https://unsplash.com/photos/LJ9KY8pIH3E
Alexa skills
Williams, Mark
Thu 13/01/2022 13:01
Brandao-Neto, Jose
1: Make an Amazon developer account at https://developer.amazon.com/alexa/console/ask
2: Click Create Skill.
3: Give your skill a name, eg Liam's Jokes. Select the language of your Alexa device.
4: Choose Custom model, and Alexa Hosted (Python), and click Create Skill.
5: Choose Start From Scratch, then Continue.
6:
When its finished, click Invocation Name, and change it to Liam's Jokes
if it is not already that, click Save Model, then Build Model.
7: Go to Code at the top.
8: Delete all the stuff that is already there, its way too complicated.
The whole thing is to take json in, and spit json out. The json comes into a function called like this:
def lambda_handler(event, context):
Dont worry about event and context yet. To get a joke, do something like:
import requests
r = requests.get("https://icanhazdadjoke.com/", headers={"Accept": "application/json"})
joke = r.json()['joke']The function needs to return something like:
return {
'version': '1.0',
'sessionAttributes': {},
'response': {
'outputSpeech': {
'type': 'PlainText',
'text': joke
},
'shouldEndSession': True
}
}Then you should be able to put joke as the
9: When you have finished the python, click Save then Deploy.
10:
Because we have used the requests package, we need to make sure it is
available. Go to the requirements.txt file, and replace the contents
with
requests==2.27.1
11: Click Save, then Deploy
12: Click Test at the top. Change the dropdown so it says "Development"13: Type "Alexa, launch Liam's Jokes" in the box
Debugging
is very hard, took me an hour to figure out I needed to do step 10, and
I've done it before. But hopefully that gets you started.
Mark
