Will De Michele, Kla Tantithamthavorn · 2026 · arXiv (Cornell University)
DOI: 10.48550/arxiv.2610.27381 · cited by 0 works (OpenAlex) · open-access PDF
Agentes de codificação de IA agora contribuem para repositórios de software em escala, mas suas práticas de documentação permanecem sem caracterização. Analisamos 29,669 solicitações de puxão geradas por IA de 2,585 repositórios no conjunto de dados do AIDev, medindo a co-intercâmbio de documentação: se as modificações de arquivos de código são acompanhadas de uma atualização de arquivos de documentação no mesmo PR. Descobrimos que 63,2% dos PRs de IA mudam de código sem atualizar qualquer documentação. Isso varia acentuadamente por tipo de tarefa: 55,2% para recursos e 77,5% para refactorização. Arquivos de código não documentados são seguidos de uma correção para o mesmo arquivo de 8,6 a 11,7% mais frequentemente do que arquivos documentados em 30 para janelas de
Machine translation (Meta M2M100); the English original follows.
AI coding agents now contribute to software repositories at scale, yet their documentation practices remain uncharacterised. We analyse 29,669 AI-generated pull requests from 2,585 repositories in the AIDev dataset, measuring documentation co-change: whether code-file modifications are accompanied by a documentation-file update in the same PR. We find that 63.2% of AI PRs change code without updating any documentation. This varies sharply by task type: 55.2% for features and 77.5% for refactoring. Undocumented code files are followed by a fix to the same file 8.6 to 11.7% more often than documented files across 30 to 90-day windows. A raw PR-level comparison shows the opposite because documented PRs are 3.9 times larger and therefore attract more fixes. Controlling for size exposure reverses this result. Documented changes also merge faster despite their larger size (median: 40 seconds versus 184 seconds), although merging is largely automated in this dataset. Documentation drift also compounds: within 90 days, a file changed without documentation is later re-documented at roughly half the rate of a file changed with documentation (24% versus 48%). This suggests that documentation debt tends to accumulate rather than be revisited.
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
Topics: Documentation, Merge (version control), Computer science, Internal documentation, Task (project management), Coding (social sciences)
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.27381, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange