The authors built SkillScriptBench after surveying more than 35,000 GitHub-hosted agent Skill roots.
Agents Repair Executable Skill Packages Better With Code-Structure Guidance
A 350-task benchmark shows that linking maintenance requests to relevant code structures improves both repair rates and consistency.
Academic
Yuxuan Liu · Haoran Li · Yuhao Zhang · Jiahe Guo · Hongyu Luo · Wenbin Hu · +7 more
The Hong Kong University of Science and Technology · Harbin Institute of Technology, Shenzhen, China · Harbin Institute of Technology, Harbin, China
Research Digest··3 min read
Liu et al.
Why this paper
From The Hong Kong University of Science and Technology and 2 others
In one line
SkillScriptBench shows AST-guided, scope-limited editing substantially improves reliable repair of executable agent skills while coordinating scripts and documentation.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ·No stated limitations found
- ✓Reports numbers on named benchmarks
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§