How to develop, in terms of validating your robots actions

I’m indoctrinated through blue-prism devolping thriough a Citrix enviroment and i’m wondering how sure you can be of the type into activities. in a normal program indentifiable through AA, how much validation would you use?

Normal developing procedure:

  1. validate you are the place you expected to be - else exception
  2. take the action
  3. Validate the action have been taken


  1. verify that the window frame of AX 2009 - “kunder” are open and activated - by using Element Exists - else throw
  2. Send “Ctrl + n” - by using type into
  3. verify that “ny kunde” window is there - else try step 2 again up to 3 times at max - else throw

I know that this is a somewhat religion - some will control and some will only control in case faults arise - but what is your organizations standpoints on this? how forgiving are your organization (not the grumpy colleague that will clean things up) in case of mass-mishandled cases?

Hi @Emil.tophoej

I suppose with the Native Citrix Support, there might be a bit less of a need to double check yourself on every step. In general, I believe robustness comes first, and then the ability to throw meaningful exception shortly after.

That’s neat, haven’t tried that yet tho (moved away from citrix)

What does robustness involve?

As in make the “happy path” work? What defines where you want to add exceptions?

Do you start with sequence or work flow normally?