I'm trying a similar approach @stalbert while I wait for the end point to be fixed with a permanent/more reliable solution...