Research

Paper

AI LLM March 02, 2026

Architecture-Aware Multi-Design Generation for Repository-Level Feature Addition

Authors

Mingwei Liu, Zhenxi Chen, Zheng Pei, Zihao Wang, Yanlin Wang, Zibin Zheng

Abstract

Implementing new features across an entire codebase presents a formidable challenge for Large Language Models (LLMs). This proactive task requires a deep understanding of the global system architecture to prevent unintended disruptions to legacy functionalities. Conventional pipeline and agentic frameworks often fall short in this area because they suffer from architectural blindness and rely on greedy single-path code generation. To overcome these limitations, we propose RAIM, a multi-design and architecture-aware framework for repository-level feature addition. This framework introduces a localization mechanism that conducts multi-round explorations over a repository-scale code graph to accurately pinpoint dispersed cross-file modification targets. Crucially, RAIM shifts away from linear patching by generating multiple diverse implementation designs. The system then employs a rigorous impact-aware selection process based on static and dynamic analysis to choose the most architecturally sound patch and avoid system regressions. Comprehensive experiments on the NoCode-bench Verified dataset demonstrate that RAIM establishes a new state-of-the-art performance with a 39.47% success rate, achieving a 36.34% relative improvement over the strongest baseline. Furthermore, the approach exhibits robust generalization across various foundation models and empowers open-weight models like DeepSeek-v3.2 to surpass baseline systems powered by leading proprietary models. Detailed ablation studies confirm that the multi-design generation and impact validation modules are critical to effectively managing complex dependencies and reducing code errors. These findings highlight the vital role of structural awareness in automated software evolution.

Metadata

arXiv ID: 2603.01814
Provider: ARXIV
Primary Category: cs.SE
Published: 2026-03-02
Fetched: 2026-03-03 04:34

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.01814v1</id>\n    <title>Architecture-Aware Multi-Design Generation for Repository-Level Feature Addition</title>\n    <updated>2026-03-02T12:50:40Z</updated>\n    <link href='https://arxiv.org/abs/2603.01814v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.01814v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Implementing new features across an entire codebase presents a formidable challenge for Large Language Models (LLMs). This proactive task requires a deep understanding of the global system architecture to prevent unintended disruptions to legacy functionalities. Conventional pipeline and agentic frameworks often fall short in this area because they suffer from architectural blindness and rely on greedy single-path code generation. To overcome these limitations, we propose RAIM, a multi-design and architecture-aware framework for repository-level feature addition. This framework introduces a localization mechanism that conducts multi-round explorations over a repository-scale code graph to accurately pinpoint dispersed cross-file modification targets. Crucially, RAIM shifts away from linear patching by generating multiple diverse implementation designs. The system then employs a rigorous impact-aware selection process based on static and dynamic analysis to choose the most architecturally sound patch and avoid system regressions. Comprehensive experiments on the NoCode-bench Verified dataset demonstrate that RAIM establishes a new state-of-the-art performance with a 39.47% success rate, achieving a 36.34% relative improvement over the strongest baseline. Furthermore, the approach exhibits robust generalization across various foundation models and empowers open-weight models like DeepSeek-v3.2 to surpass baseline systems powered by leading proprietary models. Detailed ablation studies confirm that the multi-design generation and impact validation modules are critical to effectively managing complex dependencies and reducing code errors. These findings highlight the vital role of structural awareness in automated software evolution.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.SE'/>\n    <published>2026-03-02T12:50:40Z</published>\n    <arxiv:comment>18 pages, 6 figures</arxiv:comment>\n    <arxiv:primary_category term='cs.SE'/>\n    <author>\n      <name>Mingwei Liu</name>\n    </author>\n    <author>\n      <name>Zhenxi Chen</name>\n    </author>\n    <author>\n      <name>Zheng Pei</name>\n    </author>\n    <author>\n      <name>Zihao Wang</name>\n    </author>\n    <author>\n      <name>Yanlin Wang</name>\n    </author>\n    <author>\n      <name>Zibin Zheng</name>\n    </author>\n  </entry>"
}