agent-framework/docs/design/guardrails.md

Guardrails

The design goal is to provide a flexible and extensible way to implement guardrails and a built-in set of guardrails that can be used for common use cases.

NOTE: this is work in progress.

Guardrails can be template-based to adapt to different input data types, which include:

• Message for agent messages.
• ToolCall for tool call requests.
• ToolResult for tool call results.

Guardrails are added to other components such as ModelClient and MCPServer as hooks that are called before and after the main logic of the component.

For example, the ModelClient has methods to add input and output guardrails.

model_client = ModelClient(...)
model_client.add_input_guardrails([
    PIIGuardrail[Message](...),
    SensitiveDataGuardrail[Message](...),
])
model_client.add_output_guardrails([
    HarmfulContentGuardrail[Message](...),
])

Another example to show how to use a guardrail with an MCP server:

guardrail = PIIGuardrail(
    config={
        "rules": [
            {
                "type": "email",
                "action": "block"
            },
            {
                "type": "phone",
                "action": "block"
            }
        ]
    }
)

mcp_server = MCPServer(...)
mcp_server.add_output_guardrail(guardrail)