Model-Driven Discipline for Multi-Agent LLMs: Requirement-to-Verification Generation of Traceable System Models

2026-07-18

AI Summary

This work addresses the challenges of poor traceability, difficult maintenance, and complex change management in traditional model-driven engineering—particularly acute in safety-critical systems—stemming from tool heterogeneity and model fragmentation. To overcome these limitations, the paper introduces RADIANT, a novel approach that integrates multi-agent large language models into an end-to-end model-driven development pipeline. RADIANT automatically generates heterogeneous models encompassing conceptual, domain-specific linguistic, structural, and behavioral aspects directly from requirements models, while establishing element-level executable traceability links. These links enable automated change impact analysis and formal verification. Experimental results demonstrate that RADIANT significantly outperforms single-agent baselines in generating syntactically valid and executable models. User studies further indicate a 10–15× improvement in development efficiency and strong cross-domain transferability.

Abstract

Software complexity is a long-standing challenge for system engineers. Model-Driven Engineering (MDE) addresses it by treating models as first-class artefacts, but a typical MDE process spans many tools and produces heterogeneous models of different system aspects, making traceability, maintenance, and change management difficult. We propose RADIANT, an engineering methodology that combines MDE with Multi-Agent Large Language Models (LLMs) for complete model-based system development, with a focus on safety-critical systems. From a carefully specified requirement model, RADIANT automatically generates heterogeneous models across engineering phases -- a concept model, a domain-specific modelling language, a conforming system model, and a behaviour model -- together with executable, element-level traceability links, on top of which it provides exact, automated change-impact analysis. Generated behaviour models are translated into CSP and formally verified (e.g.\ for deadlock freedom and convergence) with a counterexample-driven repair loop. Evaluating RADIANT across three LLMs, we find that the multi-agent decomposition reliably improves the \emph{syntactic validity} of the generated formal artefacts over a single-agent baseline -- and their \emph{executability} where the model's code generation permits -- while gains in semantic accuracy are model-dependent. A six-participant study shows an order-of-magnitude ($10$--$15\times$) reduction in development time, and the unmodified pipeline transfers to a second domain.

Problem

• Model-Driven Engineering
• Multi-Agent LLMs
• Traceability
• Heterogeneous Models
• Change Management

Innovation

• Model-Driven Engineering
• Multi-Agent LLMs
• Traceability
• Formal Verification
• CSP