A Node module to simplify development of Alexa apps (Skills) using Node.js
Generation of sample utterances from the schema definition in alexa-app has been pulled out to the separate alexa-utterances module.
npm install alexa-app --saveThe alexa-app-server module is a fully-working container for multiple alexa-app skills using Node.js and Express. It lets you run and debug your apps locally, and can also be used as a full production server for your apps.
alexa-app does the dirty work of interpretting the JSON request from Amazon and building the JSON response. It provides convenience methods to more easily build the response, handle session objects, and add cards. It also makes it easy to run multiple endpoints (apps) on one Node.js server instance.
After defining your application's behavior, your caller (Express, Lambda, etc) should call the app.request() method, and pass the Alexa JSON request object. It will return a promise containing the response JSON. Your caller should then insert that into its response, whatever form it may take (see examples below).
The intent schema definition and sample utterances can be included in your application's definition, making it very simple to generate hundreds (or thousands!) of sample utterances with a few lines.
NOTE: alexa-app makes no assumptions about what context it is running in. It will run in a stand-alone Node.js app, within an HTTPS server, within an AWS Lambda function, etc. It only cares about JSON in and JSON out. It is agnostic about the environment that is using it, but it provides some convenience methods to hook into common environments.
NOTE: If you don't use AWS Lambda and decide to host an Alexa skill on your own webserver, you will need to validate that requests come from Alexa. This validation is not provided by this module. For more details on how to handle alexa request validation, look at https://github.com/mreinstein/alexa-verifier which provides the necessary code, along with an example showing how to integrate with express.
- Simplified handling of requests and generating responses
- Asynchronous handlers supported
- Easy connection into AWS Lambda or Node.js express (and can be connected to any other server)
- Auto-generation of intent schema and sample utterances
- Easier handling of session data
- Apps can be tested without running a server
See the example directory for sample implementations.
var alexa = require('alexa-app');
var app = new alexa.app('sample');
app.intent('number',
{
"slots":{"number":"NUMBER"}
,"utterances":[ "say the number {1-100|number}" ]
},
function(request,response) {
var number = request.slot('number');
response.say("You asked for the number "+number);
}
);
// Manually hook the handler function into express
express.post('/sample',function(req,res) {
app.request(req.body) // connect express to alexa-app
.then(function(response) { // alexa-app returns a promise with the response
res.json(response); // stream it to express' output
});
});Apps ("skills") define handlers for launch, intent, and session end, just like normal Alexa development. The alexa-app module provides a layer around this functionality that simplifies the interaction. Each handler gets passed a request and response object, which are custom objects for this module.
// Return the type of request received (LaunchRequest, IntentRequest, SessionEndedRequest)
String request.type()
// Return the value passed in for a given slot name
String request.slot('slotName')
// Return the value of a session variable
String request.session('attributeName')
// Session details, as passed by Amazon in the request
request.sessionDetails = { ... }
// The raw request JSON object
request.dataThe response JSON object is automatically built for you. All you need to do is tell it what you want to output.
// Tell Alexa to say something. Multiple calls to say() will be appended to each other.
// All text output is treated as SSML
response.say(String phrase)
// Empty the response text
response.clear()
// Tell Alexa to re-prompt the user for a response, if it didn't hear anything valid
response.reprompt(String phrase)
// Return a card to the user's Echo app
// For Object definition @see https://developer.amazon.com/public/solutions/alexa/alexa-skills-kit/docs/alexa-skills-kit-interface-reference#card-object
// Skill supports card(String title, String content) for backwards compat of type "Simple"
response.card(Object card)
// Return a card instructing the user how to link their account to the skill.
// This internally sets the card response.
response.linkAccount()
// Tell Alexa whether the user's session is over. By default, sessions end.
// You can optionally pass a reprompt message
response.shouldEndSession(boolean end [, String reprompt] )
// Set a session variable
// By defailt, Alexa only persists session variables to the next request. The alexa-app module
// makes session variables persist across multiple requests.
response.session(String attributeName, String attributeValue)
// Send the response as success
// You don't usually need to call this. This is only required if your handler is
// asynchronous - for example, if it makes an http request and needs to wait for
// the response, then send it back to Alexa when finished.
response.send()
// Trigger a response failure
// The internal promise containing the response will be rejected, and should be handled by
// the calling environment. Instead of the Alexa response being returned, the failure
// message will be passed.
response.fail(String message)
// Calls to response can be chained together
response.say("OK").send();Your app can define a single handler for the Launch event, the SessionEnded event, the Messaging.MessageReceived event and multiple Intent handlers.
app.launch(function(request,response) {
response.say("Hello World");
response.card("Hello World","This is an example card");
});This special type of intent is called when the skill is invoked outside the context of a session.
In this scenario, a payload is sent to the skill, accessible via request.data.
Use the request.data object to create notifications for users that consent to receiving messages/push notifications.
app.messageRecevied(function(reuqest,response) {
var user = request.data.context.System.user;
if (user.permissions.consentToken) {
// Presence of `user.permissions.consentToken` indicates the user
// has allowed messaging/push notifications via the companion app.
// Proceed in creating a notification using the contents of `request.data`.
} else {
// Since this is outside the context of a user session,
// we can't use `response.say`.
console.log('The user: "' + user.userId + '" has not allowed notifications via the companion app.');
}
});Define the handler for multiple intents using multiple calls to intent(). Intent schema and sample utterances can also be passed to intent(), which is detailed below. Intent handlers that don't return an immediate response (because they do some asynchronous operation) must return false. See example further below.
app.intent('buy', function(request,response) {
response.say("You bought a "+request.slot("item"));
});
app.intent('sell', function(request,response) {
response.say("You sold your items!");
});app.sessionEnded(function(request,response) {
// Clean up the user's server-side stuff, if necessary
logout( request.userId );
// No response necessary
});In addition to specific event handlers, you can define functions that will run on every request.
Executed before any event handlers. This is useful to setup new sessions, validate the applicationId, or do any other kind of validations.
app.pre = function(request,response,type) {
if (request.sessionDetails.application.applicationId!="amzn1.echo-sdk-ams.app.000000-d0ed-0000-ad00-000000d00ebe") {
// Fail ungracefully
response.fail("Invalid applicationId");
}
};Note: The post() method still gets called, even if the pre() function calls send() or fail(). The post method can always override anything done before it.
The last thing executed for every request. It is even called if there is an exception or if a response has already been sent. The post() function can change anything about the response. It can even turn a response.fail() into a respond.send() with entirely new content. If post() is called after an exception is thrown, the exception itself will be the last argument.
app.post = function(request,response,type,exception) {
// Always turn an exception into a successful response
response.clear().say("An error occured: "+exception).send();
};alexa-app makes it easy to define your intent schema and generate many sample utterances. Optionally pass your schema definition along with your intent handler, and extract the generated content using the schema() and utterances() functions on your app.
Pass an object with two properties: slots and utterances.
app.intent('sampleIntent',
{
"slots":{"NAME":"LITERAL","AGE":"NUMBER"},
"utterances":[ "my {name is|name's} {names|NAME} and {I am|I'm} {1-100|AGE}{ years old|}" ]
},
function(request,response) { ... }
);The slots object is a simple Name:Type mapping. The type must be one of Amazon's supported slot types: LITERAL, NUMBER, DATE, TIME, DURATION
[Custom slot types](https://developer.amazon.com/public/solutions/alexa/alexa-skills-kit/docs/alexa-skills-kit-interaction-model-reference#Custom Slot Type Syntax) are supported via the following syntax:
app.intent('sampleIntent',
{
"slots":{"CustomSlotName": "CustomSlotType" },
"utterances":[ "airport {information|status} for {-|CustomSlotName}" ]
},
function(request,response) { ... }
);This will result in an utterance list of the following:
sampleIntent airport information for {CustomSlotName}
sampleIntent airport status for {CustomSlotName}
Note that the "CustomSlotType" type values must be specified in the Skill Interface's Interaction Model for the custom slot type to function correctly.
The utterances syntax allows you to generate many (hundreds or even thousands) of sample utterances using just a few samples that get auto-expanded. Any number of sample utterances may be passed in the utterances array. Below are some sample utterances macros and what they will be expanded to.
"my favorite color is {red|green|blue|NAME}"
=>
"my favorite color is {red|NAME}"
"my favorite color is {green|NAME}"
"my favorite color is {blue|NAME}"This lets you define multiple ways to say a phrase, but combined into a single sample utterance
"{what is the|what's the|check the} status"
=>
"what is the status"
"what's the status"
"check the status"When capturing a numeric slot value, it's helpful to generate many sample utterances with different number values
"buy {2-5|NUMBER} items"
=>
"buy {two|NUMBER} items"
"buy {three|NUMBER} items"
"buy {four|NUMBER} items"
"buy {five|NUMBER} items"Number ranges can also increment in steps
"buy {5-20 by 5|NUMBER} items"
=>
"buy {five|NUMBER} items"
"buy {ten|NUMBER} items"
"buy {fifteen|NUMBER} items"
"buy {twenty|NUMBER} items""what is your {favorite |}color"
=>
"what is your color"
"what is your favorite color"Several intents may use the same list of possible values, so you want to define them in one place, not in each intent schema. Use the app's dictionary.
app.dictionary = {"colors":["red","green","blue"]};
...
"my favorite color is {colors|FAVEORITE_COLOR}"
"I like {colors|COLOR}"To get the generated content out of your app, call the schema() and utterances() functions. See examples/express.js for one way to output this data.
// Returns a String representation of the JSON object
app.schema() =>
{
"intents": [
{
"intent": "MyColorIsIntent",
"slots": [
{
"name": "Color",
"type": "LITERAL"
}
]
}
]
}
app.utterances() =>
MyColorIsIntent my color is {dark brown|Color}
MyColorIsIntent my color is {green|Color}
MyColorIsIntent my favorite color is {red|Color}
MyColorIsIntent my favorite color is {navy blue|Color}
WhatsMyColorIntent whats my color
WhatsMyColorIntent what is my color
WhatsMyColorIntent say my color
WhatsMyColorIntent tell me my color
WhatsMyColorIntent whats my favorite color
WhatsMyColorIntent what is my favorite color
WhatsMyColorIntent say my favorite color
WhatsMyColorIntent tell me my favorite color
WhatsMyColorIntent tell me what my favorite color isThe response.card(Object card) method allows you to send "Home Cards" on the Alexa app, the companion app available for Fire OS, Android, iOS, and desktop web browsers.
The full specification for the card object passed to this method can be found here
Card's do not support SSML
Note: If you just want to display a card that presents the user to link their account call response.linkAccount() as a shortcut.
Display text only, aka Simple:
response.card({
type: "Simple",
title: "My Cool Card", //this is not required for type Simple
content: "This is the\ncontent of my card"
});Display text and image, aka Standard:
Note: make sure to read the restrictions on hosting the images. Must support CORS AND SSL cert signed by an Amazon approved cert authority.
response.card({
type: "Standard",
title: "My Cool Card", //this is not required for type Simple OR Standard
text: "Your ride is on the way to 123 Main Street!\nEstimated cost for this ride: $25",
image: { //image is optional
smallImageUrl: "https://carfu.com/resources/card-images/race-car-small.png", //One must be specified
largeImageUrl: "https://carfu.com/resources/card-images/race-car-large.png"
}
});Handler functions should not throw exceptions. Ideally, you should catch errors in your handlers using try/catch and respond with an appropriate output to the user. If exceptions do leak out of handlers, they will be thrown by default. Any exceptions can be handled by a generic error handler which you can define for your app. Error handlers cannot be asynchronous.
app.error = function(exception, request, response) {
response.say("Sorry, something bad happened");
};If you do want exceptions to bubble out to the caller (and potentially cause Express to crash, for example), you can throw the exception.
app.error = function(exception, request, response) {
console.log(exception);
throw exception;
};If an intent handler will return a response later, it must return false. This tells the alexa-app library not to send the response automatically. In this case, the handler function must manually call response.send() to finish the response.
app.intent('checkStatus', function(request,response) {
http.get("http://server.com/status.html", function(res) {
// This is async and will run after the http call returns
response.say(res.statusText);
// Must call send to end the original request
response.send();
});
// Return false immediately so alexa-app doesn't send the response
return false;
});Amazon has documentation on how to setup your Alexa app to run in AWS Lambda.
Apps built using alexa-app have a built-in "handler" method to handle calls from AWS Lambda. You don't need to do anything different to make them work within Lambda, other than to setup the Lambda Function correctly and make sure that the Handler is set to "index.handler", which is the default value.
For backwards compatibility, or if you wish to change the Handler mapping to something other than index.handler, you can use the lambda() function. See example/lambda.js.
var app = new alexa.app('sample');
app.intent( ... );
// Connect the alexa-app to AWS Lambda
exports.handler = app.lambda();var express = require('express');
var alexa = require('alexa-app');
var bodyParser = require('body-parser');
var express_app = express();
var app = new alexa.app('sample');
app.launch(function(request,response) {
response.say("Hello World");
});
// This call defines a post() and optionally a get() handler in express, mapped to the alexa-app
// express_app: The express app instance to map to
// path: The path prefix to map to
// enableDebug: If false, don't mapy a GET handler. Default: true
// Debugging GET requests call express' render() method using 'test'
app.express( express_app, "/echo/", false );
// Now POST calls to /echo/sample in express will be handled by the app.request() function.
// GET calls will not be handledapp.messages.NO_INTENT_FOUND = "Why you called dat intent? I don't know bout dat";See the code for default messages you can override.
app.launch(function(request,response) {
response.session('number',42);
response.say("Would you like to know the number?");
response.shouldEndSession(false);
});
app.intent('tellme', function(request,response) {
response.say("The number is "+request.session('number'));
// Clear only the 'number' attribute from the session
response.clearSesssion('number');
});
// The session variables can be entirely cleared, or cleared by key
app.intent('clear', function(request,response) {
response.clearSession(); // or: response.clearSession('key') to clear a single value
response.say("Session cleared!");
});By default, alexa-app will persist every request session attribute into the response. This way, any session attributes you set will be sent on every subsequent request, as is typical in most web programming environments. If you wish to disable this feature, you can do so.
var app = new alexa.app('test');
app.persistentSession = false;When mapped to express, the default endpoint for each app is the name of the app. You can customize this using the second parameter to the app() method.
var app = new alexa.app('hello','myEndpointName');All named apps can be found in the alexa.apps object, keyed by name. The value is the app itself.
See CHANGELOG for details.
Copyright (c) 2016 Matt Kruse
MIT License, see LICENSE for details.