Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description goes beyond the annotations by explaining immediate effects (status becomes 'invited'), future effects (becomes 'active' on first sign-in), and edge cases (re-inviting a pending invite refreshes it, re-inviting an active member is refused). This gives the agent valuable expectations for side effects and failure modes, while the annotations already note it is not read-only or destructive.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.