Go to https://developer.amazon.com/alexa/console/ask/test/amzn1.ask.skill.ad1669b4-291c-4daa-9fbb-fa32b8ea3078/development/de_DE/, make a sample request, and look for accessToken
in JSON input.
Copy it and paste it into token.sh
.
Then:
source private/token.sh
To run tests:
ginkgo -r
New code gets pushed via scripts/deploy-code.sh
. That doesn't make the code available in Prod, i.e. it's not customer facing, but it can be tested in the Alexa developer console, given that the endpoint there is pointing to the non-prod endpoint. To publish the deployed code in prod, there is publish-version
. That will create a new version in Lambda and also move the prod alias to that latest version.
Workflow:
- Make changes
- Deploy changes
scripts/deploy-code.sh
- Verify changes work in the console
- When everything looks good, deploy to production:
scripts/publish-version.sh
Workflow:
- Make sure files in
skill-package
are in sync with what's online:mv skill-package skill-package.old ask smapi export-package -s amzn1.ask.skill.ad1669b4-291c-4daa-9fbb-fa32b8ea3078 --stage development git diff
- When trying things with new Lambda service version too, change endpoints to latest lambda version:
scripts/make-endpoint-unqualified.sh
. Deploy? --> yes. - Iterate on changes in model (in the web console) and Lambda service
- Verify changes work in the console
- When everything looks good:
mv skill-package skill-package.old ask smapi export-package -s amzn1.ask.skill.ad1669b4-291c-4daa-9fbb-fa32b8ea3078 --stage development git diff scripts/make-endpoint-prod.sh
- Commit changes.
- Submit for certification and publication in web console.