Metraipolis › archive › dc_06cc2fda1eb2

EvoCodeAgent: Project-Level Code Generation via Architecture-Code Co-Evolution and Experience-Guided Repair — arXiv:2611.06328

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.06328
License (as archived): CC0-1.0
Archived: 2026-11-05T18:26:27.928Z
SHA-256 of the archived text: 09a055c15d50b8eb67dd92fe332907a7a3a4ee59d33170274da05dc7cbb6ab4c
Stable URL of this page: https://metraipolis.com/documento/dc_06cc2fda1eb2

How to cite

EvoCodeAgent: Project-Level Code Generation via Architecture-Code Co-Evolution and Experience-Guided Repair — arXiv:2611.06328 — arXiv (en), https://arxiv.org/abs/2611.06328 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_06cc2fda1eb2 (Metraipolis archive dc_06cc2fda1eb2, sha256 09a055c15d50b8eb67dd92fe332907a7a3a4ee59d33170274da05dc7cbb6ab4c)

BibTeX

@misc{dc06cc2fda1eb2,
  title = {EvoCodeAgent: Project-Level Code Generation via Architecture-Code Co-Evolution and Experience-Guided Repair — arXiv:2611.06328},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.06328},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_06cc2fda1eb2; sha256 09a055c15d50b8eb67dd92fe332907a7a3a4ee59d33170274da05dc7cbb6ab4c},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.06328. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Project-level code generation aims to automatically synthesize complete software repositories from high-level requirements, producing coherent multi-file systems with well-structured modules and consistent cross-file dependencies. Despite the strong capabilities of large language models (LLMs) in specification understanding and code synthesis, existing approaches still exhibit several critical limitations. First, many methods rely on static architectural designs that cannot adapt during iterative generation, leading to a growing misalignment between architecture and implementation. Second, they lack mechanisms to accumulate and reuse experience across iterations, resulting in recurring errors and reduced efficiency. Third, they often depend on standalone LLM-based evaluation of intermediate artifacts, which suffer from limited transparency and potential unreliability under long-context reasoning. To address these challenges, we propose EvoCodeAgent, a state-centric framework for project-level code generation. EvoCodeAgent maintains an evolving architectural state initialized from structured project specifications and continuously reconciles predicted relations from architecture and skeletons with realized relations extracted from generated code. It further uses tool-grounded verification evidence to support reliable state updates and abstracts reusable repair experiences from previous revision attempts to guide subsequent generation. Extensive experiments on project-level benchmarks of different scales, including DevBench and CodeProjectEval, demonstrate that EvoCodeAgent consistently outperforms baseline approaches. Under the GPT-5 setting, it achieves unit test pass rates of 60.48% and 27.73%, respectively, yielding up to a 53.5% improvement over the strongest baseline.

Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.SE, nuovi lavori del giorno).
Categorie arXiv: Software Engineering (cs.SE).
Pagina del lavoro: https://arxiv.org/abs/2611.06328 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-05T18:16:03.883Z dalla pagina https://arxiv.org/list/cs.SE/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).