HarnessRisk Benchmark
Explore score history ↗HarnessRisk evaluates agent harness safety across six lifecycle phases with 128 sandboxed cases. It measures Utility, Attack Success Rate, Persistence, and Detection for each trajectory.
- Released
- 2026-08-18
- Institutions
- University of Central Florida · University of North Carolina at Chapel Hill · Michigan State University
- Evaluates
- Agent safety, Agents, Cybersecurity, Persistent-state safety, Prompt-injection resistance, Safety, Safety & Trustworthiness, Tool-use safety
- Openness
- unknown
- Importer
- Claire Radar
- Review status
- source-reviewed
HarnessRisk paper, code and dataset
No reported scores are on record for this benchmark yet.