In this paper, we describe an ontologically well-founded agentbased discrete event simulation framework with a high-level rulemulation language and an abstract simulator architecture and execution model, which allows both basic discrete event simulations without agents and complex agent-based simulations with (possibly distorted) perceptions and (possibly false) beliefs. The framework includes the simulation language AORSL, a Java simulator with a statistics and a 2D visualization tab, a player for the 3D rendering of simulation logs, and a Web simulation server. It is available on www.AOR-Simulation.org. Categories and Subject Descriptors I.2.0 [ARTIFICIAL INTELLIGENCE]: Cognitive simulation Keywords Simulation, cognition, beliefs.