Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare destructiveHint=true, signaling the tool's destructive nature. The description adds no further behavioral traits (e.g., loss of form state, reloading resources). With annotations bearing the burden, a score of 3 is appropriate for not augmenting beyond them.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.