← all datasets

SKILLVETBENCH

Emerging
2papers using it
2026first seen

SKILLVETBENCH is a live public leaderboard that evaluates the security of community-contributed skills for open-source LLM agents using an LLM-as-Judge to provide a multi-dimensional risk assessment.

Papers using SKILLVETBENCH (2)

SKILLVETBENCH dataset β€” papers, benchmarks & downloads Β· AI Agents