> ## Documentation Index
> Fetch the complete documentation index at: https://gpars.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Security considerations

> Security properties and implementation guidance for GPARS deployments.

## Security properties

<CardGroup cols={2}>
  <Card title="No self-authorization" icon="ban">
    The user's security policy is the sole authorization authority. Agents cannot self-authorize or grant themselves permissions.
  </Card>

  <Card title="Verified identity" icon="fingerprint">
    Agent identity is verified by infrastructure under the user's control at the plane boundary — agents cannot self-assert their identity to the Action Plane.
  </Card>

  <Card title="Continuous enforcement" icon="shield-halved">
    The enforcement point applies security policy on every operation before it reaches MCP servers.
  </Card>

  <Card title="Audit surface" icon="list-check">
    Every EMO crosses the MCP boundary, creating a natural audit point for all agent actions.
  </Card>
</CardGroup>

By externalizing EMOs, agents avoid executing untrusted or privileged actions internally. Authorization denials are standard and structured, enabling agents to handle them gracefully without leaking policy details.

## Implementation notes

<Info>
  GPARS v0.1 intentionally leaves implementation choices open. The following are suggestions, not requirements.
</Info>

* **Enforcement point implementations:** reverse proxies, network isolation, Unix socket permissions, containers, runtime sandboxes, or OS-level isolation.
* **Identity verification mechanisms:** mTLS, API keys, process-level isolation, or other methods appropriate to your deployment.
* **Policy management engines:** static config files, RBAC engines, or interactive approval prompts.