paper-with-me

Papers

RepoMaster: Autonomous Exploration and Understanding of GitHub Repositories for Complex Task Solving

2025-05-27 · Huacan Wang, Ziyi Ni, Shuo Zhang, Shuo Lu, Sen Hu, Ziyang He, Chen Hu, Jiaye Lin, Yifu Guo, Yuntao Du, Pin Lyu

The ultimate goal of code agents is to solve complex tasks autonomously. Although large language models (LLMs) have made substantial progress in code generation, real-world tasks typically demand full-fledged code repositories rather than simple scripts. Building such repositories from scratch remains a major challenge. Fortunately, GitHub hosts a vast, evolving collection of open-source repositories, which developers frequently reuse as modular components for complex tasks. Yet, existing frameworks like OpenHands and SWE-Agent still struggle to effectively leverage these valuable resources. Relying solely on README files provides insufficient guidance, and deeper exploration reveals two core obstacles: overwhelming information and tangled dependencies of repositories, both constrained by the limited context windows of current LLMs. To tackle these issues, we propose RepoMaster, an autonomous agent framework designed to explore and reuse GitHub repositories for solving complex tasks. For efficient understanding, RepoMaster constructs function-call graphs, module-dependency graphs, and hierarchical code trees to identify essential components, providing only identified core elements to the LLMs rather than the entire repository. During autonomous execution, it progressively explores related components using our exploration tools and prunes information to optimize context usage. Evaluated on the adjusted MLE-bench, RepoMaster achieves a 110% relative boost in valid submissions over the strongest baseline OpenHands. On our newly released GitTaskBench, RepoMaster lifts the task-pass rate from 24.1% to 62.9% while reducing token usage by 95%. Our code and demonstration materials are publicly available at https://github.com/wanghuacan/RepoMaster.

📄 PDF Abstract BibTeX arXiv:2505.21577

Code (1)

wanghuacan/repomaster 공식 구현

Tasks

Code Generation

Similar Papers 제목 키워드 기반

RepoMasterEval: Evaluating Code Completion via Real-World Repositories

2024-08-07 · Qinyun Wu, Chao Peng, Pengfei Gao, Ruida Hu 외

With the growing reliance on automated code completion tools in software development, the need for robust evaluation benchmarks has become critical. However, existing benchmarks focus more on code generation tasks in fun…

Code CompletionCode GenerationDescriptive

GitTaskBench: A Benchmark for Code Agents Solving Real-World Tasks Through Code Repository Leveraging

2025-08-26 · Ziyi Ni, Huacan Wang, Shuo Zhang, Shuo Lu 외 arxiv

Beyond scratch coding, exploiting large-scale code repositories (e.g., GitHub) for practical tasks is vital in real-world software development, yet current benchmarks rarely evaluate code agents in such authentic, workfl…

EnvX: Agentize Everything with Agentic AI

2025-09-09 · Linyao Chen, Zimian Peng, Yingxuan Yang, Yikun Wang 외 arxiv

The widespread availability of open-source repositories has led to a vast collection of reusable software components, yet their utilization remains manual, error-prone, and disconnected. Developers must navigate document…

Speech RecognitionCode Generation

Active Zero: Self-Evolving Vision-Language Models through Active Environment Exploration

2026-02-11 · Jinghan He, Junfeng Fang, Feng Xiong, Zijun Yao 외 arxiv

Self-play has enabled large language models to autonomously improve through self-generated challenges. However, existing self-play methods for vision-language models rely on passive interaction with static image collecti…

LEGION: Harnessing Pre-trained Language Models for GitHub Topic Recommendations with Distribution-Balance Loss

2024-03-09 · Yen-Trang Dang, Thanh-Le Cong, Phuc-Thanh Nguyen, Anh M. T. Bui 외

Open-source development has revolutionized the software industry by promoting collaboration, transparency, and community-driven innovation. Today, a vast amount of various kinds of open-source software, which form networ…