/foxycli

Primary LanguageJavaScriptMozilla Public License 2.0MPL-2.0

Foxy client Build Status

This is the an agent that handles the voice processing for the "Hey Foxy" application.

Prerequisites


The following are prerequisites for running the Foxy Client:

Install node (if you didn't use nvm)

(If you already installed node via nvm you can skip this step)

Follow the directions from NodeJS to install on your platform.

Install software

First, clone this repo:

$ git clone https://github.com/tamarahills/foxycli.git

Install dependencies

Install portaudio and sox:

$ brew install portaudio sox
$ pip install pyaudio

Download the Snowboy binaries at http://docs.kitt.ai/snowboy/#downloads

Install the node packages

$ cd foxycli
$ npm install

Download the "Hey Foxy" model from https://snowboy.kitt.ai/dashboard:

  1. mkdir foxycli/resources
  2. Search on "Hey Foxy"
  3. Record a few samples while you are there (Optional)
  4. Click on the "download icon"
  5. Save "Hey_Foxy.pmdl" to the foxycli/resources directory.

Start the agent

$ cd foxycli
$ node index.js

Start the extension

See instructions at https://github.com/tamarahills/foxyext to install and start the extensions. To verify that Foxy is working, you can use the following commands:

Hey Foxy, what's the weather in Seattle
Hey Foxy, set a timer for ten minutes
Hey Foxy, play some NPR
Hey Foxy, play some (jazz|classical|rock)
Hey Foxy, add that to pocket (requires pocket integration)
Hey Foxy, turn the kitchen light on (requires integration with https://github.com/mozilla-iot/gateway)

Running the tests

$ cd foxycli
$ npm test

Troubleshooting Known Issues