Typing Multi-Agent Systems via Commitments

This work presents an agent typing system, that differently than most of other proposals relies on notions that are typical of agent systems instead of relying on a functional approach. Specifically, we use commitments to define types. The proposed typing includes a notion of compatibility, based on subtyping, which allows for the safe substitution of agents to roles along an interaction that is ruled by a commitment-based protocol. Type checking can be done dynamically when an agent enacts a role. The proposal is implemented in the 2COMM framework and exploits Java annotations. 2COMM is based on the Agent & Artifact meta-model, exploit JADE and CArtAgO, by using CArtAgO artifacts in order to reify commitment protocols.