PydanticAI integration

Action security for PydanticAI.

Add action policy to typed tools and agent execution.

Get the Python SDK on PyPITry the local demoRead the documentation

Get the package and source.

Connect the execution boundary.

Attach the firewall to the agent and decorate the intended tools. Type validation describes an argument’s shape; action policy evaluates its business context.

An example to test

A correctly typed refund request names the wrong recipient. Check role, purpose and destination before calling the payment system.

Illustrative workflow, not a customer deployment or a measured result.

Before connecting live policy

  1. Identify the handler that actually performs the side effect.
  2. Provide trusted identity, purpose and final action arguments.
  3. Configure the tenant’s assigned runtime and keep API credentials server-side.
  4. Test allowed, blocked and unavailable paths before production.

Integration example

Adapt this existing SDK pattern to your application. Names such as the agent, handler and session refer to your own objects. Follow the package documentation for compatible versions and complete setup.

# Install PydanticAI support.
python -m pip install "agenticdome-python-sdk[pydanticai]"

from agenticdome_sdk.pydantic import CyberSecFirewall

firewall = CyberSecFirewall()
firewall.attach_to_agent(customer_support_agent)
@firewall.secure_tool
async def refund_customer(ctx, customer_id: str, amount: int) -> dict: ...

agenticdome-demo --framework pydanticai --scenario refund_hijack

Python demo commands evaluate fixed inputs against a bundled local policy; they do not instantiate the selected framework or prove a live integration. The TypeScript client and OpenClaw runtime plugin need their documented tenant configuration.

Read the integration researchUnderstand runtime verification

See the decision before the action.

Try a local scenario, then bring your workflow to a deployment review. Start with one agent and one tool, and see exactly where the action can be stopped.

Try the local demo Book a deployment review