paperJuly 2026Unreviewed

Trusted Credentials, Untrusted Behavior: Benchmarking LLM-Agent Security in High-Performance Computing

Jie Li

Abstract

Large language model (LLM) agents are starting to take on routine work in high-performance computing (HPC), including monitoring Slurm jobs, diagnosing failed builds, inspecting simulation output, and coordinating scientific workflows. To do this work, an agent commonly acts under its user's credentials and inherits the user's access to files and the scheduler. This arrangement creates a failure mode that ordinary account-level controls do not capture. Adversarial instructions in a log, tool des

Cite

@misc{li2026trusted,
  title = {{Trusted Credentials, Untrusted Behavior: Benchmarking LLM-Agent Security in High-Performance Computing}},
  author = {Jie Li},
  year = {2026},
  month = jul,
  eprint = {2607.18485},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2607.18485}
}