Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations cover safety (readOnlyHint=true, destructiveHint=false, idempotentHint=true, openWorldHint=true), so the agent knows it's a safe, repeatable read operation. The description adds minimal behavioral context—it implies a GET request but doesn't detail response format, pagination, or error conditions. With annotations doing heavy lifting, the description adds little extra value, warranting a baseline score.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.