🇪🇸 Español

Allows Amazon Alexa to control Google Chromecast

This skill supports controlling a single Chromecast or multiple Chromecasts in different rooms. Each Alexa device can be set to control a different room. This is done by matching the room name to your Chromecast device's name. E.g. If your Chromecast is named: "Master Bedroom TV", then set the Alexa room to control to "Master Bedroom"

The following will then pause the Chromecast in the Master Bedroom:

Alexa, ask Chromecast to pause

You can control another room by saying something like:

Alexa, ask the Chromecast to pause in the Media Room

To change the room a particular Alexa device controls you can say:

Alexa, ask the Chromecast to set the room

Here are some example voice commands:

Alexa, tell Chromecast to play

Alexa, tell Chromecast to play songs by Macklemore

Alexa, tell Chromecast to play maroon 5 playlist

Alexa, tell Chromecast to play The Matrix trailer

Alexa, tell Chromecast to set the volume to 5

Alexa, tell Chromecast to stop

Or:

Alexa, ask the Chromecast in the Media Room to stop

Alexa, ask the Chromecast to play in the Media Room

Alexa skills run in the cloud, but this skill needs to be on your local network to control the Chromecast. This skill implements a hybrid approach: the command is handled by Alexa on AWS, which sends a notification to your local server.

The Lambda component is in src/lambda, and the local component is in src/local.

Both the Chromecast and the Raspberry Pi (or whatever the local notification handler will run on) MUST be on the same network in order for the Chromecast to be discoverable.

Installation requires a UNIX environment with:

• BASH
• Python 3.7
• Pip

1. Create an Amazon Web Services account
2. Run aws-setup.sh to create a Role, Lambda Function, and SNS Topic. (It will run aws configure, so have an key id and access key ready)

3. Go to ASK Console and choose "Create Skill"
4. Select "Custom" and "Provision your own", then click "Create skill". On the template screen just use the "Hello World Skill" template
5. Click on "Interaction Model" in the left menu, then "JSON Editor"
6. Copy and paste the content from config/interaction_model.json into the editor, then click "Save Model"
7. Click on "Endpoint" in the left menu. Enter the Lambda function ARN by the aws-setup.sh. Click "Save Endpoints"
8. Click on "Invocation" in the left menu. Click on "Build Model"
9. Click on the "Test" tab. Enter

10. Install local dependencies with sudo pip3 install -r ./src/local/requirements.txt
11. Run ./start.sh to start the listener, or ./docker-start.sh to run in an interactive docker session. Or ./docker-start.sh -d to run as a service. The service attempts to AWS SNS using UPNP. If UPNP is disabled in your network you can specify a port. ./start.sh -p 30000 To see other options run ./start.sh -h or ./docker-start.sh -h.

When run you should see something like the following:

2020-07-12 11:10:40,688 - root - INFO - Starting Alexa Chromecast listener...
2020-07-12 11:10:40,688 - local.ChromecastSkill - INFO - Finding Chromecasts...
2020-07-12 11:10:45,696 - pychromecast - INFO - Querying device status
2020-07-12 11:10:45,727 - pychromecast - INFO - Querying device status
2020-07-12 11:10:45,767 - local.ChromecastSkill - INFO - Found Media Room TV
2020-07-12 11:10:45,768 - local.ChromecastSkill - INFO - Found Living Room TV
2020-07-12 11:10:45,769 - local.ChromecastSkill - INFO - 2 Chromecasts found
2020-07-12 11:10:45,809 - botocore.credentials - INFO - Found credentials in environment variables.
2020-07-12 11:10:46,967 - local.SkillSubscriber - INFO - Listening on http://123.123.123.123:30000
2020-07-12 11:10:46,968 - local.SkillSubscriber - INFO - Subscribing for Alexa commands...
2020-07-12 11:10:47,344 - local.SkillSubscriber - INFO - Received subscription confirmation...
2020-07-12 11:10:47,431 - local.SkillSubscriber - INFO - Subscribed.

12. Say "Alexa ask Chromecast to play" The skill will take you through any required room setup.

./start.sh

The skill subscriber can be run with docker:

./docker-start.sh - for an interactive session ./docker-start.sh -d - to run as a service

The skill subscriber (local) uses these environment variables:

• AWS_SNS_TOPIC_ARN - AWS SNS Topic ARN (can be found in the .env file after running aws-setup.sh)
• AWS_ACCESS_KEY_ID - AWS User Access Key
• AWS_SECRET_ACCESS_KEY - AWS Secret Access Key
• AWS_DEFAULT_REGION - AWS Lambda and SNS Region (e.g. eu-west-1)

If you have run aws configure, you will not need to set AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, or AWS_DEFAULT_REGION.

Sets up an AWS environment for the Alexa Skill:

1. Creates an IAM role for Alexa (with permissions for SNS)
2. Creates an SNS topic to communicate over
3. Creates an S3 persistent store for persisting the room to Alexa device mapping
4. Creates a Lambda function

Creates a lambda-bundle.zip, which can be uploaded to an AWS Lambda function.

Runs build-lambda-bundle and automatically uploads the bundle to AWS Lambda.

When the local service starts it searches for Chromecasts on the network. If there are no ChromeCasts found, it will exit. To fix this, you must confirm that the Chromecast is on and working, make sure you can access it from your phone, and make sure that everything is on the same network. To debug, a tool to search and list found ChomeCasts is provided at ./search-chromecasts (make sure to make it executable with chmod +x ./search-chromecasts).

If the local listener fails to subscribe (no subscribe messages or an error) then the Chromecasts won't receive commands from Alexa By default the local listener uses UPNP and a dynamic port to establish an external connection, you can override this if required.

1. Check UPNP is enabled/allowed on your network
2. If UPNP is not enabled or working try and manually specify a port, and ensure your firewall/router is configured to allow external access to this port e.g. to use port 30000 run ./start.sh -p 30000 or ./docker-start.sh -p 30000
3. Log into the AWS console and check the SNS topic is setup, and check the Cloud Watch logs for your the lambda function for any errors.

1. Try redeploying the lambda skill. ./aws-update-lambda.sh
2. If that didn't work go to the AWS Console and check the CloudWatch logs associated with the lambda function

1. Check the local listener output, it should show the received command and any error that was encountered
2. To check the docker service logs run something like docker logs alexa_chromecast --since=30m, which shows the logs for the last 30 minutes