Check and learn the policy applied

• In Chrome, click on F5 Workforce AI security Admin Portal - No Login required, SSO done
• On the left menu, click on Default Policy.

• You can start by the Services tab, and check the services that are enabled and allowed. As you can see a service can be allowed but in Monitoring mode, meaning no action is taken, only monitoring and logging.

• In the MCP Tools tab, check the tools allowed. You can see we only allow Github with List tools. And also check the default actions

  Note

  As you can see, read-only tools are allowed even if they are not explicitily allowed in the policy. And all other tools are blocked.

• In the Sensitive Data tab, you can see the sensitive entities enabled.

• The Classifiers are set in 2 separated sections. The intent is set on the left menu and the classifiers are enabled into the policy. As you can see, 4 classifiers are enabled in blocking mode.

• Click on the left menu Classifiers to see the list of classifiers and select one by clicking on the right arrow.
  • You can see that 100 messages has been added to learn the Classifier on what to detect.

• Then, last things to see today, are the insights. Click on the left menu Insights to see the list of insights.

---

Note

This is the end of the lab guide. Feel free to navigate and explore the Surepath Admin Portal. You can also send your own prompts to try triggering BLOCK or REDACT.