary to keep these up-to-date after the initial setup of the Interaction Model.
To establish initial Slot values, you can use this webapp and point it at the instance of Kodi that has the largest number of items in its library.
You can also get the information from running python generate_custom_slots.py in the project directory, assuming you have a local copy of the code and proper configuration in kodi.config. This will communicate with the Kodi instance defined in the [DEFAULT] section in kodi.config and generate a series of files from which you can create your Slots.
For each Slot, you need to click Add Slot Type, name it according to the output from one of the above generators, and fill the values with their contents.
Ultimately, you will end up adding the following Slots:
- MOVIES
- MOVIEGENRES
- SHOWS
- SHOWGENRES
- MUSICVIDEOS
- MUSICVIDEOGENRES
- MUSICARTISTS
- MUSICALBUMS
- MUSICSONGS
- MUSICGENRES
- MUSICPLAYLISTS
- VIDEOPLAYLISTS
- ADDONS
If one of your slots is empty, you can just enter the word 'Empty' or something so that it'll save.
Next, you'll have to paste the contents of IntentSchema.json into the Intent Schema field.
Then, paste the contents of SampleUtterances.en.txt or SampleUtterances.de.txt, depending on your native language, in the Sample Utterances field.
Enter your endpoint here. It will always be HTTPS and you should choose a region that is geographically close to you:
If you are using Heroku or Lambda, select the middle option:
If you are self-hosting, select one of the outer options depending on how you've chosen to generate your certificate.
Setup is complete at this point.
From here, you can use the simulator or any other Alexa-enabled device to test requests. Note that the simulator has a few nuances and isn't identical to an actual device, so if a request fails in the simulator, please try it on a real Alexa-enabled device.
If you can't manage to get valid responses at this point, see the Getting Help section.
To verify that incoming requests are only allowed from your own copy of the skill, you can set the skill_id configuration variable to your own Application ID; e.g., amzn1.ask.skill.deadbeef-4e4f-ad61-fe42-aee7d2de083d
Setting the timezone configuration variable will make it so when you ask how long something has left playing, it'll also tell you when it will end according to your local wall-clock time.
Setting scheme to https allows you to talk to Kodi securely, but this requires that you set up a reverse HTTPS proxy.
By default, the skill allows very generic queries such as, play 99 red balloons or shuffle the office. These very generic commands can be slow however, and may cause timeouts. If these timeouts bother you, you can direct the skill to provide help playing media more specifically instead when it encounters these kinds of requests, by disabling deep_search.
As of version 2.6 of the skill, it can now control more than one instance of Kodi. The skill determines which instance to talk to by determining which Echo device received the command.
You set up the mapping in the kodi.config file. There are a few examples there with dummy device IDs.
If a device ID isn't explicitly present in the config file, it will utilize the details in the [DEFAULT] section. So, for example, if you wanted most of your devices to send commands to Kodi in your living room, you would set the [DEFAULT] section to point at that instance. For any that you want to override -- say, office and master bedroom -- you would define override sections with those device IDs.
Further, for override sections, if a variable isn't defined, it will inherit it from the [DEFAULT] section. Thus, if the only thing you need to change is address and port, define just those in the override. You do not need to copy all of the other variables as well.
When you send a request to the skill, it will log an entry on the skill's server that will look something like this:
Sending request to http://mydomain.com:8080/jsonrpc from device amzn1.ask.device.AEFDXCGLSFJFNGCVF8SDJF90FID9G284JDJFGJGHH83358DJFFGIGD734JJDFGK211GDFFHHH23HGFJTYEFGJRT56KJDHDFJ5546DJDFFSWOPPP677P88P873EHZNZDFEIRTYIN2239NDFGIH724JDFKS2AA
For AWS Lambda/Zappa deployments, you can access your logs with:
zappa tail dev
To generate your override sections, you will want to tail (watch the end of) the log file and send any request (like, Alexa, ask Kodi what is playing?) from the Echo device you wish to override. Look for a line that looks like the above. The device ID is everything from amzn1.ask.device. to the end of the line. Copy this text and paste it to the end of the kodi.config file, placing it within square brackets [], like so:
[amzn1.ask.device.AEFDXCGLSFJFNGCVF8SDJF90FID9G284JDJFGJGHH83358DJFFGIGD734JJDFGK211GDFFHHH23HGFJTYEFGJRT56KJDHDFJ5546DJDFFSWOPPP677P88P873EHZNZDFEIRTYIN2239NDFGIH724JDFKS2AA]
Anything in square brackets denotes a new section. In this section, you can override whatever variables you'd like. In this example, this Echo device is my Echo Dot in the office, so I would do something like:
# Office Echo Dot
[amzn1.ask.device.AEFDXCGLSFJFNGCVF8SDJF90FID9G284JDJFGJGHH83358DJFFGIGD734JJDFGK211GDFFHHH23HGFJTYEFGJRT56KJDHDFJ5546DJDFFSWOPPP677P88P873EHZNZDFEIRTYIN2239NDFGIH724JDFKS2AA]
address = office-dot
Every time you make a request to find and/or play a media item, the skill will ask Kodi for all of the media items it has in its library of that media type. For example, if you have a song library of 30,000 items and you ask for a particular song, the skill must retrieve the 30,000 songs from your library to match what you asked for with what you have.
As you can imagine, sending such a large response over the internet to your skill can take some time. How long this takes is a function of your available internet upload bandwidth vs how large your library is.
For those with particularly large libraries or very slow uplinks, we provide the option to cache these responses somewhere closer to the skill.
When utilizing caching, you will need to make a voice request to clear out the old cache objects every time you make changes to your Kodi library. 'Changes' include adding a new item, removing an item, or editing an item. In practice, most people simply add and remove items in their libraries, which requires instructing Kodi to notice the changes anyway. So, if you get into the habit of using voice requests for this:
Alexa, ask Kodi to clean library
Alexa, ask Kodi to update library
everything will sort itself out. If you forget to use the voice commands after a library change, the skill will not pick up your changes/additions. In this case, just re-do the library update/clean via voice.
The configuration for the cache buckets is defined in the device sections of kodi.config. This allows you to use more than one bucket if you wish to do so. It is not necessary though even if you have more than one Kodi library: each cache object is named such that they won't collide. However, the clean library and update library commands will clear out all objects in the bucket -- even those that don't pertain to the library you're updating or cleaning. Therefore, it is up to you to decide if it's worth the extra configuration effort to use more than one bucket.
If you only have one library or you are unsure, it will be simplest to just use one bucket by defining all of the cache configuration in the [DEFAULT] section of kodi.config.
Regardless of the location where you store the cached responses, bear in mind that you may incur charges to do so.
We allow storage of cache objects currently in either an Amazon S3 bucket or on an ownCloud/nextCloud server.
For most people, an Amazon S3 bucket will be the right choice here. But please note that Amazon may charge you for this service. It is very inexpensive, however, and it's highly likely that you'll fall under the Free Usage Tier if you're only using it for this skill.
For Heroku users and those self-hosting the skill, the backend you choose will mostly depend on which ends up being faster for you and whether or not your usage falls under the free tier for Amazon S3. If cost is not a consideration, feel free to try all backends and see what works best for you.
If you are self-hosting the skill local to your Kodi installation(s), you may wonder if this has any benefit for you. It will, if you already have or want to set up an ownCloud/nextCloud server local to your Kodi installation(s). Since the requests the skill makes to Kodi involve a database query as well as the hit to its relatively-slow-web-server, retrieving the responses from a cache can shave off a little time. How much really depends on your hardware and database configuration; the difference will probably be small, but if you've read this far, such an optimization might be worth the effort for you.
For all cache backends, you should use credentials that only provide access to the cache bucket and nothing else. The primary concern here is that the credentials are stored in kodi.config, which is stored in your skill deployment. If you can't secure the skill deployment, you should create a user that only has access to the cache.
A secondary consideration is that this can put your library items in public view. The cache objects themselves will not directly divulge personally identifying information, but your account credentials can tie the data to you if someone gained access to them.
The first thing you need to do is sign up for Amazon S3 if you haven't already. Next, create a unique bucket in which the skill can store cache objects. It is recommended that you use a randomized name for the bucket if possible.
For the bucket you create, you definitely should not use your root credentials, and ideally, you shouldn't use admin credentials either. If you deployed the skill via AWS Lambda, 'admin credentials' would be the the keys for the kodi-alexa user described there.
It would be best to create another user with only the ability to list buckets and administer the cache bucket via the IAM Management Console.
However, if you don't use S3 for anything else that you consider important or private, you're welcome to use the kodi-alexa user credentials for the rest of this.
For each bucket you have created, you will need to add the following to either the [DEFAULT] section or the appropriate device section:
cache_bucket = your_bucket_name
s3_cache_aws_access_key_id = your_access_key_id
s3_cache_aws_secret_access_key = your_secret_key
For example:
[DEFAULT]
cache_bucket = kodi-alexa-7ce4fea901
s3_cache_aws_access_key_id = AKQJ56NSL3XGGLQYDT0L
s3_cache_aws_secret_access_key = iCmj7ag+mdxkcDLcst/gAMDXrjkp0j7toMMDrrkJ
# Office Echo Dot
[amzn1.ask.device.AEFDXCGLSFJFNGCVF8SDJF90FID9G284JDJFGJGHH83358DJFFGIGD734JJDFGK211GDFFHHH23HGFJTYEFGJRT56KJDHDFJ5546DJDFFSWOPPP677P88P873EHZNZDFEIRTYIN2239NDFGIH724JDFKS2AA]
cache_bucket = kodi-alexa-4fae019a1f
You will need to provide credentials for a user that can create the bucket (folder/directory) that you specify at the root of each URL you provide.
For each bucket, you will need to add the following to either the [DEFAULT] section or the appropriate device section:
cache_bucket = your_bucket_name
owncloud_cache_url = https://my.owncloudserver.com/owncloud/
owncloud_cache_user = username
owncloud_cache_password = password
For example:
[DEFAULT]
cache_bucket = kodi-alexa-7ce4fea901
owncloud_cache_url = https://my.owncloudserver.com/owncloud/
owncloud_cache_user = username
owncloud_cache_password = password
# Office Echo Dot
[amzn1.ask.device.AEFDXCGLSFJFNGCVF8SDJF90FID9G284JDJFGJGHH83358DJFFGIGD734JJDFGK211GDFFHHH23HGFJTYEFGJRT56KJDHDFJ5546DJDFFSWOPPP677P88P873EHZNZDFEIRTYIN2239NDFGIH724JDFKS2AA]
cache_bucket = kodi-alexa-4fae019a1f
If you're interested in chatting with us about the development of the skill, we are on Slack.