abstract
The goal of this project is to develop a system for explaining, debugging, and validating the behavior of BDI (Belief-Desire-Intention) agents through the reimplementation of a multi-level explainability framework, using a set of declarative sifting patterns. The system aims to bridge the gap between technical implementation details and human-understandable explanations of agent behavior by providing a DSL that allows to extract higher-level narratives out of raw execution traces.
outcomes
