An extremely simple way of implementing voiceflow in an alexa skill. This skill works using the voiceflow dialog manager api, allowing for advanced ChatGPT conversations for free in alexa. Read carefully and do not change the names of anything unless specified. Also make sure to hit the save and build buttons in the alexa skill after making any significant changes, especially after you finish your skill.
First, you want to make a voiceflow account and create an assistant. Then, you need to set up your assistant to something that looks similar to the photo above. The assistant will output a welcome message (that you can change to whatever you want), then capture all input and send it to an ai which will speak a response out loud and end the assistant. Currently, the assistant does not loop so everytime you want to ask a question you will have to start the skill again. Go into the integrations tab of your voiceflow assistant found on the side bar and copy your api key. Keep it stored, it will be important later, also remember to run your assistant and hit the publish button after you make any changes to the assistant for them to be applied in the alexa skill.
Log into your alexa account (the same amazon account you use for your alexa) and create a skill. Give it a name, language, etc. Select other as a type of experience and scroll down until you find the hosting settings. Select Alexa hosted (Python) and continue creating it. Make sure to change the invocation name of your skill after you create it too, in this case i am going to set it to Alex Git but you can set it to anything you want as long as it is within the amazon guidelines for invocation names.
Create an intent called InfoIntent and give it a slot called CatchAll with the type AMAZON.SearchQuery. You can give it many different utterances (which are ways of activating the response), but all of them most follow the structure of: 'PHRASE' {CatchAll}, by example: Search for {CatchAll}, or a keyword like Example {CatchAll}. Do keep in mind that {CatchAll} is the text that is going to be sent to the ai, so if you say 'Search for Domino's pizza' and you have that as an utterance, the AI will only receive Domino's pizza as a prompt. Delete the default HelloWorldIntent from the skill since we will not need it.
After that, get the latest release from Releases and upload the zip file that you downloaded to the code tab of your skill, as shown in the image below. Select every single folder of the zip file from the upload tab.
After that, replace YOUR_API_KEY in the code with your actual api key that you saved earlier from the voiceflow project. Remember to hit save and then deploy in your project after you finish updating the code.
Now your code should work and you can test it in the test tab of your skill. For some reason, amazon does not let you test on some browsers like opera gx so i recommend using google chrome. Try it out with your browser of choice and see if it works. After you finish building your skill, if you used the same account on your echo device, you should be able to use it just by calling the invocation name that you set on your skill. Try waiting 20-30 minutes after building before testing on your echo device just to make sure it works.