Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden and does add one useful behavioral trait: it applies to 'all text nodes' in a scope, indicating bulk application. However, it does not disclose whether existing font overrides are overwritten, whether changes are reversible, or how 'scope' is interpreted (e.g., recursive).
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.