Create a StopIntent for Google Assistant with Dialogflow

by Kaan Kilic on Sep 10, 2018

By default, Google Assistant immediately exits the app when a user says things like "exit", "cancel", or "stop". If you want to return a last response, you need to override this default behavior. Read on to learn how it works.


Jovo comes with a default 'END' handler that lets you clean up some data and send a final response after a person wants to exit your voice app. For Alexa Skills, the AMAZON.StopIntent and SessionEnded requests are mapped to END by default. Learn more here: Routing > 'END' Intent.

Google Assistant, however, doesn't offer this behavior by default, but rather exits the conversation by just playing an earcon. Read more here: Actions on Google Documentation > Conversation Exits.

Create a StopIntent or CancelIntent

Alexa already comes with a required AMAZON.StopIntent that is triggered when a user wants to stop the conversation. For Google Assistant, we need to add a new intent with a specific event actions_intent_CANCEL to our Dialogflow Agent. Any incoming request with this event will then be mapped to the 'END' intent in your Jovo app.

You can call it whatever you want, for example StopIntent or CancelIntent (notice that there is no difference on Google Assistant between "stop" and "cancel" in comparison to AMAZON.StopIntent or AMAZON.CancelIntent).

There are two alternatives to add the intent:

Add it manually in Dialogflow

You can add it by going to the Dialogflow Console and creating a new intent. Select "Events" and type in actions_intent_CANCEL. No need to add any training phrases.

Add actions_intent_CANCEL to Dialogflow

Don't forget to scroll all the way down and enable webhook fulfillment:

Enable webhook fulfillment for the intent

You can now save the intent and train the agent.

Add it to the Jovo Language Model

If you want to maintain your language model in your code base (recommended), you can choose different strategies.

For example, you can add the intent to the dialogflow specific block in your language model file:

Alternatively, you can add an intent to the general Jovo intents with specific blocks for both alexa and dialogflow:

Learn more about how to deploy your Dialogflow Agent here: Deploy a Dialogflow Agent with the Jovo CLI.

Kaan Kilic

Technical Content Marketing Associate at Jovo

Comments and Questions

Any specific questions? Just drop them below or join the Jovo Community Forum.

Join Our Newsletter

Be the first to get our free tutorials, courses, and other resources for voice app developers.