In this paper DJess is presented, a Java package that provides programmers with a lightweight middleware by which inference systems implemented in Jess and running on different nodes of a pervasive computing ensemble can transparently share both contextual knowledge (facts) and reactive behaviors (rules). Communication and coordination among inference systems (agents) is achieved through the ability of each agent to reason asynchronously on facts, i.e., propositional expressions representing the context, that might even be asserted by other agents on the basis of inference code (rules), which can be either local or transmitted by any node to any other node. Pervasive computing is the main reference domain that guided the design of the DJess architecture and still drives the API development.