Reproduce “Trajectory-Guided Fault Localization for Agent Skill Evolution” 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

Yu Ge, Linna Xie, Zhong Li, Yu Pei, Tian Zhang

SkillMorph improves code agents’ reusable skills by linking recurring failure-associated actions in execution histories to candidate passages before proposing edits. It abstracts traces, compares action evidence across runs and tasks, then uses an LLM to localize and evaluate revisions. On CannBot, 14/15 tasks passed all three trials, versus 11/15 for CAP-Evolve; both solved 14/15 at least once, in these settings.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.