This work proposes WebDesignIter, a framework built around a persistent knowledge graph (WebAppArchKG) that fuses repository structure with design knowledge and keeps both in sync across development cycles, and outperforms every general-purpose coding agent Claude Code, OpenHands, SWE-Agent, Codex CLI on every model configuration.
Abstract
Front-end development accumulates change after change at the repository level, weaving complex cross-file dependencies that current LLM coding agents tuned for single-shot tasks cannot reliably track across multiple iterations, leading to functional regressions and code that resists maintenance. We argue the missing piece is design knowledge: architectural principles, module responsibilities, and structural constraints that developers lean on to keep code readable, maintainable, and evolvable as a system scales. To operationalize this, we propose WebDesignIter, a framework built around a persistent knowledge graph (WebAppArchKG) that fuses repository structure with design knowledge and keeps both in sync across development cycles. WebDesignIter works in two stages: design-informed planning pulls historical context and architectural overviews from WebAppArchKG to produce an implementation plan with corresponding test scripts, and design-aware generation executes that plan through targeted diff-based patches, validated by sandbox execution and automatic syntax repair. On Web-Bench, WebDesignIter delivers an average Pass@2 gain of 9.55 percentage points across nine foundation models over existing baselines. More importantly, WebDesignIter outperforms every general-purpose coding agent Claude Code, OpenHands, SWE-Agent, Codex CLI on every model configuration, posting the highest Pass@1 and Pass@2 while consuming 2530 fewer input tokens. Ablation singles out design knowledge as the most impactful component: stripping it drops Pass@1 by 11.40 percentage points, a degradation far larger than removing code-graph retrieval, patch-based generation, or sandbox verification, confirming that design knowledge provides a fundamentally more efficient and reliable path to repository-level code generation.
Results show that CODENS produces highly relevant and well-grounded answers, while qualitative feedback highlights a remaining challenge in concise, documentation-oriented synthesis.
Abdelhak Kelious, Chyrine Tahri, Eliot Bardet· Proceedings of the 2026 ACM...· 0 citations
Repo0 is presented, a continuous structural evolution framework for zero-to-all code generation that maintains an explicit architectural state instantiated as a Dual-Directed-Acyclic-Graph (Dual-DAG), consisting of a requirement-level DAG, a component-level DAG, and their alignment relation.
Silin Chen, Haoyi Teng, Xiaodong Gu et al.· 0 citations
Repository-level code translation is critical for modernizing legacy systems, yet existing approaches based on large language models (LLMs) operate at the file level and fail to scale to codebases with complex inter-file dependencies. This limitation is evident in our industrial setting, where we aim to migrate a production repository (STAR) from Java to Kotlin, but file-level approaches produce fragmented results and fail to achieve end-to-end correctness. In this paper, we show that the primary cause of failure at the repository level is dependency inconsistency. Through an empirical study on open-source and industrial systems, we find that most errors arise from unresolved cross-file dependencies that cannot be effectively addressed by iterative feedback alone. We propose a dependency-aware incremental migration framework that elevates the unit of translation from individual files to dependency-consistent batches. Our approach constructs a dependency graph, groups interdependent files, and performs batched translation with iterative compile- and test-driven validation. We evaluate our method on a 51K line of code (LOC) industrial system and multiple repositories across interoperable language pairs (Java-Kotlin, Java-Scala, and C#-F#). On the STAR repository, file-level approaches achieve 38.16% compilation and 9.39% test success, whereas our approach achieves 100% compilation and test success across the evaluated settings, converging within a small number of iterations. These results show that dependency-aware batching improves scalability and reliability in repository-level code translation.
Sivajeet Chand, Alexander Pretschner, Steve Haupt et al.· 1 citation
RUCACoder is proposed, a closed-loop multi-agent framework with a Retriever for hierarchical repository exploration, a Verifier for reranking and validation, and a Coder for feedback-driven script synthesis that consistently outperforms strong retrieval and generation baselines.
Kaitao Lin, Songwen Gong, Adam Jatowt et al.· Annual International ACM SIG...· 1 citation
Large language models are increasingly capable of synthesizing executable frontend projects, yet existing benchmarks still treat web generation as a static evaluation problem. We argue that frontend artifacts demand a different paradigm: they are interactive rather than static, admit diverse yet equally valid implementations, and evolve faster than rigid pipelines can accommodate. To address these gaps, we present LiveEvalBench, an automated framework that reformulates web-generation evaluation as an agentic, adaptive, and extensible process. LiveEvalBench instantiates evaluation as a collaborative review workflow, in which a Build Engineer, a Code Engineer, and a UI Tester collectively gather evidence across the full lifecycle of a frontend project, from deployment and code inspection to browser-based interaction. To handle implementation diversity, an adaptive protocol couples shared rubrics for cross-model comparability with implementation-grounded criteria tailored to each artifact. The framework further supports incremental integration of new evaluator roles and assessment dimensions without pipeline redesign. Experiments across diverse real-world web-generation scenarios show that LiveEvalBench aligns closely with human expert judgment and provides fine-grained insights into frontier models'web generation capabilities. Code is available at https://github.com/wyysteelhead/LiveEvalBench
Yiyao Wang, Zhen Wen, Ying Tang et al.· 0 citations
: The growing adoption of large language models (LLMs) in software engineering has introduced new opportunities but also risks in the software maintenance lifecycle. While LLMs can generate entire codebases from natural language prompts, such automatically generated or rapidly prototyped code often accumulates structural debt, making systematic refactoring increasingly urgent. This work investigates LLMs as metric-driven refactoring assistants rather than code generators. Six models (ChatGPT, Claude, Gemini, Grok, DeepSeek, and Qwen) were evaluated on two types of Java projects: three controlled applications with manually inflated structural metrics, and three real-world applications from public GitHub repositories. Using MetricsReloaded in IntelliJ IDEA, we measured four CK metrics: complexity (WMC), cohesion (LCOM), coupling (CBO), and inheritance depth (DIT). Results indicate that LLMs significantly reduce complexity and coupling, improving class simplicity and modularity. However, cohesion improvements remained limited, with LCOM proving especially elusive. Inheritance depth showed strong reductions in synthetic high-metric applications but minimal change in real projects. ChatGPT produced the most consistent and structurally stable refactoring outputs in real applications, though occasional cohesion deterioration occurred. These findings suggest that while LLMs are valuable assistants for structural improvement, their interventions require careful monitoring to avoid unintended trade-offs.
Tindwende Sawadogo, Fadel Touré· Proceedings of the 21st Inte...· 0 citations