What problem does it solve? Modeling agent cognition with beliefs, desires, and intentions requires a formal, queryable structure that plain text or ad-hoc data models cannot provide. This Skill converts external RDF context into BDI mental states grounded in a formal ontology, enabling deliberative reasoning, explainability, and semantic interoperability across multi-agent systems. ## Core Features & Use Cases - T2B2T Pipeline: Implements the Triples-to-Beliefs-to-Triples paradigm, translating incoming RDF into beliefs and projecting deliberation results back into RDF for downstream systems. - Formal BDI Ontology Patterns: Provides class hierarchies, object properties, OWL restrictions, and DOLCE alignment for Belief, Desire, Intention, Plan, WorldState, and Justification entities. - Framework Integration: Includes translation patterns for SEMAS production rules, JADE/JADEX agent platforms, FIPA ACL messaging, and Logic Augmented Generation pipelines that constrain LLM outputs with ontological validation. - Use Case: An agent receives a payment-request notification as RDF triples; the Skill guides generating a belief grounded in that world state, a motivated desire, a committed intention with a task-ordered plan, and a final world state assertion once the plan executes. ## Quick Start Use the bdi-mental-states skill to model an agent's beliefs, desires, and intentions from the attached RDF context and generate the corresponding Turtle triples.
