Swarms are a metaphor for complex dynamic systems. In swarms, large numbers of individuals locally interact and form non-linear, dynamic interaction networks. Ants, wasps and termites, for instance, are natural swarms whose individual and group behaviors have been evolving over millions of years. In their intricate nest constructions, the emergent effectiveness of their behaviors becomes apparent. Swarm-based computational simulations capture the corresponding principles of agent-based, decentralized, self-organizing models. In this work, we present ideas around swarm-based developmental systems, in particular swarm grammars, a swarm-based generative representation, and our efforts towards the unification of this methodology and the improvement of its accessibility.