The fix for an agent that keeps getting your design system wrong isn't a better prompt.

It's putting the rules where the agent reads them.

In the repo. In skills. In checks it can run itself.

Writing a course about exactly this.

What sells it for me is the refusal telling the agent exactly where the token should go. A bare "no hex" error usually just sends it looping. Does guard.mjs run only as an agent hook, or in pre-commit too so it catches humans as well?

A CPU runs four stages in under a nanosecond. Fetch, decode, execute, write back.

Try explaining that in a paragraph.

So I made a 3D exploration where every scroll is one stage. A piece turns from a technical drawing into 3D and locks into place, until the chip is complete.

When people can see what your product does, they don't need a paragraph to believe it.

Nice!