GitHub · Python
MSKazemi/aobench
Role-aware, permission-enforced benchmark for LLM agents that operate HPC systems — SLURM, telemetry, RBAC, docs, facility. A policy violation hard-fails the task. 88 tasks (10 categories x 5 roles), 29 reproducible environments with 6 rebuilt from real Marconi100 ExaData, scored on 7 weighted dimensions.
- Stars
- 9
- Forks
- 14
- Stars gained in 24 hours
- 0
- Forks gained in 24 hours
- 0
Topics: agent-benchmark, agent-evaluation, ai-agents, ai-safety, aiops, benchmark, evaluation-framework, high-performance-computing, hpc, langfuse
Latest data point: Nov 22, 2026 · 9 stars · 14 forks