Reasoning about Organisation: Shaping the Infrastructure

Exploiting artificial intelligence within systems calls for support at any level — models, technologies and methodologies. In this article, we discuss the requirements for an agent infrastructure aimed at the support of automated reasoning over organisation issues. As a reference example, we experiment on the notion of agent coordination context — as embodied by the TuCSoN infrastructure for agent coordination — made available to intelligent agents as a first-class organisation abstraction amenable of inspection and run-time modification.