Currently, if a user has matched an intent, if using a dialog model, you can elicit a specific slot.

There are times when you want to continue to elicit that slot, for example if you are playing a game that requires the user and Alexa to go back and forth with responses with the user filling the same elicited slot each time.

However if the user says Help, or let's say you have another intent for asking the score or any other question, that moves to another intent, you cannot elicit that previous intent's slot anymore. For any skill, a user saying Help will cause this problem, even if you are only wanting to elicit the one time.

At that point, you then have to prompt the user to say the right thing to get back to the previous intent in order to elicit again. This is clumsy and disruptive to the flow.

Of course you can't just decide to use any intent you want the user to use, but in the case of the user actually previously intending an intent, all this feature request is proposing is to allow to go immediately back to that previous intent and elicit slots from that intent.

So right now the updated Intent must matches exactly with the intent that the skill receives, but this feature request is for allowing it to match either the intent it receives or the previous intent it had received.