In the response do you still need to include an empty slots subnode?

		"invocationResponse": {
			"body": {
				"version": "1.0",
				"response": {
					"canFulfillIntent": {
						"canFulfill": "YES",
						"slots": {}
					}
				}
			}
		},

or is something like this okay?

		"invocationResponse": {
			"body": {
				"version": "1.0",
				"response": {
					"canFulfillIntent": {
						"canFulfill": "YES"
					}
				}
			}
		},

I have a private Alexa for business skill that I wrote. Can you perhaps confirm or deny that this intent is invoked on private skills? I put print statements in my code (Im emulating this all in Python). When I try invoke the skill without using the skill name just by using utterances, I would expect this function to be invoked underneath. The print statements log to cloud watch, so I was expecting to see data there.

However, I am not seeing data in cloud watch and it doesn't seem to be working at all. Meaning none of my utterances are matching up with intents and getting passed to the skill without the invocation name.

Can someone provide a sample perhaps in python so I can check my work, and\or advise on what the json response should look like when we dont need to worry about slots?

The documentation states For each slot, a skill is optionally expected to return values for the canUnderstand and canFulfill fields. Optionally to me means that you simple can omit the data and just return an empty slots subnode but I am not sure.

def canFulfillIntent(fullfill_request,session):
    print("We hit the fullfill intent")
    lsta = []
    intent = fullfill_request['intent']
    intent_name = str(fullfill_request['intent']['name'])
    print("Intent Name Is: " + intent_name)
    if intent_name in lsta:
        #we are good return yes
        Decision = "YES"
        print("FOUND INTENT")
        return build_response_fulfill(build_fulfill_response(Decision))
    else:
        #return no    
        Decision = "NO"
        print("DID NOT FIND INTENT")
        return build_response_fulfill(build_fulfill_response(Decision))

Hi,

I can't test this sample with vocal console in Amazon. When i say "open voice experts and tell me about Jedidiah Esposito", it's work fine but if i just say "tell me about Jedidiah Esposito", alexa answer me "Sorry, I'm not sure."
How can i do?

Thanks

When using "ask-sdk-core" I kept getting the "handlerInput.responseBuilder.withCanFulfillIntent is not a function" error. When changed to "ask-sdk" I was able to get the following (expected) response:

"response": {
"canFulfillIntent": {
"canFulfill": "YES",
"slots": {
"voiceExpert": {
"canUnderstand": "YES",
"canFulfill": "NO"
}
}
}
}

Here is my "npm ls --depth=0" output:
8c85900feb20:custom $ npm ls --depth=0
[email protected] ~/canFulfillPaul/lambda/custom
├── [email protected]
├── [email protected]
└── [email protected]

Hi,

Your code seems to rely on the entity resolution information for the slots which is not implemented by Amazon according to the response to this question on the forum.

https://forums.developer.amazon.com/questions/172553/is-entity-resolution-json-present-in-a-canfulfilli.html

So is the info present or not? Does anybody have a real request or we just keep copy-pasting fantasy JSON into the simulator?