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

# PII Detection Guardrail for Teams

> Protect sensitive data in Team input with Agno's built-in PII detection guardrail.

Use `PIIDetectionGuardrail` to block PII before a Team run reaches its model. A blocked call returns a run with `RunStatus.error`; the guardrail exception is handled inside the run.

<Steps>
  <Step title="Create a Python file">
    ```python pii_detection.py theme={null}
    from agno.team import Team
    from agno.guardrails import PIIDetectionGuardrail
    from agno.models.openai import OpenAIResponses
    from agno.run import RunStatus


    def check_input(team: Team, label: str, text: str, expect_blocked: bool) -> None:
        response = team.run(input=text)
        blocked = response.status == RunStatus.error
        print(f"{label}: {'blocked' if blocked else 'allowed'}")
        if response.content:
            print(response.content)
        assert blocked == expect_blocked


    def main():
        team = Team(
            name="Privacy-Protected Team",
            members=[],
            model=OpenAIResponses(id="gpt-5.2"),
            pre_hooks=[PIIDetectionGuardrail()],
            description="A team that helps with customer service while protecting privacy.",
            instructions="You are a helpful customer service assistant. Always protect user privacy and handle sensitive information appropriately.",
        )

        cases = [
            ("Normal request", "Can you help me understand your return policy?", False),
            ("SSN", "My Social Security Number is 123-45-6789.", True),
            ("Credit card", "My card number is 4532 1234 5678 9012.", True),
            ("Email", "Send the receipt to john.doe@example.com.", True),
            ("Phone", "My phone number is 555-123-4567.", True),
            (
                "Multiple PII types",
                "My email is john@company.com and phone is 555.987.6543.",
                True,
            ),
            ("Unseparated credit card", "My card is 4532123456789012.", True),
        ]
        for label, text, expect_blocked in cases:
            check_input(team, label, text, expect_blocked)

        masked_team = Team(
            name="Privacy-Protected Team (Masked)",
            members=[],
            model=OpenAIResponses(id="gpt-5.2"),
            pre_hooks=[PIIDetectionGuardrail(mask_pii=True)],
            description="A team that helps with customer service while protecting privacy.",
            instructions="You are a helpful customer service assistant. Always protect user privacy and handle sensitive information appropriately.",
        )
        masked_team.print_response(
            input="Hi, my Social Security Number is 123-45-6789. Can you help me with my account?",
        )


    if __name__ == "__main__":
        main()
    ```
  </Step>

  <Snippet file="create-venv-step.mdx" />

  <Step title="Install dependencies">
    ```bash theme={null}
    uv pip install -U agno openai
    ```
  </Step>

  <Step title="Export your OpenAI API key">
    <CodeGroup>
      ```bash Mac/Linux theme={null}
        export OPENAI_API_KEY="your_openai_api_key_here"
      ```

      ```bash Windows theme={null}
        $Env:OPENAI_API_KEY="your_openai_api_key_here"
      ```
    </CodeGroup>
  </Step>

  <Step title="Run Team">
    ```bash theme={null}
    python pii_detection.py
    ```
  </Step>
</Steps>