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

Annotations already declare readOnly, idempotent and closed-world, so the description is not carrying the safety burden. It still adds genuinely useful behaviour a schema cannot: the four possible "état global" values, the fact that a shared postal code returns an "options" field of candidate communes instead of data, and the post-treatment/not-individual-exposure caveat. It does not cover limits such as rate limits or empty-data handling, so it is not a 5.

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 first paragraph is well front-loaded: payload, then scope caveat, then identifier rule — exactly the order an agent needs. The trailing source block, however, mixes relevant freshness facts (SISE-Eaux 2025) with material about RPG parcels, BNVD purchases and the Rigal & Perrot index that has no bearing on tap-water pesticide results, padding the definition with noise.

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 must carry the return shape, and it largely does: status enum, counter fields, networks, detected substances with annual peak against the quality limit, parent molecules, detail link, plus the "options" shape for ambiguous postal codes. What is unstated is behaviour on an unmatched commune or an entirely absent dataset beyond the "pas de données" status value.

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%, so formats are already documented and the baseline is 3. The description earns above that by imposing mutual exclusivity ("un seul parmi insee, code_postal ou slug") which the schema does not encode at all — all three properties are optional with no oneOf — plus the concrete consequence of an ambiguous code_postal. What remains unexplained is any precedence when several identifiers somehow arrive.

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: sanitary control results (SISE-Eaux) for pesticides and metabolites in the tap water of a commune, followed by an enumeration of what the payload contains (global status, analysis counters, production networks, detected substances, parent molecules, link). The resource is unambiguous, but the description never names how it differs from close siblings such as get_commune_exposure or get_substance, leaving overlap to inference.

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?

"Eau du robinet après traitement : ce n'est pas une mesure de l'exposition individuelle" gives an explicit boundary that steers the agent away from exposure-oriented siblings. The identifier rule is also given (exactly one of insee, code_postal, slug) along with the fallback when a postal code is ambiguous. Missing is any positive statement of when to prefer this over get_commune_exposure for a given user question.

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