Search papers, labs, and topics across Lattice.
This paper introduces SkillAlchemy, a framework for creating reusable agent skills from open-world materials, addressing the challenge of skill generation when human authorship and model priors are unavailable. By leveraging contrastive evidence to identify implicit requirements and compiling candidate procedures into a grammar-guided skill package, SkillAlchemy significantly enhances the skill creation process. Experimental results show a 19.9 percentage point improvement in pass rates over no-skill execution and an 8.6 percentage point increase compared to the best automated baseline, achieving performance on par with human-curated skills.
SkillAlchemy boosts agent skill creation efficiency, achieving nearly 20% higher success rates without relying on human authorship.
Agent skills are reusable procedural artifacts that extend language agents with specialized workflows, tool conventions, and domain behaviors at inference time. However, creating reliable skills still depends largely on human authorship, model priors, or execution traces. These sources are often unavailable for unfamiliar tasks, suggesting the need to create skills from open-world materials. In this paper, we study open-world skill creation: given an underspecified skill brief and a source-access specification, a creator must discover behavior-relevant requirements omitted by the brief and determine how broadly each source-derived procedure is justified. We propose SkillAlchemy, an admission-centered framework for source-grounded skill creation. SkillAlchemy identifies implicit requirements through contrastive evidence, admits candidate procedures based on evidence-supported scope, and compiles the admitted content into a grammar-guided skill package. Extensive experiments across 87 SkillsBench v1.1 tasks demonstrate that SkillAlchemy improves pass rate over no-skill execution by 19.9 percentage points and the strongest automated baseline by 8.6 percentage points, while achieving performance comparable to human-curated skills.
