Repository-Level Code Generation

Follow this topic Ask about this topic

Repository-level code generation studies how AI systems can implement features, fix issues, or build projects while respecting the code, interfaces, and conventions already present. Common approaches retrieve relevant files with semantic search or dependency graphs, organize requirements into specifications or contracts, and coordinate specialized agents; testing and static analysis provide feedback for refinement. Benchmarks evaluate both repository repair and projects built from scratch, often using hidden tests and real execution environments. A central challenge is that passing tests does not guarantee sound architecture or maintainability, while agents must also find useful, current context without spending excessive time or tokens exploring a codebase.

Latest papers

43 papers7 papers in the last four weeks, up 40% on the four weeks before. 0.0% of all new papers.

Leveraging LLMs for Multi-File DSL Code Generation: An Industrial Case Study
Apr 27, 2026Sivajeet Chand, Kevin Nguyen, Peter Kuntz +1LLM Fine-Tuning Code Generation
Technical University of Munich · Munich, Germany · BMW Group
Contract-Coding: Towards Repo-Level Generation via Structured Symbolic Paradigm
Apr 10, 2026Yi Lin, Lujin Zhao, Yijie ShiCode Generation Repository-Level Code Generation
State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications, Beijing, 100876, China
Beyond Functional Correctness: Design Issues in AI IDE-Generated Large-Scale Projects
Apr 7, 2026Syed Mohammad Kashif, Ruiyin Li, Peng Liang +4Software Engineering AI Coding Agents
School of Computer Science, Wuhan University, China · School of Mathematical and Computational Sciences, Massey University, New Zealand · School of Computer Science, Nanjing University of Science and Technology, China +2
Persistent Cross-Attempt State Optimization for Repository-Level Code Generation
Apr 4, 2026Ruwei Pan, Jiangshuai Wang, Qisheng Zhang +6Repository-Level Code Generation Persistent Memory for Language Models
Chongqing University · Peking University · Harbin Institute of Technology (Shenzhen)
Toward Executable Repository-Level Code Generation via Environment Alignment
Apr 4, 2026Ruwei Pan, Junlei Shen, Linhao Wu +5Code Generation Repository-Level Code Generation
Chongqing University · Peking University · Harbin Institute of Technology (Shenzhen)
Needle in the Repo: A Benchmark for Maintainability in AI-Generated Repository Edits
Mar 29, 2026Haichao Zhu, Qian Zhang, Jiyuan Wang +2Software Engineering AI Coding Agents
Independent Researcher · UC Riverside · Tulane University
Learning to Commit: Generating Organic Pull Requests via Online Repository Memory
Mar 27, 2026Mo Li, L. H. Xu, Qitai Tan +2Repository-Level Code Generation Memory-Augmented Language Model Agents
Tsinghua University
ATime-Consistent Benchmark for Repository-Level Software Engineering Evaluation
Mar 27, 2026Xianpeng, Sun, Haonan Sun +5Software Engineering Benchmarks Temporal Data Leakage
Simon
ReCUBE: Evaluating Repository-Level Context Utilization in Code Generation
Mar 26, 2026Jiseung Hong, Benjamin G. Ascoli, Jinho D. ChoiRepository-Level Code Generation Code Generation Evaluation
Language Technologies Institute, Carnegie Mellon University · Computer Science, Emory University
HCAG: Hierarchical Abstraction and Retrieval-Augmented Generation on Theoretical Repositories with LLMs
Mar 19, 2026Yusen Wu, Xiaotie DengRepository-Level Code Generation Retrieval-Augmented Generation
CodeTaste: Can LLMs Generate Human-Level Code Refactorings?
Mar 4, 2026Alex Thillen, Niels Mündler, Veselin Raychev +1Automated Code Refactoring Automated Software Engineering
Department of Computer Science, ETH Zurich · LogicStar.ai, Zurich, Switzerland
SWE-Adept: An LLM-Based Agentic Framework for Deep Codebase Analysis and Structured Issue Resolution
Mar 1, 2026Kang He, Kaushik RoyCode Localization Multi-Agent Coding
Electrical and Computer Engineering, Purdue University
TENET: One Step Toward Test-Driven Development for Repository-Level Code Generation
Sep 29, 2025Yiran Hu, Shanchao Liang, Nan Jiang +2Code Generation Agentic Code Generation
Purdue University

Apr 27, 2026J/K move · Enter open · S save

Leveraging LLMs for Multi-File DSL Code Generation: An Industrial Case Study

Sivajeet Chand, Kevin Nguyen, Peter Kuntz, Alexander Pretschner

Technical University of Munich · Munich, Germany · BMW Group

LLM Fine-Tuning Code Generation Repository-Level Code Generation LLM Code Generation

Open paper AskPDF arXiv

Large language models (LLMs) perform strongly on general-purpose code generation, yet their applicability to enterprise domain-specific languages (DSLs) remains underexplored, especially for repository-scale change generation spanning multiple files and folder structures from a single natural-language (NL) instruction. We report an industrial case study at BMW that adapts code-oriented LLMs to generate and modify project-root DSL artifacts for an Xtext-based DSL that drives downstream Java/TypeScript code generation. We develop an end-to-end pipeline for dataset construction, multi-file task representation, model adaptation, and evaluation. We encode DSL folder hierarchies as structured, path-preserving JSON, allowing single-response generation at repository scale and learning cross-file dependencies. We evaluate two instruction-tuned code LLMs (Qwen2.5-Coder and DeepSeek-Coder, 7B) under three configurations: baseline prompting, one-shot in-context learning, and parameter-efficient fine-tuning (QLoRA). Beyond standard similarity metrics, we introduce task-specific measures that assess edit correctness and repository structural fidelity. Fine-tuning yields the most significant gains across models and metrics, achieving high exact-match accuracy, substantial edit similarity, and structural fidelity of 1.00 on our held-out set for multi-file outputs. At the same time, one-shot in-context learning provides smaller but consistent improvements over baseline prompting. We further validate practical utility via an expert developer survey and an execution-based check using the existing code generator.