Reproduce “SkillFlow: Scalable and Efficient Agent Skill Retrieval System” 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
Fangzhou Li, Pagkratios Tagkopoulos, Ilias Tagkopoulos
SkillFlow retrieves relevant agent skills from a large shared library through a staged search that narrows candidates with dense retrieval, reranking, and selection. On SkillsBench, GPT-5-mini Pass@1 rose from 9.2% without skills to 16.4% with retrieved skills, reaching 84.1% of the oracle score. On Terminal-Bench, use on 70.1% of tasks did not significantly improve Pass@1 over the no-skill baseline.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.