Connect a bot to Cortana
Cortana is a speech-enabled channel that can send and receive voice messages in addition to textual conversation. A bot intended to connect to Cortana should be designed for speech as well as text. A Cortana skill is a bot that can be invoked using a Cortana client. Publishing a bot adds the bot to the list of available skills.
To add the Cortana channel, open the bot in the Azure Portal, click the Channels blade, and then click Cortana.
General bot information
All fields marked with an asterisk (*) are required. Bots must be published on the Bot Framework before they can be connected to Cortana.
Currently, general bot information is only available as READONLY on the Azure portal. These fields can now be managed from the Store Workspace - see documentation for details
Select a custom icon to represent this bot. This icon is displayed in the Cortana canvas when this skill is invoked and anywhere skills are discoverable, such as the Microsoft store. The image must be a PNG, 60 x 60 pixels, and no larger than 30kb.
The name of this skill as displayed to the user at the top of Cortana's visual UI.
The name that Cortana will recognize and use to invoke this skill when spoken aloud by the user. See the Invocation Name Guidelines for more information on how to choose this phrase.
Description of the skill. This is used where skills are discoverable, like the Microsoft Store.
This summary will be used to describe the skill in Cortana’s Notebook.
Register a bot to the Knowledge Store
Scroll down to Discovery and Management
After you click Register, you will be redirected to the Store portal.
Click Yes to allow the Store to access your account information
Upon acceptance, you'll have the option of going through the First Time User Experience tutorial. At this point, your bot will be connected to the cortana channel.
The Store workspace allows you to create and publish new skills for Cortana, explore existing skills created by other developers, access free samples and tutorials to build botlets, skills and more. See the Store Overview for more details.
Enable speech recognition priming
If your bot uses a Language Understanding (LUIS) app, register the LUIS application ID.
Click the Settings tab and then under Configuration, enter the LUIS application ID in the Speech recognition priming with LUIS text box. This helps your bot recognize spoken utterances that are defined in your LUIS model.
For more information on how to configure speech priming, click here.