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

Annotations already declare readOnlyHint, openWorldHint and destructiveHint, and the description goes well beyond them: the 25,000-row platform limit, the JSON hard cap, CSV's startRow pagination that stops when the API rejects the next page, the summary.truncated flag, and CSV writing rows to a file and returning the path. These are non-obvious failure modes an agent could not infer from structured fields. It stops short of describing pagination tokens or timing for JSON.

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?

Purpose is front-loaded in the first sentence and every subsequent sentence carries distinct information about limits or format behavior. The middle sentence about CSV pagination is long and clause-heavy, which slightly hurts readability, but nothing is redundant.

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?

For a 12-parameter, no-output-schema analytics tool, the description covers the highest-risk aspects: result-set completeness, truncation signaling, and file-vs-inline output. It leaves a few parameters unexplained, but those are documented (or at least enumerated) in the schema, so an agent has enough to call it correctly.

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?

Schema description coverage is 67%, and the description compensates for the biggest gap: `format` has no schema description at all, and the text explains that csv writes to a file and returns a path. It also adds the cross-parameter rule that the hour dimension requires data_state=hourly_all (set automatically when unset), which the schema does not state. Remaining params (search_type, aggregation_type, chart, range) are left to the schema.

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?

States a specific verb and resource — "Query Google Search performance data grouped by dimensions (query, page, country, device, searchAppearance, date, hour) with filters" — and enumerates the dimensions, so an agent knows exactly this is the raw Search Analytics retrieval tool rather than a derived report like report_quick_wins or report_ctr_outliers. It is distinguishable from every sibling without opening a schema.

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 gives real operational guidance — use format=csv when you need more than the 25,000-row JSON cap, and check summary.truncated before treating results as complete — but gives no routing advice about when to call this versus the many report_* siblings that presumably wrap it. Usage context is implied rather than contrasted against alternatives.

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