I’m currently working on the system to allow you to create custom actions that can be attached to tiles, nfc tags or geofences. Its much more detailed than what the current app has. The question is could the voice command system use this same system. I know with the current app the commands can be complicated to say because they are generic to be able to support multiple units and most of the services that can be attached to devices. I’ve been thinking maybe voice commands should work like the actions mentioned above, in that you select the exact action you want to do and then just type what you want the voice command too be for that action. I have not experimented with this idea yet but it seems pretty logical and would IN THEORY allow users to create exactly the voice commands they want. what do you all think?
I don’t see why not. Let’s say you create an scene called “Turn Off Lights”, you could tell Cortana, “Grasshopper, turn off lights.” That should send the command to Grasshopper to run the action. I’m not sure how the voice commands are built from the API perspective, but I would imagine it could work the same way.
Hopefully i’ll be able to get the beta out soon. 
Don’t tempt me with such news!