Self-Evolving Software Agents

TL;DR

This paper presents self-evolving software agents that combine BDI reasoning with LLMs, enabling autonomous goal and behavior evolution in dynamic environments.

Contribution

It introduces a novel BDI-LLM architecture with an automated evolution module for autonomous software evolution.

Findings

Agents can autonomously discover new goals from experience.

Agents generate executable behaviors with minimal prior knowledge.

The approach demonstrates feasibility but faces limits in behavioral inheritance and stability.

Abstract

Autonomous agents can adapt their behaviour to changing environments, but remain bound to requirements, goals, and capabilities fixed at design time, preventing genuine software evolution. This paper introduces self-evolving software agents, combining BDI reasoning with LLMs to enable autonomous evolution of goals, reasoning, and executable code. We propose a BDI-LLM architecture in which an automated evolution module operates alongside the agent's reasoning loop, eliciting new requirements from experience and synthesizing corresponding design and code updates. A prototype evaluated in a dynamic multi-agent environment shows that agents can autonomously discover new goals and generate executable behaviours from minimal prior knowledge. The results indicate both the feasibility and current limits of LLM-driven evolution, particularly in terms of behavioural inheritance and stability.