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

Annotations already declare readOnly, idempotent, non-destructive, non-open-world, so the safety profile is covered. The description adds real behavioral context beyond that: the figures are scoped to the last twelve full months ('mois pleins'), a data-quality signal is included, and it tells the agent to surface the measurement date, which matters for interpreting stale figures. It does not discuss failure modes or what happens if the pharmacy is unidentified.

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 priority instruction ('À appeler en premier') is front-loaded, followed by the payload list, then the exclusions and a one-line English restatement. The middle sentence is a dense enumeration, but each listed field (PAMP valuation, margin, data quality, software-change date) earns its place by telling the agent what it can answer with.

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 zero parameters and no output schema, the description carries the full burden of describing what comes back, and it does so completely: identity fields, historical context, twelve months of financials, and a data-quality indicator plus the reference date. Nothing an agent needs in order to decide to call it or to interpret the answer is missing.

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 tool takes zero parameters, so the baseline of 4 applies; the description correctly implies no input is needed by framing the call as 'call it first' to discover which pharmacy the user is talking about. No parameter meaning is missing because there are none.

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 concrete verb+resource (identify the pharmacy and return its identity plus twelve full months of key figures), enumerating the exact payload: name, address, phone, identifiers, software-change date, stock value at PAMP, revenue, margin in euros and percent, and data quality. It also fixes its position against the siblings ('À appeler en premier') and excludes medical/patient/market-price territory, so an agent can separate it from bilan_mensuel, ventes_mois or chercher_produit 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?

Usage is explicit: call it first, and use it whenever a pharmacist talks about their pharmacy, sales, stock or margin, to establish which pharmacy and at what date figures were measured. Negative guidance is also explicit (not for a medical question, dosage, patient record or market price). It stops short of naming the sibling tools to redirect to, so routing beyond the exclusion list 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.