Reproduce “Generation of Web Apps with Agentic IDEs: An Empirical Assessment” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Manuel Marceca, Maria Teresa Rossi, Leonardo Mariani
The authors evaluate whether IDE-integrated agents can generate complete web applications from natural-language specifications. They compare GitHub Copilot, Cursor, and Windsurf across five applications, using iterative validation and correction. All projects passed the study’s checks, but generated tests accounted for 49.6% of requirement-level errors; the distributed Task Queue was hardest, limiting conclusions to these tasks.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.