Bridging Behavior and Implementation: Automated Java Glue Code Generation for Behavior-Driven Development

2026-07-21

AI Summary

This work addresses the high manual effort required in Behavior-Driven Development (BDD) for writing glue code, which demands simultaneous comprehension of natural-language behavior specifications and underlying system implementations. To overcome this challenge, the paper introduces AutoGlue, the first hierarchical multi-agent framework leveraging large language models to automatically generate BDD glue code for Java projects through a behavior-first workflow. AutoGlue innovatively integrates a behavior interpreter with developer agents that collaboratively parse behavioral intent, retrieve project-aware contextual information, and synthesize executable glue code. Experimental evaluation on eight open-source Java projects demonstrates that AutoGlue outperforms few-shot prompting by 58.7% in API F1 score and 43.7% in CodeBLEU, with 46.1% of the generated step definitions being directly usable without modification.

Abstract

Behavior-Driven Development (BDD) helps technical and non-technical stakeholders share a common understanding of software requirements through natural-language scenarios. Glue code makes these scenarios executable by mapping each step to the corresponding project code. However, developing and maintaining glue code requires knowledge of both the intended behavior and the underlying codebase, making it a labor-intensive part of BDD as requirements evolve. Although large language models (LLMs) have shown strong code generation capabilities, their use for automated glue code generation remains unexplored. This task requires reasoning over underspecified behavior, related BDD artifacts, and large project codebases. We present AutoGlue, a hierarchical multi-agent framework for automated Java glue code generation. AutoGlue follows a behavior-first workflow that separates behavior interpretation, context retrieval, and code generation. A Behavior Interpreter derives the intent of a step from its scenario context, while a Developer agent retrieves relevant BDD artifacts and project code before generating the final glue code. We evaluate AutoGlue on 1,307 steps from eight open-source Java projects. Compared with few-shot prompting, AutoGlue improves API F1 by 58.7% and CodeBLEU by 43.7%. It produces directly usable glue code for 46.1% of the evaluated steps, while most partially correct outputs require only minor revisions, such as adding missing actions or refining parameters. Ablation results show that behavior interpretation and project-aware context retrieval both contribute substantially to generation quality. These findings demonstrate that LLMs can effectively connect natural-language behavior specifications with project code and support specification-driven software development.

Problem

• Behavior-Driven Development
• glue code
• code generation
• natural-language scenarios
• software requirements

Innovation

• automated glue code generation
• behavior-driven development
• large language models
• multi-agent framework
• Java