Jiajun Wu, Jian Yang, Yaxin Du, Shukai Liu, Linzheng Chai, Tuney Zheng, Wei Zhang, Haitao Yang · 2026 · arXiv
DOI: 10.48550/arXiv.2610.31118 · open-access PDF
Agen pengkodean frontend biasanya dinilai pada satu pertanyaan: apakah target yang diminta dibangun. Perubahan rekayasa nyata membawa kewajiban kedua, karena sisanya dari situs harus terus bekerja. Kami menyebutnya kewajiban kedua kewajiban website kontinuitas, dan memperkenalkan ProductWebBench, benchmark yang skor baik pada repositori frontend hidup: 320 perubahan tugas atas kode yang ada dan 80 tugas long-horizon Build. Setiap tugas membekukan snapshot repositori, persyaratan tahap, bukti browser dan pemeriksaan deterministik sebelum evaluasi, dan memasuki hanya sekali referensi melewati setiap cek dan plausible-
Machine translation (Meta M2M100); the English original follows.
Frontend coding agents are usually graded on one question: did the requested target get built. Real engineering changes carry a second obligation, because the rest of the site has to keep working. We call that second obligation website continuity, and introduce ProductWebBench, a benchmark that scores both on living frontend repositories: 320 Change tasks over existing code and 80 long-horizon Build tasks. Every task freezes a repository snapshot, staged requirements, browser evidence and deterministic checks before evaluation, and enters only once a reference passes every check and plausible-
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.31118, 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