Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already convey read-only/idempotent hints, but the description adds valuable behavioral context: return format (one row per filing, filer names, max percent), exclusions (no item narrative text), the meaning of event_revision null, and data delay. This goes well beyond the structured annotations and sets accurate expectations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.