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

TL;DR

RAIM introduces a localization mechanism that conducts multi-round explorations over a repository-scale code graph to accurately pinpoint dispersed cross-file modification targets and employs a rigorous impact-aware selection process based on static and dynamic analysis to choose the most architecturally sound patch and avoid system regressions.

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.

Figures (6)

• Figure 1: Motivating example on instance pylint-dev__pylint-8190.
• Figure 2: Overview of RAIM. 1 Parse the code repository into a structure tree; 2 Localize initial suspicious files; 3 Analyze module call relationships of suspicious files; 4 Localize feature-relevant files; 5 Generate query text for feature-relevant functions; 6 Retrieve feature-relevant functions from neighbor nodes; 7 Update candidate functions with retrieved results; 8 Generate multiple feature-relevant designs; 9 Localize line-level editing positions; 10 Generate diverse patches based on multi-design; 11 Perform code change impact analysis; 12 Evaluate patches based on impact analysis results.
• Figure 3: The example of RAIM on instance pylint-dev__pylint-8190.
• Figure 4: Construction of Feature-relevant Code Subgraph.
• Figure 5: Performance comparison between RAIM and Agentless across LLMs on NoCode-bench Verified.