UnitTenX: Generating Tests for Legacy Packages with AI Agents Powered by Formal Verification

2025-10-06

AI Summary

To address low test coverage and difficulty in validating critical paths in legacy code, this paper proposes a synergistic test generation method integrating AI multi-agent systems, formal verification, and large language models (LLMs). The approach employs specialized agents for code comprehension, constraint modeling, test generation, and formal verification, augmented by static analysis and symbolic execution to overcome LLM limitations in logical consistency and boundary-condition reasoning. The resulting end-to-end system automatically generates high-coverage, high-assurance unit tests, significantly improving defect detection while enhancing code readability and documentation. Experimental evaluation across multiple industrial-scale legacy projects demonstrates an average 32.7% improvement in branch coverage and a 91.4% critical-path verification rate. This work delivers a verifiable, scalable, and automated testing solution for safety-critical legacy system refactoring.

Technology Category

• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Multiagent Systems: Multiagent Learning
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

This paper introduces UnitTenX, a state-of-the-art open-source AI multi-agent system designed to generate unit tests for legacy code, enhancing test coverage and critical value testing. UnitTenX leverages a combination of AI agents, formal methods, and Large Language Models (LLMs) to automate test generation, addressing the challenges posed by complex and legacy codebases. Despite the limitations of LLMs in bug detection, UnitTenX offers a robust framework for improving software reliability and maintainability. Our results demonstrate the effectiveness of this approach in generating high-quality tests and identifying potential issues. Additionally, our approach enhances the readability and documentation of legacy code.

Problem

• Generates unit tests for legacy code using AI agents
• Combines formal verification with large language models
• Improves test coverage and software maintainability

Innovation

• AI multi-agent system generates unit tests
• Combines formal methods with large language models
• Automates test generation for legacy codebases