Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With only readOnlyHint and openWorldHint supplied, the description carries real weight: it discloses the statistical method, the confidence-interval treatment, the three grouping dimensions of the output, and an explicit warning that the result is informational rather than a determination. That last point is the kind of misuse-prevention context annotations cannot express. It omits the data scope (which claims, what time window) and any refresh/cost characteristics.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

Two sentences, no filler, with the metric and its grouping dimensions front-loaded ahead of the consumer and the caveat. The density costs some readability — 'dx-1' and 'log-log confidence intervals' appear without explanation — but nothing is wasted.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

There is no output schema, so the description is the only place an agent can learn the return shape, and it gives only the grouping keys plus the notion of a duration with intervals. It does not say what the metric values look like, what the data window is, or how fresh the index is — gaps that matter for a zero-parameter analytic call.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema has zero parameters, so there is nothing for the description to disambiguate; baseline is 4. It implicitly signals that the call takes no filters, meaning the aggregation scope is fixed server-side, though it does not state what that fixed scope is.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a concrete computed artifact — a Kaplan-Meier decay index measuring how long claims of each check type, claim kind and host class stay true — rather than restating the tool name. The closing 'Display only, not a verdict' distinguishes it from verdict-producing siblings such as verify_claim or get_status. It stops short of an explicit verb and leans on unexplained jargon ('dx-1'), but the resource is unmistakable.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

It names its consumers ('the recheck scheduler and an agent choosing how long to trust a claim'), which implies a clear when-to-use condition. The exclusion 'not a verdict' hints at what it should not be used for. However, it never names the alternative tool an agent should call instead when it actually needs a truth determination.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.