Automated imagery pattern recognition engineers apply established automated workflows to their tasks with limited to no examination of assumptions built into the individual functional blocks. This paper identifies numerous assumptions implicit to workflow blocks inclusive of task definition, data acquisition, data processing, training/inferencing, generating decisions, and assessing confidence. The effect that each assumption has on the established workflow is accompanied by a worked example using downlooking, panchromatic electro-optical images, acquired from a satellite platform. The information is presented so that engineers can utilize the identification of these assumptions to re-examine their processing procedures, resulting in robust and resilient systems.