II.
Capability overview
Reference · livecapability:supports-guardrails
Supports guardrails overview
The framework or agent supports input/output validation guardrails that run in parallel with agent execution. Guardrails can inspect and validate user input before processing and agent output before delivery, enabling safety and compliance enforcement.
Attributes
displayName
Supports guardrails
description
The framework or agent supports input/output validation guardrails
that run in parallel with agent execution. Guardrails can inspect
and validate user input before processing and agent output before
delivery, enabling safety and compliance enforcement.
appliesToNodeKinds
- AgentVersion
- AgentRuntimeImpl
category
safety
Outgoing edges
None.
Incoming edges
supports2
- agent-core-impl:openai-agents-sdk.core@current·AgentCoreImplOpenAI Agents SDK Core (current)
- agent-version:openai-agents-sdk@current·AgentVersion