Reproduce “CompSkill: Compiling Execution Experience for Skill Retrieval” 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
Qi Shen, Ziyue Zhu, Yongqing Jiang, Zhen Xia, Li Shengjie, Shuai Zhao, Bo Cheng, Xiaobao Wu and 2 more
CompSkill selects procedural guides using outcomes from comparable past executions by the same agent, rather than changing the guides or agent. It recalls guides beyond semantic search and calibrates rankings from matched comparisons. On 332 ToolQA tasks, GPT-5.6-luna completed 256 versus 245 with RankGPT, using one document each (Holm-corrected p = 0.012); shared templates and development-set reuse limit the finding.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.