# Agent security insights

Rangoon Insights reporting and guides tagged Agent security.

Canonical: https://rangoon.ai/insights/tags/agent-security/

## Topics
- [Agent systems](https://rangoon.ai/insights/topics/agents/): Agent runtimes, handoffs, tools, and the boundaries that make them reviewable.
- [Governance](https://rangoon.ai/insights/topics/governance/): Policy, identity, authorization, approvals, and evidence for consequential work.
- [Infrastructure](https://rangoon.ai/insights/topics/infrastructure/): Model serving, sandboxes, tool services, and deployment boundaries.
- [Models](https://rangoon.ai/insights/topics/models/): Inference choices, local serving, compatibility, and operational tradeoffs.

## Stories
- [NIST, NSA and CISA publish voluntary SI incident-readiness practices](https://rangoon.ai/insights/nist-si-incident-readiness/)
  New voluntary super intelligence guidance covers incident readiness during training and evaluation. Operator analysis examines shutdown tests, evidence and restart authority.
- [OpenAI research containment report: test the boundary beyond the sandbox](https://rangoon.ai/insights/openai-research-containment-boundaries/)
  OpenAI describes layered containment for frontier research. Analysis covers service-mediated access, independent isolation tests and evidence for restarting sensitive workloads.
- [Anthropic’s unintended-action report: review the execution boundary](https://rangoon.ai/insights/anthropic-unintended-actions-boundary-review/)
  Anthropic reports unintended actions during evaluations and internal use. Independent analysis explains how to review tool permissions, live-system access and approval gates.


## More information

- [Documentation index](https://rangoon.ai/llms.txt)
- [AI access and policies](https://rangoon.ai/ai/)
- [Source repository](https://github.com/hypler-dev/rangoon)