Jarvis: LLM-Assisted Plan and Belief Generation via Artifact Introspection

• Lucas C. Raupp UFSC
• Gustavo Paulo UFSC
• Ana Laura S. Chagas UFSC

DOI: https://doi.org/10.5753/wesaac.2026.31351

Resumo

BDI agents traditionally rely on plan libraries authored at design time, which limits their ability to operate in environments containing artifacts the programmer did not anticipate. We present Jarvis, a JaCaMo-native gateway that lets agents acquire plans and beliefs from a large language model at runtime, grounded on reflective introspection of the CArtAgO artifacts present in the workspace. The LLM receives the discovered operation signatures of each artifact and produces an AgentSpeak plan library; a validator filters the response against the discovered operations before two internal actions insert the surviving plans and beliefs into the agent’s mental state. We illustrate the approach with a three-agent case study in which an agent learns to operate an air-conditioner artifact entirely from LLM-generated knowledge and serves comfort requests issued by peer agents.

Referências

Aguzzi, G., Ciatto, G., Ferrando, A., Gatti, A., and Mascardi, V. (2025). LLMs as agents, LLMs at the service of agents, or agents at the service of LLMs? Extended abstract, 26th Workshop “From Objects to Agents” (WOA 2025). Introduces the PlanchBDI proposal unifying chattification and LLM-based plan generation under BDI control.

Boissier, O., Bordini, R. H., Hübner, J. F., Ricci, A., and Santi, A. (2013). Multi-agent oriented programming with JaCaMo. Science of Computer Programming, 78(6):747–761.

Bordini, R. H., Hübner, J. F., and Wooldridge, M. (2007). Programming Multi-Agent Systems in AgentSpeak using Jason. John Wiley & Sons.

Bratman, M. E. (1987). Intention, Plans, and Practical Reason. Harvard University Press, Cambridge, MA.

Ciatto, G., Aguzzi, G., Battistini, R., Baiardi, M., Burattini, S., Ricci, A., et al. (2025). Exploiting GenAI for plan generation in BDI agents. Frontiers in Artificial Intelligence and Applications, 413:3495–3502. Proceedings of the 28th European Conference on Artificial Intelligence (ECAI 2025).

Gatti, A., Mascardi, V., and Ferrando, A. (2025). ChatBDI: think BDI, talk LLM. In Proceedings of the 24th International Conference on Autonomous Agents and Multiagent Systems (AAMAS), pages 2541–2543.

Hong, S., Zhuge, M., Chen, J., Zheng, X., Cheng, Y., Wang, J., Zhang, C., Yau, S., Lin, Z., Zhou, L., et al. (2024). MetaGPT: Meta programming for a multi-agent collaborative framework. In 12th International Conference on Learning Representations (ICLR).

La Malfa, E., La Malfa, G., Marro, S., Zhang, J., Black, E., Luck, M., Torr, P., and Wooldridge, M. (2025). Large language models miss the multi-agent mark. In Advances in Neural Information Processing Systems (NeurIPS), Position Paper Track, volume 38.

Rao, A. S. and Georgeff, M. P. (1995). BDI agents: From theory to practice. In Proceedings of the First International Conference on Multi-Agent Systems (ICMAS), pages 312–319.

Ricci, A., Piunti, M., and Viroli, M. (2011). Environment programming in multi-agent systems: An artifact-based perspective. Autonomous Agents and Multi-Agent Systems, 23(2):158–192.