Erebots framework enables flexible AI conversation agents to work together

EREBOTS 4.0: A framework for conversational assistants with A2A agents and MCP tools

Multiagent Systems

Summary

Conversational assistants like chatbots often use multiple AI agents that either rely on fixed rules or language models, but combining these effectively has been a challenge. The authors introduce EREBOTS 4.0, a new open-source software framework that helps different types of AI agents communicate and work together smoothly. It uses protocols to let language-model agents plan tasks and call specialized tools that provide exact answers. This setup makes it easier to add agents and track conversations while ensuring user privacy and task history.

What this means in practice

Authors

Elia Pacioni, Davide Calvaresi, Andrei C. Coman, Ismael Jaggi, Valerio Crocetti, Gaetano Manzo, Michael Ignaz Schumacher

Abstract

In recent years, multi-agent systems have shifted from rule-based, symbolic agents to sub-symbolic agents built on language models. However, several limitations remain unresolved. The new frontier is neuro-symbolic agents, which pair language models with deterministic tools and explicit protocols. Without a common framework to standardize and compose them, each study rebuilds the same infrastructure, and results are hard to replicate. This paper presents EREBOTS 4.0, an open-source, hybrid-AI, multi-agent framework that fills this gap for conversational assistants. Language-model agents plan and route requests, while deterministic tools compute exact results over the Model Context Protocol (MCP). An orchestrator selects one or more specialist agents per turn and calls them over the Agent2Agent protocol (A2A). A new agent, local or remote, is added by listing its URL in a configuration file. The framework also manages identity, user context with consent flags, task states and traces. Its A2A server passed 97 tracked requirements in the official A2A Technology Compatibility Kit, with no failures.