Skill Package API Reference

You can use the Skill Package API to interact with your skill as a single package, rather than by component.

Skill package

A skill package consists of at least one component:

  • Manifest – The skill manifest contains metadata about the skill and configuration, including the display name, skill capabilities (video, smart home, custom), and endpoint for the skill service code (whether AWS Lambda or a separate web endpoint). The name of the skill manifest file is skill.json. Every skill must have a manifest.

The following components of a skill package are optional:

  • Image assets – These images are associated with your skill, appearing as icons in the Alexa Skills Store, or your in-skill products.
  • In-skill products – These define the in-skill products that users can purchase in your skill. They specify the type of purchase and other details such as pricing.
  • Interaction models – These define the voice user interface for your skill which includes the invocation name, intents, slots, and so forth. Only custom skills include interaction models, with one model for each locale that the skill supports. The name of the interaction model is <locale>.json, for example en-US.json. In the skill package format, the interaction models must be in a folder named custom, which itself must be in a folder named interactionModels.

Your skill might also contain account linking information, which is currently not supported by the Skill Package API. You can still use the Skill Package API, but you will also have to use SMAPI to manage account linking.

If you interact with your skill only through the Alexa Skills Kit developer console, these components still exist behind the scenes. If you interact with your skill through the Alexa Skills Kit Command Line Interface (ASK CLI) or the Skill Management API (SMAPI), you see these components as individual files. A skill package combines these components into a single file so that you can work with them as a unit.

Skill package format

A skill package file combines all of the skill components into a zipped file with a .zip extension. The skill package components are organized in the following way:

SkillPackageName.zip
├── assets
│   └── images
│       ├── en-US_largeIcon.png
│       ├── en-US_smallIcon.png
│       ├── isp1
│       │   ├── en-US_largeIcon.png
│       │   └── en-US_smallIcon.png
│       └── isp2
│           ├── en-US_largeIcon.png
│           └── en-US_smallIcon.png
├── interactionModels
│   └── custom
│       └── en-US.json
├── isps
│   ├── isps.json
│   ├── isp1.json
│   └── exampleDirectory
│       └── isp2.json
└── skill.json

To create and import a skill using the skill package format:

  1. Organize your skill components as shown in the preceding diagram, then zip the top-level folder.
  2. Create an upload URL.
  3. Use the URL that you created in the previous step to upload the .zip file to Amazon S3.
  4. After you upload the skill package, either to S3 or your own URL, you can:

You can also update an existing skill by exporting it in the skill package format, editing it locally, then using the preceding steps to upload and import it.

Manage in-skill products in a skill package

You can manage in-skill products (ISPs) by including them in your skill package. When you include ISPs in a skill package, you must include a folder named isps which contains a file named isps.json, as shown in the preceding section. This file contains information about ISPs to create or update, and ISPs that are associated (linked) with the skill, as shown in the following example.

{
  "isps": {
    "Unique ISP Name 1": {
      "path": "file://isps/isp1.json"
    },
    "Unique ISP Name 2": {
      "path": "file://isps/exampleDirectory/isp2.json"
    }
  },
  "associations": [
    "amzn1.adg.UUID3",
    "amzn1.adg.UUID4"
  ]
}

The following list describes each object in the isps.json file:

isps
Contains one or more ISPs, each with a unique name that you create. Each ISP contains a path that specifies the location of the ISP schema file relative to the skill.json file (you can consider skill.json the package root). You can organize the ISP files in whatever directory structure you want, using each ISP's path value to specify the organization.
associations
Contains a list of product IDs, which are unique identifiers for an ISP. Each product ID in the list represents an ISP that is associated (linked) with your skill. To get the product ID for an ISP, use the In-Skill Product Management API or the developer console.

When you include ISPs in a skill package that you import into an existing skill, the existence of isps and associations in isps.json determines what the Skill Package API does:

  • When isps.json contains isps or associations that are new, the Skill Package API creates or associates the ISPs accordingly.
  • When isps.json contains updates to isps that previously existed in the skill package, the Skill Package API updates them accordingly.
  • When isps.json omits associations that previously existed in the skill package, the Skill Package API disassociates them accordingly. The Skill Package API does not delete ISPs.

If you don't want the Skill Package API to take these actions, you can omit the isps folder from the skill package, and continue to manage ISPs separately.

For more information about ISPs, see In-Skill Purchasing Overview and In-Skill Purchasing Schemas.

Use eTag to avoid overwriting a skill package

When you use the Skill Package API to import a skill package to an existing skill, the imported skill package overwrites the existing skill's components. In some cases, you might want to avoid overwriting an existing skill. For example, consider the following scenario in which a graphic designer and a developer work on the same skill simultaneously:

  1. The graphic designer and the developer each use the Skill Package API to export separate copies of the skill to their own laptops.
  2. The graphic designer adds new images to the skill package, then uses the Skill Package API to upload and import the skill package to the existing skill.
  3. The developer updates the skill's interaction model and adds a new ISP, then uses the Skill Package API to upload and import the skill package to the existing skill. This import overwrites the changes that the graphic designer made, because the developer did not have these changes in his or her local copy of the skill package.

To avoid this scenario, you can use the skill package's eTag. You can think of an eTag as an opaque version identifier for a skill package. You get an eTag in the body of the response when you:

You can include an eTag value in the If-Match HTTP request header when you import a package for an existing skill.

Consider the preceding scenario with the use of eTags:

  1. The graphic designer and the developer each use the Skill Package API to export separate copies of the skill to their own laptops. They each save the eTag value in addition to the skill package .zip file.
  2. The graphic designer adds new images to the skill package, then uses the Skill Package API to upload and import the skill package to the existing skill. The graphic designer's import request includes the If-Match HTTP header with the eTag value, which matches the server-side eTag value for the skill package. The import succeeds, which results in a new eTag value for this latest version of the skill package.
  3. The developer updates the skill's interaction model and adds a new ISP, then uses the Skill Package API to upload and import the skill package to the existing skill. The developer's import request includes the If-Match HTTP header with the eTag value, which does not match the server-side eTag value for the skill package, so the import request fails.
  4. The developer exports the skill package to obtain the latest version and the corresponding eTag value. This version includes the previous updates that the graphic designer made.
  5. The developer compares his or her local changes to the latest version, and reconciles the differences.
  6. The developer uses the Skill Package API to upload and import the skill package to the existing skill. The developer's import request includes the If-Match HTTP header with the eTag value, which now matches the server-side eTag value for the skill package, so the import request succeeds.

API endpoint and authorization

The API endpoint is https://api.amazonalexa.com. Each API request must have an Authorization header whose value is the access token retrieved from Login with Amazon.

Create a new upload URL

Use this API to create a presigned upload URL that you can use to upload a skill package to Amazon S3. To upload a skill package to S3, you can:

  • Send an HTTP PUT request to the presigned URL that you receive in the response to this API.
  • Use an AWS SDK. For more information, see Uploading Objects Using Presigned URLs in the Amazon S3 documentation.

You can also bypass this API entirely by uploading your skill package to a publicly-accessible URL on the internet.

After you upload the skill package, either to S3 or your own URL, you can:

Request

POST /v1/skills/uploads

Response

201 Created

The response body includes the uploadUrl value and the expiry date, as shown in the following sample.

{
  "uploadUrl": "string",
  "expiresAt": "2018-10-11T19:28:34.525Z"
}

Errors

Code Description
401 The authorization token is invalid, expired, or doesn't have access to the resource.
429 Exceeds the permitted request limit.
500 Internal Server Error.
503 Service Unavailable.

Create a new skill package

Use this API to submit a request to create a new skill package.

Request

POST /v1/skills/imports

The request body takes the following form:

{
  "vendorId": "string",
  "location": "string"
}

Response

202 Accepted

The value of the Location header is a relative URL that should be used to track the status of the import request.

Errors

Code Description
400 Server cannot process the request due to a client error.
401 The authorization token is invalid, expired, or doesn't have access to the resource.
413 Payload Too Large.
429 Exceeds the permitted request limit.
500 Internal Server Error.
503 Service Unavailable.

Import a package for an existing skill

This API submits a request to import a zipped package file for an existing skill specified by skillId.

Request

POST /v1/skills/{skillId}/imports

To avoid overwriting a previous update to the skill, you can optionally include the If-Match HTTP header with an eTag value. For more information, see Use eTag to avoid overwriting a skill package.

The request body takes the following form, with the location value being the presigned URL that you have previously obtained, or a public URL from which the zipped skill package can be downloaded. For more information, see Create upload URL.

{
  "location": "string"
}

Response

202 Accepted

The value of the Location header is the URL that you specified in the request.

Errors

Code Description
400 Server cannot process the request due to a client error.
401 The authorization token is invalid, expired, or doesn't have access to the resource.
409 Conflict.
413 Payload Too Large.
429 Exceeds the permitted request limit.
500 Internal Server Error.
503 Service Unavailable.

Get the status for a specified importId

Obtains the importStatus for the specified importId.

Request

GET /v1/skills/imports/{importId}

Response

200 OK

The status is one of:

  • FAILED
  • IN_PROGRESS
  • SUCCEEDED

The skill.resources object's status is one of:

  • FAILED
  • IN_PROGRESS
  • ROLLBACK_FAILED
  • ROLLBACK_IN_PROGRESS
  • ROLLBACK_SUCCEEDED
  • SKIPPED
  • SUCCEEDED

The skill.resources object's action is one of:

  • ASSOCIATE
  • CREATE
  • UPDATE
  • DISASSOCIATE

Response body

{
  "status": "enum",
  "errors": [
    {
      "code": "string",
      "message": "string"
    }
  ],
  "warnings": [
    {
      "code": "string",
      "message": "string"
    }
  ],
  "skill": {
    "skillId": "string",
    "eTag": "string",
    "resources": [
      {
        "name": "string",
        "status": "enum",
        "action": "enum",
        "errors": [
          {
            "code": "string",
            "message": "string"
          }
        ],
        "info": [
          {
            "message": "string"
          }
        ],        
        "warnings": [
          {
            "code": "string",
            "message": "string"
          }
        ]
      }
    ]
  }
}

Errors

Code Description
401 The authorization token is invalid, expired, or doesn't have access to the resource.
404 The resource being requested is not found.
429 Exceeds the permitted request limit.
500 Internal Server Error.
503 Service Unavailable.

Create an export request for an existing skill

Creates an export request for an existing skill. You can then get the status of the export request, including a URL to download the exported skill.

Request

The stage value must be either live (for a live, published skill) or development (for a skill in the development stage).

POST /v1/skills/{skillId}/stages/{stage}/exports

If your skill has been hidden or removed, after previously being published, the stage value is still live. For more information, see Hide or remove a skill. You can re-submit a hidden or removed skill for re-certification.

Response

202 Accepted

The response has no body. To get the status of the export request, including a URL to download the exported skill, send a GET request to the path in the response's Location header. For more information, see Get the status for an export request.

Errors

Code Description
401 The authorization token is invalid, expired, or doesn't have access to the resource.
404 The resource being requested is not found.
409 The request could not be completed due to a conflict with the current state of the target resource.
429 Exceeds the permitted request limit.
500 Internal Server Error.
503 Service Unavailable.

Get the status for an export request

Gets the status of an export request that you created previously. When the status of the export request is SUCCEEDED, the response contains a URL to download a zip archive of the exported skill.

Request

Send a GET request to the path contained in the Location header of the response to a previous export request. The exportId is provided in this path, so you don't need to do anything extra to obtain it.

GET /v1/skills/exports/{exportId}

Response

200 OK

The body of the response contains one of these statuses: FAILED, IN_PROGRESS, or SUCCEEDED. When the status is SUCCEEDED, the skill.location value contains a URL that you can use to download a zip archive of the exported skill.

Example response body

{
  "skill": {
    "expiresAt": "1550521786056",
    "location": "URL to download the exported skill",
    "eTag": "string"
  },
  "status": "SUCCEEDED"
}

Errors

Code Description
401 The authorization token is invalid, expired, or doesn't have access to the resource.
404 The resource being requested is not found.
429 Exceeds the permitted request limit.
500 Internal Server Error.
503 Service Unavailable.
Amazon Alexa
Amazon Appstore
Amazon Web Services
Amazon Dash Replenishment