Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations only include openWorldHint=false, so the description carries the burden of behavioral disclosure. It does add meaningful info: calling it once saves the original config and subsequent calls do not overwrite that restore point, implying idempotent behavior. However, it does not state whether the tool mutates state, what permissions are required, or what the output represents.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.