Metraipolis › archive › dc_cce8247fec35

Do Coding Agents Reuse Existing Code or Reinvent the Wheel? — arXiv:2609.35357

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.35357
License (as archived)
CC0-1.0
Archived
2026-10-07T21:03:29.403Z
SHA-256 of the archived text
3e72faac2d9eb9f6e85c64bed0ec4543bffe4833c99e1b5ec8502d8858b5ffa5
Stable URL of this page
https://metraipolis.com/documento/dc_cce8247fec35

How to cite

Do Coding Agents Reuse Existing Code or Reinvent the Wheel? — arXiv:2609.35357 — arXiv (en), https://arxiv.org/abs/2609.35357 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_cce8247fec35 (Metraipolis archive dc_cce8247fec35, sha256 3e72faac2d9eb9f6e85c64bed0ec4543bffe4833c99e1b5ec8502d8858b5ffa5)
BibTeX
@misc{dccce8247fec35,
  title = {Do Coding Agents Reuse Existing Code or Reinvent the Wheel? — arXiv:2609.35357},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.35357},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_cce8247fec35; sha256 3e72faac2d9eb9f6e85c64bed0ec4543bffe4833c99e1b5ec8502d8858b5ffa5},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.35357. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Coding agents are increasingly deployed for iterative development on real repositories, yet existing evaluation barely answers a basic question: \emph{do coding agents reuse existing code or reinvent the wheel?} The question matters: every duplicated implementation is a fix applied twice and agents produce code far faster than humans can audit, so redundancy accumulates unsupervised. Thus, we present \textbf{RepoReuse}, a multi-turn benchmark for auditing code reuse in real repositories, where requirements are revealed turn by turn and the workspace accumulates across turns. It is built by a fully automated pipeline combining AST-based dependency graphs, guided evidence collection, and execution-verified task synthesis, and scales readily to new repositories. Beyond pass rates, we measure the reuse rate together with recall and cross-turn structural redundancy. An audit over 3{,}000 turns shows that agents progressively stop exploring relevant repository code, reuse their own history less even when it is fully in the workspace, and leave duplicated logic in 50.8\% of task chains by turn~5---all while pass rates barely move. Such deficiencies are invisible to pass rates, underscoring the need to evaluate code generation beyond functional correctness.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Software Engineering (cs.SE) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.35357 — 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 2026-10-07T20:29:03.323Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).