I just realized a helpful openclaw use case.
Giving AI agents to your kids more safely.
You can tweak exactly how you want the agent to act, what’s allowed and what’s not…
Much better than ...
Or just teach them to think for themselves before letting LLMs outsource their intellects.