In the quest for simplified development methodologies that allow for the inclusion of AI agents, I found a gap. This article presents a lightweight approach tailored for cross-functional teams. The core principle revolves around a clear handover process: tasks should not transition silently; each transfer must be a conscious act accompanied by context. This methodology is designed to function effectively whether the team consists of three members, fifteen, or includes AI agents among its participants. It is a concept open to critique and discussion.
As the head of a client and analytics solutions development center with over 50 professionals across seven cross-functional teams, I have experienced various methodologies like Scrum, Kanban, and Scrumban. Each has its strengths, yet they all share a common shortcoming. None adequately address what happens during the transition of a task from one role to another. For instance, when a developer marks code as 'Ready for Testing,' what follows? Does the tester see it? Are they aware of the context? What if the tester is unavailable? Who is accountable for the task? These methodologies outline iterations and roles but overlook the critical moment of handover, where operational losses often occur. With the introduction of AI agents into our workflows, new questions arise regarding task completion signals and validation of results.