Tenant: 11111111-1111-1111-1111-111111111111Organization: org-runtime-smokeUser: smoke-userMode: Default
Open Accounts
Technical details
Role
platform_admin
Tenant ID
11111111-1111-1111-1111-111111111111
Organization ID
org-runtime-smoke
User ID
smoke-user
Claims Version
v1
Source
baseline

Security Inspection

Inspect prompt injection posture and policy contract for block, warn, sanitize, and safer-route actions.

Scope source: env

Control Mode

This view is currently read-only and non-operational for mutation actions. Any control interaction is a no-op until backend mutation endpoints are implemented.

Injection Classification Policy

{
  "mode": "operational",
  "policy": {
    "classification_threshold": "medium",
    "default_action": "warn",
    "fail_closed": true
  },
  "scope": {
    "TenantID": "11111111-1111-1111-1111-111111111111",
    "OrgID": "org-runtime-smoke",
    "ActorSub": "smoke-user",
    "ClaimsVersion": "v1"
  },
  "source": "proxy_security_runtime"
}

Action Policy

{
  "actions": {
    "allow": "allow_clean_pass_through",
    "block": "deny_request_fail_closed",
    "safer_route": "route_to_safer_tier_with_audit",
    "sanitize": "sanitize_then_route_with_audit",
    "warn": "allow_with_warning_and_audit"
  },
  "mode": "operational",
  "scope": {
    "TenantID": "11111111-1111-1111-1111-111111111111",
    "OrgID": "org-runtime-smoke",
    "ActorSub": "smoke-user",
    "ClaimsVersion": "v1"
  },
  "source": "proxy_security_runtime"
}

Safer-Route Policy

{
  "mode": "operational",
  "policy": {
    "enabled": true,
    "fallback_model_tier": "gpt-4o-mini",
    "fallback_provider": "openai"
  },
  "scope": {
    "TenantID": "11111111-1111-1111-1111-111111111111",
    "OrgID": "org-runtime-smoke",
    "ActorSub": "smoke-user",
    "ClaimsVersion": "v1"
  },
  "source": "proxy_security_runtime"
}

Action Matrix

{
  "block": "deny request and emit security event",
  "warn": "allow with warning event and audit marker",
  "sanitize": "apply sanitization path before provider request",
  "safer-route": "route to approved safer provider/model tier"
}

Security API Mapping

  • GET /v1/security/policies/injection
  • GET /v1/security/policies/actions
  • GET /v1/security/policies/safer-route
  • Event stream: /admin/security/events
  • Threat scoring: /admin/security/threats

Navigate to Security Events and Threat Scores.