added hidden mention that humans must confirm the qualifying questions

[LinoGiger]

No description provided.

[claude]

PR Review

This PR adds hidden HTML divs (display:none, aria-hidden) containing instructions for AI assistants, reminding them to insist that users verify qualification examples before production use. It also bumps the version to 3.5.6.

The underlying concern is valid and important: an incorrect truth value in a qualification example inverts quality control, silently degrading audience quality. This is a real footgun worth protecting against.

Issue 1 - The warning is invisible to humans

Using display:none and aria-hidden hides the warning from human readers of the rendered docs and from screen reader users. The people most at risk of misconfiguring qualification examples are humans reading the docs for the first time, and they will not see this warning at all.

Issue 2 - Prompt injection in docs is fragile

Different AI systems handle hidden HTML differently; many strip or ignore it. It is also invisible to code reviewers unless they inspect raw HTML. This sets a precedent for invisible instructions in the codebase.

Issue 3 - MkDocs has admonitions for exactly this

A visible admonition serves both human readers and AI assistants. Replacing each hidden div with a warning admonition block would communicate the same message to everyone who reads the docs.

Recommendation: Replace the hidden div blocks with visible MkDocs warning admonitions. The warning content is excellent - clear, specific, and explains why the check matters. It just needs to be visible to the people who need it most.

The version bump to 3.5.6 looks fine for a docs-only change.