Hi everyone. Just got my nano-claw instance running on my home server this week. I'm trying to understand the guardrails better.

I want to create a rule that stops the agent from making any network calls to domains I haven't explicitly allowed. I saw the default configs, but they seem focused on content, not network control. Is this something I can do in the `config.yml`?

Specifically:
- Where in the flow do I hook this check?
- Do I need to write a custom action?
- How do I log the blocked attempt without exposing too much private data from the agent's context?

My setup is Docker on Ubuntu, using the basic python toolkit. Any pointers to the right part of the docs would help a lot. Still finding my way around.