Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare the tool read-only, idempotent, and non-destructive. The description adds only the optional filtering behavior, which is also captured in the schema. It does not add new behavioral context like pagination or auth requirements, but the output schema covers return values.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.