Towards Exception Safety Code Generation with Intermediate Representation Agents Framework

TL;DR

Seeker is a multi-agent framework that enhances exception safety in LLM-generated code by decomposing exception handling into specialized agents, utilizing a comprehensive knowledge base, and improving robustness and accuracy in real-world programming scenarios.

Contribution

The paper introduces Seeker, a novel multi-agent system with an IR approach and a knowledge base to improve exception handling in code generated by LLMs, significantly reducing fragility.

Findings

Outperforms state-of-the-art baselines in exception handling precision by up to 37%

Improves overall code robustness by 38% as measured by expert review

Achieves a 28% success rate in real-world issue fixes, surpassing prior methods' 19%

Abstract

Large Language Models (LLMs) often struggle with robust exception handling in generated code, leading to fragile programs that are prone to runtime errors. We propose Seeker, a novel multi-agent framework that enforces exception safety in LLM generated code through an Intermediate Representation (IR) approach. Seeker decomposes exception handling into five specialized agents: Scanner, Detector, Predator, Ranker, and Handler that collaboratively analyze code, detect fragile segments, retrieve best practice exception strategies, and inject robust handling code. We also introduce Common Exception Enumeration (CEE), a comprehensive knowledge base derived from official documentation, technical practices, and real world code, to standardize exception handling strategies. Seeker also incorporates a Deep Retrieval-Augmented Generation (Deep RAG) algorithm to efficiently navigate the exception inheritance hierarchy, cutting down search overhead by 93% while improving accuracy in identifying relevant exceptions. We evaluate Seeker on 15 open source Java projects and multiple benchmarks. Seeker outperforms state of the art baselines, improving exception handling precision by up to 37% and overall code robustness by 38% as measured by expert code review. It significantly closes the gap between LLM and human developers in exception management, achieving a 28% success rate on real world issue fixes (SWE bench) versus 19% by prior methods. Our framework preserves functional correctness of code while proactively handling errors, demonstrating a practical, generalizable solution for safer code generation. In this paper, we discuss the novelty of using intermediate representation and multi-agent collaboration for exception handling, and outline how Seeker can be extended to other programming languages and complex software engineering tasks, aligning LLM-generated code with industrial standard.