The goal of this paper is to present a logical framework for the formalizat
ion of agents' mutual beliefs in a Multi Agent system. The approach is base
d on a combination of extensional specifications of beliefs and context-bas
ed (finite) presentation of the specifications by employing a particular cl
ass of Multi Context systems. The extensional specification provides a set-
theoretic characterization of beliefs in terms of sets closed under certain
conditions. Its finite presentation is provided by using as constructors i
nference rules inside a Multi Context system. The resulting framework allow
s for capturing many relevant cases of real (not omniscient) agents, which
are very common in Multi Agent scenarios embedded in real world environment
s. In order to substantiate this claim, two Multi Agent scenarios are forma
lly specified in detail in the specification framework. (C) 1999 John Wiley
& Sons, Inc.