Back on the Amazon page, click the Test tab. Let's first test by typing the request, so I write "Alexa, ask MacroFab to blink an LED at 35 hertz". If successful, Alexa should respond with the text from step 10.

Now we have to deal with the fact that Alexa is a computer and not a person. Instead of typing the request, try holding the microphone button and saying your request. You may be surprised to see that Alexa misinterprets your words and cannot respond. What you can do now is copy what Alexa hears into a new Intent and Utterance as explained in Step 4. For me, I wanted to say "Alexa, ask MacroFab to blink an LED at 35 hertz", but she heard "Alexa, ask macro fab to blink and l.e.d. at {rate} hurts" so I changed the Invocation and Intents to match. 

Now Alexa should be able to respond to your voice and you can even try talking to a real Alexa device now.

For this step, I'm going to connect a Particle Photon to Losant so I can physically blink an LED. It's possible to use any device that can connect to the web (like an ESP-32), but I'm using a Particle device because Losant makes it so easy.

1. Create an account at https://www.particle.io
2. Register your device and connect it to WiFi
3. In Particle Build, copy the following code and flash it to your device. It sets up a cloud-exposed function that will blink the onboard LED at the frequency that it is set to.

   // onboard LED of Photon
   int led = D7;
   
   // how long to delay between blinks
   int rate = 0;
   
   int setBlinkRate(String rateString);
   
   void setup() {
       // set pin to output
       pinMode(led, OUTPUT);
     
       // register the cloud function
       Particle.function("setBlinkRate", setBlinkRate);
   }
   
   void loop() {
       // wait until rate is changed to something
       if (rate != 0)
       {
           // set LED high
           digitalWrite(led, HIGH);
           
           // delay
           delay(rate);
           
           // set LED low
           digitalWrite(led, LOW);
           
           // delay
           delay(rate);
       }
   }
   
   // this function automagically gets called upon a matching POST request
   int setBlinkRate(String rateString)
   {
       // convert string sent from Losant to integer
       float temp = rateString.toFloat();
       
       // invert to go from frequency to rate
       temp = 1/temp;
       
       // then multiply to go from seconds to milliseconds for delay()
       temp = temp*1000;
       
       // and divide by 2 to blink at 50% duty cycle
       rate = temp/2;
   }

4.  At Losant, go to the Integrations tab and add an Integration
5. Choose a name. Select "Particle" as the Integration type. Copy the access token from the Particle Build Settings tab into the Access Token field on Losant. Save.
6. Back on the Losant Workflow, drag a "Particle Call" node into the workspace and connect the Webhook Node output to this Particle Call input. For the integration, select the Particle device you made in the previous step. If you copied my code, then the Function Name is "setBlinkRate". You'll need to copy the ID of your Particle device from https://console.particle.io/devices and enter it into the "Device Name or ID" field. For the Function Argument, choose "Payload Path". Copy "data.body.request.intent.slots.rate.value" into the "Argument Payload Path". This is the location of the variable sent from Alexa to Losant. The Losant workspace should now look like this:

That's it! You should now be able to speak to the Alexa and have the LED on the Particle Photon blink at the frequency you stated.