critique-toylisted

# critique-toy Reviews a markdown-prose artifact composed in the toy domain's own grammar (`bench/generator/README.md`, "Worked example: the toy domain") against its three TOY-* criteria: a passive-voice recast that deletes the actor (TOY-ACTIVE), a hedging phrase stacked ahead of an otherwise direct statement (TOY-HEDGE), and a subheading left with no body before the next heading of equal or higher level (TOY-ORPHAN). This skill is `docs/internal/skill-template.md`'s own committed fixture: it exists to prove the template builds end to end (S-04 spec, Behavior/Examples), not as a seventh launch skill. It lives under `skills/_template-fixture/` rather than directly under `skills/` precisely so it is never discovered as a shipped component, and it is never registered in `library.json`. ## Contract Every finding this skill emits conforms to `contract/critique-contract.schema.json`. See `docs/reference/critique-contract.md` for the field contracts a schema cannot check on its own: location navigable unaided, evidence quoted or measured rather than characterized, violation naming the breach, fix actionable and specific. ## Protocol Follow these four passes in order. Do not skip ahead to severity or fixes while still sweeping. 1. **Inventory.** Map the artifact's structure (sections, headings, components, whatever the artifact type has). No judgments yet, no findings yet. This pass exists so the sweep in step 2 does not anchor on whatever was noticed first. 2. **Criteri