I'm trying to generate insights off of a file I upload to dropbox at midnight ET every night.  This file contains a simple list of tickers and a trading signal (-1, 0, 1).  

     By using the algorithm parameter Update or OnSecuritiesChanged, I'm able to download my file and parse it for signals.  (I bascially copied the logic I used to create my universe in my ScheduledUniverseSelectionModel).

      However, since these signals are generated before market open, I'd like to be able to use them in a custom execution model that places auction orders.  As these must be submitted more than 15 minutes before the open, I can't generate my insights in Update and use them for my ideal ExecutionModel (Update only runs with each new data tick, and those are only generated post-open). 

   Is there a way to implement using a Dropbox signals file like I am envisioning in the algorithm framework?