Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already provide read-only, idempotent, open-world, and non-destructive hints, lowering the bar. The description adds value by specifying exactly what information is retrieved (fields, geometry type, record count, capabilities), going beyond the annotations to clarify the tool's output behavior. This is useful additional context for an agent.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.