Tutorial: Add Your First Visual Response to a Custom Skill

This tutorial shows you how to use Alexa Presentation Language (APL) to add a visual response to a custom skill. The response displays a "Hello World" message by using a responsive template for the layout and a data source for the content.

Prerequisites

You must have an Amazon developer account and a working custom skill. The tutorial provides code examples that build on the Node.js "Hello World" sample skill available on GitHub: skill-sample-nodejs-hello-world. Follow the directions provided in the git repository to set up the skill before you begin this tutorial.

Make sure that you can log into the developer console and use the test simulator to invoke your skill. Verify that you can invoke one of your intents and get a response without errors. If you use the Hello World sample, make sure you can invoke the HelloWorldIntent.

Steps to add a visual response to your custom skill

APL works within the request and response interface for a custom skill. When your skill receives a request, such as a LaunchRequest or IntentRequest, you can provide an APL document and data source as part of your response. Alexa displays your response on the screen.

In this tutorial, you create a new APL document and data source, and then update an intent handler in the skill code to include the APL content in the response. The tutorial walks through the following steps:

1. Configure the skill to support the APL interface.
2. Use the APL authoring tool to create and save an APL document to display.
3. Create a data source to contain the content to display within the document, and then write data-binding expressions to bind properties of the document to the data source.
4. Update the skill code to send the document and data source to Alexa as part of a response.
5. Test the skill in the developer console to see the new visual response.

Step 1: Enable the APL interface in the developer console

To send a response with APL content, your skill must support the Alexa.Presentation.APL interface. Enable this option in the developer console.

To enable the APL interface in the developer console

1. Open the developer console, and click Edit for the skill you want to configure.
2. Navigate to the Build > Interfaces page.

3. Enable the Alexa Presentation Language option.

   This action makes it possible for your skill to send Alexa APL content to display.

4. In the list of viewport profiles, select all the profiles.

   Explain what selecting all the profiles means

   Selecting a viewport profile in this list means that your APL response supports that profile and therefore works on devices that use that profile.

   Clearing a viewport profile in the list means that your skill doesn't support that profile. Users who use your skill on an unsupported device see scaled version of your visual response. This might not be a good customer experience.

   Whenever possible, adapt your APL responses to work across all the different viewport profiles and then select to support all the profiles in the APL interface.

   This tutorial uses a responsive template that works across devices. Therefore, you can select all the profiles.

   For more about supporting different viewport profiles, see Select the Viewport Profiles Your Skill Supports.

5. Click Save Interfaces and then Build Model to re-build your interaction model.

Step 2: Create a new APL document in the authoring tool

An APL document defines a template for your visual response. In the following sections, you create a new APL document in the authoring tool and update the document to display a line of text.

To create a new APL document

1. In the developer console for your skill, make sure you're still on the Build page.
2. In the sidebar, click Multimodal Responses. This action opens the APL authoring tool in a new window or browser tab.
3. Select Visual.
4. Click Create Visual Response. This action opens a page with several templates you can use.
5. Click Blank Document to create a new APL document with no content.

6. Click the Save icon. When prompted, enter the name "HelloWorldDocument".

   As you complete the tutorial steps, click Save frequently to save your work. The authoring tool doesn't save automatically.

An APL package defines a set of layouts, resources, and styles that you can import into your APL documents. The alexa-layouts package provided by Amazon contains a set of responsive components and templates. These layouts automatically work on viewports with different modes, sizes, and shapes. When you decide to create a visual response, review the available pre-built templates and components first.

To use a package in your document, you must first import it.

To import the alexa-layouts package

1. In the authoring tool, click the Code View tab to display the JSON for the APL document.

   Show the initial APL document code

   The starter code for the document looks like the following.

    {
        "type": "APL",
        "version": "2024.3",
        "license": "Copyright 2021 Amazon.com, Inc. or its affiliates. All Rights Reserved.\nSPDX-License-Identifier: LicenseRef-.amazon.com.-AmznSL-1.0\nLicensed under the Amazon Software License  http://aws.amazon.com/asl/",
        "settings": {},
        "theme": "dark",
        "import": [],
        "resources": [],
        "styles": {},
        "onMount": [],
        "graphics": {},
        "commands": {},
        "layouts": {},
        "mainTemplate": {
            "parameters": [
                "payload"
            ],
            "items": []
        }
    }
   

2. In the import array, add the following object.

    {
      "name": "alexa-layouts",
      "version": "1.7.0"
    }
   

   Note: For packages provided by Amazon, you specify the name and version. For your own packages or packages shared by other developers, you specify the name, version, and source, where source is a URL for the package.

   What are the APL packages provided by Amazon?

   Amazon provides three packages you can use in your APL documents:

   • alexa-layouts – Contains a set of responsive components and templates.
   • alexa-styles – Contains a set of styles and resources.
   • alexa-viewport-profiles &ndash Defines resources for a set of viewport profiles you can use to define conditional logic in your APL document.

   The alexa-layouts package also imports both alexa-styles and alexa-viewport-profiles.

   For more about these packages and a reference to the resources inside them, see Alexa Design System for APL.

After making the change, your JSON should look like the following example:

{
  "type": "APL",
  "version": "2024.3",
  "license": "Copyright 2021 Amazon.com, Inc. or its affiliates. All Rights Reserved.\nSPDX-License-Identifier: LicenseRef-.amazon.com.-AmznSL-1.0\nLicensed under the Amazon Software License  http://aws.amazon.com/asl/",
  "settings": {},
  "theme": "dark",
  "import": [
    {
      "name": "alexa-layouts",
      "version": "1.7.0"
    }
  ],
  "resources": [],
  "styles": {},
  "onMount": [],
  "graphics": {},
  "commands": {},
  "layouts": {},
  "mainTemplate": {
    "parameters": [
      "payload"
    ],
    "items": []
  }
}

In an APL document, the mainTemplate property specifies the layout to display when the document initially displays on the screen. The mainTemplate.items property contains an array of items to display. In the following steps, you add the AlexaHeadline responsive template to the items array.

To add the JSON for AlexaHeadline

1. Make sure that the authoring tool is still in Code View.

2. In the mainTemplate.items array, add the following block of code:

    {
      "type": "AlexaHeadline",
      "primaryText": "Display this text"
    }
   

   The preview pane updates to show "Display this text" centered in the viewport.

   What are the 'type' and 'primaryText' properties?

   The type property for the object identifies the name of the item to display. The items array can contain APL components, layouts defined within your document, or layouts defined in an imported package. Each one of these items always has a type that you specify when placing the item in your document.

   In this example, AlexaHeadline is the name of a layout defined in the alexa-layouts package.

   The primaryText property is a parameter defined on the AlexaHeadline template. The template has several other properties. For more about this template, see Alexa Headline.

After making the change, your JSON should look like the following example:

{
  "type": "APL",
  "version": "1.8",
  "license": "Copyright 2021 Amazon.com, Inc. or its affiliates. All Rights Reserved.\nSPDX-License-Identifier: LicenseRef-.amazon.com.-AmznSL-1.0\nLicensed under the Amazon Software License  http://aws.amazon.com/asl/",
  "settings": {},
  "theme": "dark",
  "import": [
    {
      "name": "alexa-layouts",
      "version": "1.4.0"
    }
  ],
  "resources": [],
  "styles": {},
  "onMount": [],
  "graphics": {},
  "commands": {},
  "layouts": {},
  "mainTemplate": {
    "parameters": [
      "payload"
    ],
    "items": [
      {
        "type": "AlexaHeadline",
        "primaryText": "Display this text"
      }
    ]
  }
}

Step 3: Connect the document to a data source

Although the document you've created so far would work in a skill, it's better to separate the presentation from the content to display. This approach is important when the data to display within the template changes based on some condition in the skill. For example, imagine a skill that asks the user for their name, and then later displays that name on the screen. Each time the device renders the document, the data to display might be different.

The correct way to pass data from your skill to the document is to place the content in a data source and then use a data-binding expression to connect that content to your APL document. From your code, you send an identical document in each response, but with an updated data source.

To add a data source to the authoring tool

1. In the authoring tool, click the Data button to open the data sources pane.

   The data sources pane lets you simulate passing a data source to your document. When you add the document to your code later, you send the data source as part of the directive.

2. Replace the contents of this pane with the following code:

    {
      "helloWorldDataSource": {
        "primaryText": "Hello World!",
        "secondaryText": "Welcome to Alexa Presentation Language!",
        "color": "@colorTeal800"
      }
    }
   

   The preview pane doesn't change because you haven't yet bound any template properties to the data source properties.

When building your own APL documents, you design the structure of the data source.

To add the data source to your document and write data-binding expressions

1. In the authoring tool, click the APL button to return to the document pane. Make sure you're still in Code View.

2. In the parameters array, remove the existing payload item and replace it with "helloWorldDataSource".

    {
      "mainTemplate": {
        "parameters": [
          "helloWorldDataSource"
        ],
        "items": [
          {
            "type": "AlexaHeadline",
            "primaryText": "Display this text"
          }
        ]
      }
    }
   

   Explain what the parameters array is for

   The mainTemplate.parameters array contains the data sources you want to connect to your document. For each item in the array, Alexa identifies a data source with the same name, and then adds that data source to the data-binding context. The data-binding context is an object that contains the collection of data available to an APL document for data binding. You access the data-binding context with data-binding expressions in the form ${...}.

   Therefore, if you add helloWorldDataSource to mainTemplate.parameters, the data-binding context contains the property helloWorldDataSource that refers to the corresponding data source. You can then write expressions to access the data source properties, such as ${helloWorldDataSource.primaryText}.

   If you leave mainTemplate.parameters set to "payload", you can still bind to the data source, but you must use a longer expression, such as ${payload.helloWorldDataSource.primaryText}.

3. For the primaryText property, change "Display this text" to the following expression:

    ${helloWorldDataSource.primaryText}
   

   The text shown on the preview pane changes to "Hello World!"

   What does this expression mean?

   A data-binding expression is an expression that Alexa dynamically evaluates when displaying your document. You use these expressions to bind a property in your document to a value in a data source.

   Data-binding expressions begin with a dollar sign ($) and then enclose the expression with curly brackets ({...}). For details about the syntax, see Data-Binding Syntax.

   In an earlier step, you added "helloWorldDataSource" to the mainTemplate.parameters property. This action makes the helloWorldDataSource data source object available to your document in the data-binding context. The data-binding context is an object that contains the collection of data available to an APL document for data binding.

   Therefore, you can access the primaryText property for the object in the data source with the expression: ${helloWorldDataSource.primaryText}.

   For more technical details about how APL evaluates data-binding expressions, see APL Data-Binding Evaluation.

4. Add the secondaryText and backgroundColor properties, binding each one to the corresponding data in the data source:

    {
      "type": "AlexaHeadline",
      "primaryText": "${helloWorldDataSource.primaryText}",
      "secondaryText": "${helloWorldDataSource.secondaryText}",
      "backgroundColor": "${helloWorldDataSource.color}"
    }
   

   The preview pane updates to show the two blocks of text and a teal background.

See the full finished document  

{
  "type": "APL",
  "version": "1.8",
  "license": "Copyright 2021 Amazon.com, Inc. or its affiliates. All Rights Reserved.\nSPDX-License-Identifier: LicenseRef-.amazon.com.-AmznSL-1.0\nLicensed under the Amazon Software License  http://aws.amazon.com/asl/",
  "settings": {},
  "theme": "dark",
  "import": [
    {
      "name": "alexa-layouts",
      "version": "1.4.0"
    }
  ],
  "resources": [],
  "styles": {},
  "onMount": [],
  "graphics": {},
  "commands": {},
  "layouts": {},
  "mainTemplate": {
    "parameters": [
      "helloWorldDataSource"
    ],
    "items": [
      {
        "type": "AlexaHeadline",
        "primaryText": "${helloWorldDataSource.primaryText}",
        "secondaryText": "${helloWorldDataSource.secondaryText}",
        "backgroundColor": "${helloWorldDataSource.color}"
      }
    ]
  }
}

Step 4: Add code to your skill to render the response

Now that you have a working APL document, you can add code to your skill to include the document in a response. You send Alexa the Alexa.Presentation.APL.RenderDocument directive in your response. You pass a link to the document and the full data source JSON to the RenderDocument directive in the directive properties.

To add the RenderDocument directive to the skill response

1. Open your skill code in an IDE. For an Alexa-hosted skill, navigate to the Code tab in the developer console.

2. Find the intent handler to update. For the "Hello World" sample skill, find the HelloWorldIntentHandler.

   See the HelloWorldIntentHandler (Node.js)

    const HelloWorldIntentHandler = {
        canHandle(handlerInput) {
            return Alexa.getRequestType(handlerInput.requestEnvelope) === 'IntentRequest'
                && Alexa.getIntentName(handlerInput.requestEnvelope) === 'HelloWorldIntent';
        },
        handle(handlerInput) {
            const speakOutput = handlerInput.t('HELLO_MSG');
   
            return handlerInput.responseBuilder
                .speak(speakOutput)
                //.reprompt('add a reprompt if you want to keep the session open for the user to respond')
                .getResponse();
        }
    };
   

3. To add the RenderDocument directive, add the if...else block to the handle function between the first line of the code and the return statement.

    handle(handlerInput) {
        const speakOutput = handlerInput.t('HELLO_MSG');
       
        // ADD THE NEW IF BLOCK HERE
      
        return handlerInput.responseBuilder
            .speak(speakOutput)
            //.reprompt('add a reprompt if you want to keep the session open for the user to respond')
            .getResponse();
    }
   

4. Save your updates to the code.

If…else block to return RenderDocument

if (Alexa.getSupportedInterfaces(handlerInput.requestEnvelope)['Alexa.Presentation.APL']){
    console.log("The user's device supports APL");

    const documentName = "HelloWorldDocument"; // Name of the document saved in the authoring tool
    const token = documentName + "Token";

    // Add the RenderDocument directive to the response
    handlerInput.responseBuilder.addDirective({
        type: 'Alexa.Presentation.APL.RenderDocument',
        token: token,
        document: {
            src: 'doc://alexa/apl/documents/' + documentName,
            type: 'Link'
        },
        datasources: {
            "helloWorldDataSource": {
                "primaryText": "Hello World!",
                "secondaryText": "Welcome to Alexa Presentation Language!",
                "color": "@colorTeal800"
            }
        }
    });
    
} else {
    // Just log the fact that the device doesn't support APL.
    // In a real skill, you might provide different speech to the user.
    console.log("The user's device doesn't support APL. Retest on a device with a screen")
}

Explain how the if…else block works  

The following code example shows the full HelloWorldIntentHandler with the APL code.

See the full HelloWorldIntentHandler  

const HelloWorldIntentHandler = {
    canHandle(handlerInput) {
        return Alexa.getRequestType(handlerInput.requestEnvelope) === 'IntentRequest'
            && Alexa.getIntentName(handlerInput.requestEnvelope) === 'HelloWorldIntent';
    },
    handle(handlerInput) {
        const speakOutput = handlerInput.t('HELLO_MSG');
    
        if (Alexa.getSupportedInterfaces(handlerInput.requestEnvelope)['Alexa.Presentation.APL']){
            console.log("The user's device supports APL");
        
            const documentName = "HelloWorldDocument"; // Name of the document saved in the authoring tool
            const token = documentName + "Token";
        
            // Add the RenderDocument directive to the response
            handlerInput.responseBuilder.addDirective({
                type: 'Alexa.Presentation.APL.RenderDocument',
                token: token,
                document: {
                    src: 'doc://alexa/apl/documents/' + documentName,
                    type: 'Link'
                },
                datasources: {
                    "helloWorldDataSource": {
                        "primaryText": "Hello World!",
                        "secondaryText": "Welcome to Alexa Presentation Language!",
                        "color": "@colorTeal800"
                    }
                }
            });
            
        } else {
            // Just log the fact that the device doesn't support APL.
            // In a real skill, you might provide different speech to the user.
            console.log("The user's device doesn't support APL. Retest on a device with a screen")
        }

        return handlerInput.responseBuilder
            .speak(speakOutput)
            //.reprompt('add a reprompt if you want to keep the session open for the user to respond')
            .getResponse();
    }
};

Before you can test your updates, you must deploy your skill code.

To save and deploy an Alexa-hosted skill

1. Make sure you're on the Code tab.
2. Click Save, and then click Deploy.

The RenderDocument code shown sets the document property to a link to your document saved in the authoring tool. To make APL documents saved in the authoring tool available to the skill back-end code, you must build the interaction model for the skill.

To rebuild the interaction model

1. Return to the developer console and navigate to the Build tab.
2. Navigate to one of the interaction model sections, such as Invocations > Skill Invocation Name and click Build Model.

Step 5: Test your updated skill

You can test your skill with a device or in the skill simulator in the developer console.

To test your skill in the developer console

1. In the developer console, navigate to the Test tab.
2. For the Skill testing is enabled in option, select Development.
3. To display the simulator for a device with a screen, select Device Display.

4. Under Alexa Simulator, invoke your skill, and then enter your test utterances.

   For the Hello World, skill, enter your invocation name, and then enter the utterance "hello".

5. Scroll past the Skill I/O section to see the device simulator. You should see the "Hello World" document you built earlier.

Troubleshooting the hello world skill

If the simulator doesn't display the APL comment, check the following issues.

Issue: Error stating that the URI is not valid

Symptoms

When you invoke the intent, Alexa responds with the error message "The U.R.I. configured for the specified skill is not valid."

This error occurs when the URI you provided in the document.src property for the RenderDocument directive doesn't match a document stored with your skill.

Try this

You must rebuild your interaction model to make documents saved in the authoring tool available to your skill. This error can occur when you have saved the document, but haven't yet rebuilt the model. Rebuild the model and test the skill again.

Try this

The URI you pass to the document.src property must be in the format doc://alexa/apl/documents/{documentName} where {documentName} is the name for the document in the authoring tool. In this tutorial, you should save your document as HelloWorldDocument. Therefore, the correct URI is doc://alexa/apl/documents/HelloWorldDocument.

1. Recheck your code against the examples shown in Add the RenderDocument directive to the skill response. Make sure you spelled all names correctly.
2. Return to the authoring tool and confirm that you saved the document with the correct name.

Issue: The APL content on screen is missing information or the screen is blank

Symptoms

When testing the skill, Alexa doesn't report any errors. However, the content displayed on the screen is missing information. For example, the "Hello World" line displays, but the "Welcome to Alexa Presentation Language!" line is missing. Alternatively, the entire screen is blank.

Try this

This issue occurs if the data-binding expressions in the document don't match the structure of the data source. If the entire screen is blank, your RenderDocument directive might not be configured to pass the full data source along with your document.

To make sure the data-binding expressions match the data source

1. Make sure that your call to addDirective() sets the datasources property to the JSON for the helloWorldDataSource:

    handlerInput.responseBuilder.addDirective({
        type: 'Alexa.Presentation.APL.RenderDocument',
        token: token,
        document: {
            src: 'doc://alexa/apl/documents/' + documentName,
            type: 'Link'
        },
        datasources: {
            "helloWorldDataSource": {
                "primaryText": "Hello World!",
                "secondaryText": "Welcome to Alexa Presentation Language!",
                "color": "@colorTeal800"
            }
        }
    });
   

2. Make sure that all properties in the data source match the properties used in the document. The sample hello world document uses three expressions:

   • ${helloWorldDataSource.primaryText}
   • ${helloWorldDataSource.secondaryText}
   • ${helloWorldDataSource.color}

3. In your document, make sure that the mainTemplate.parameters array contains the string helloWorldDataSource.

Next steps

• Go deeper with a course on APL – Alexa Learning Lab
• Learn more about the AlexaHeadline responsive template – AlexaHeadline
• Explore other responsive templates – Responsive Templates
• Learn more about how APL works and the different elements you build – What Makes Up an APL Visual Response?
• Learn more about using the APL authoring tool – Build Documents in the Developer Console

Last updated: Nov 28, 2023