Challenge 1 — Solution

Task: Explain why "the rules arrived automatically" and "the rules are
correct" are two different claims, using one concrete example from
earlier in this course to illustrate the gap between them.

"The rules arrived automatically" is a claim about the LOADING
mechanism working correctly - that CLAUDE.md's own @rules/ include
list successfully pulled in the full content of every referenced file
before the conversation began, exactly as Chapter 1 and this chapter's
own five-step flow describe. "The rules are correct" is a completely
separate claim, about whether the CONTENT of what got loaded is still
accurate and up to date. A loading mechanism working perfectly says
nothing about whether the information it's faithfully delivering is
still true.

The concrete example: Chapter 5's own finding about the kanji tiles
page's back-link rule. That rule loaded correctly, every single
session, with total reliability - the @rules/ mechanism never failed
to deliver it. And yet what it described was genuinely broken: the
rule pointed every kanji page's own "back to grid" link at a stale,
now-nonexistent old-site URL, breaking that link sitewide, until the
2026-08 rules review actually checked the live behaviour and caught
it. The loading mechanism had a perfect track record here - it was
the CONTENT being delivered that was wrong, and no amount of reliable
delivery would have caught that on its own.

Notes:
- This is the same distinction Chapter 8 drew between "the memory says
  X exists" and "X exists now" - just applied to the rules-loading
  system instead of an individual memory file.
- The practical implication: reliable automatic delivery is a real
  improvement over the old R1 system (nothing gets forgotten to load),
  but it doesn't remove the separate, ongoing need to actually verify
  a rule's own content is still accurate - that's still deliberate,
  human-driven work, not something the loading mechanism does for you.