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

Annotations already declare this a safe, idempotent, closed-world read computation, so the description does not need to argue safety. It adds real behavioral context instead: the output keys, that optionNetLePlusEleve is a label 'sans recommandation', and the precise conditions under which optionNetLePlusEleveFaisable is false (cash below price → partial property quote; subsidiary loan capped at cash with unmodeled complement). This explains the limits of the model, which annotations cannot.

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?

Scope and currency handling are front-loaded correctly, but the single run-on paragraph buries the output contract and the feasibility caveat mid-sentence with heavily nested parentheticals. The content earns its place, yet a short bulleted split of inputs / outputs / caveats would be far more scannable.

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 9-parameter cross-border tax model with no output schema, the description does cover return fields and caveats, but it omits the meaning of most inputs and any note on assumptions such as loan rate usage (tauxPret) or holding horizon. An agent can call it, but cannot reason about whether its inputs are being interpreted as intended.

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 only 22%, so the description must compensate. It only establishes that amounts are in GBP and that an exchange rate is supplied ('Entrées en livres sterling avec le taux de change'). It does not define tresorerieSASU, tmiFrance, horizonAnnees, loyersBrutsAnGBP, chargesAnnuellesGBP, or prixReventeGBP, and the feasibility clause only hints at tresorerie and prix indirectly.

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 opens with a specific verb and scope: 'Compare deux façons d'investir au Royaume-Uni depuis une SASU française', and enumerates the fiscal dimensions covered (droits d'acquisition, impôt sur les loyers, plus-value, IS britannique, retenue à la source). This clearly separates it from simulate_immo_espagne/belgique/portugal and the generic compare_* siblings, which do not mention UK non-resident taxation or the 2008 treaty.

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?

The scope implies when to use it (a French SASU weighing UK property held directly vs via a subsidiary), but no alternative is named and no exclusion is stated. It does not say, for example, when to prefer simulate_pv_immo_ir or a generic compare_* tool, leaving routing to inference from the country name.

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