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

The annotations already cover the safety profile (readOnly, idempotent, non-destructive, closed-world), so the description is free to add operational context — and it does: 'not cached and not counted as a visit in the metrics', the 'incompleto: true' partial-failure signal, and the guarantee that results carry no contacts and no personal names. That is genuinely useful behavior beyond the structured fields, though the response shape itself remains unspecified.

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?

The lead sentence front-loads the return scope ('up to eight Brazilian companies'), and the remaining sentences are dense but each carries distinct information (matching rules, input constraints, failure flag, exclusions, caching). The sentence fragments are choppy 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?

With no output schema, the description must carry the return burden, and it does explain count (up to eight), deduplication, the 'incompleto' flag, and what results exclude. It stops short of describing the actual response fields, but combined with rich read-only annotations this is close to complete for a one-parameter lookup.

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 coverage is 100% and the single 'q' param is documented, so the baseline is 3. The description goes further by explaining input detection ('it detects the kind of input'), the 3-character floor, the phone/e-mail requirements, and ranking rules (exact names before prefixes, no repeated CNPJ, matches legal name, trade name and partners), which meaningfully deepens what the schema alone conveys.

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 states a specific verb ('suggest') and resource ('up to eight Brazilian companies') and enumerates the accepted input kinds (CNPJ, name, partner, phone, e-mail), so the agent knows exactly what the tool produces. It does not, however, distinguish itself from the similarly named sibling 'consulta_sugerir', which by name appears to do the same thing, so the sibling differentiation is missing.

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 concrete input-side conditions (text needs at least three characters, a phone needs the area code, an e-mail needs the full address on record) which is more than most definitions. But it never says when to prefer this tool over 'consulta_cnpj', 'consulta_cadastro', or the near-identical 'consulta_sugerir', so alternative routing is left to inference.

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