In the era of computational complexity, autonomous agent-based systems stand as an established paradigm that simplifies various stages of project conception and design compared to classical approaches. This chapter explores the rich trajectory of Agent-Oriented Software Engineering (AOSE), delineating its path from conceptual foundations to contemporary applications with a specific focus on the contributions presented during Workshop on Objects and Agents (WOA). Our study begins with a retrospective analysis of how AOSE methodologies have evolved to implement increasingly advanced agents. We trace this progression from methodologies designed for simple state machine-based reactive systems through those supporting cognitive agents to sophisticated approaches enabling the development of Belief-Desire-Intention (BDI) models. This evolution culminates in contemporary methodologies that facilitate the creation of complex agents capable of real-time introspection, reflection, and self-adaptation. The second part of this chapter analyzes the core principles and practices of AOSE methodologies and their development and how they evolved during WOA events. When creating an agent methodology, we emphasise that a good principle is to follow clearly defined rules that consider appropriate abstractions for each context, making it possible to move from requirements to code when applying the methodology. Also, we underline and explore that essential tools for this are the metamodel and a notation language suitable for developing models for each phase of the methodology, as well as the selected framework or implementation platform to be used. We conclude with a forward-looking perspective on AOSE, outlining innovative directions and potential research areas.