Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnly, idempotent, and non-destructive behavior, so the description adds context beyond those: it discloses return constraints such as 'official anchors only', 'no raw quote tables', and 'trend direction and anchors only'. This clarifies what the agent will receive and what it will not, which is valuable for setting expectations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.