CapSeal: Capability-Sealed Secret Mediation for Secure Agent Execution

Abstract

Modern AI agents routinely depend on secrets such as API keys and SSH credentials, yet the dominant deployment model still exposes those secrets directly to the agent process through environment variables, local files, or forwarding sockets. This design fails against prompt injection, tool misuse, and model-controlled exfiltration because the agent can both use and reveal the same bearer credential. We present CapSeal, a capability-sealed secret mediation architecture that replaces direct secret

Categories

Framework mappings

Suggested from the entry's categories.

@misc{jin2026capseal,
  title = {{CapSeal: Capability-Sealed Secret Mediation for Secure Agent Execution}},
  author = {Shutong Jin and Ruiyi Guo and Ray C. C. Cheung},
  year = {2026},
  month = apr,
  eprint = {2604.16762},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2604.16762}
}