Reproduce “Work2Skill: Capturing Human GUI Experience for Reusable Agent Skill Induction” 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
Yuqi Huang, Hui Li, Yifei Cao, Zikun Cai, Tianci Xie, Yangfan Gao, Pinxue Guo, Chaoyun Zhang and 4 more
Work2Skill turns human GUI demonstrations into reusable procedural skills by linking actions to visual and available accessibility evidence, then letting an agent inspect that evidence while drafting and revising instructions. On 64 MyPCBench tasks, GPT-5.5 scores 85.27 versus 74.73 without skills. Across tested executors, gains appear on MyPCBench and a selected, non-random 286-task OSWorld-Verified subset.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.