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

Annotations already declare readOnlyHint=true, idempotentHint=true, destructiveHint=false and openWorldHint=true, so the safety profile is covered. The description's added value is the mandatory two-phase call pattern via decrire_outil, which is real behavioral disclosure beyond the annotations. It says nothing, though, about what a call returns, whether results are paginated, or how the free-form 'arguments' bag behaves — a notable omission for a dispatcher tool with a wildcard object parameter.

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?

Short and front-loaded with purpose, but the second and third sentences are near-verbatim duplicates ('appeler decrire_outil ... avant usage' repeated twice), so one sentence is pure waste. The phrasing 'obtenir son contrat exact' is also ambiguous (likely a literal rendering of 'contract/schema') and the 'nom' vs 'operation' naming mismatch adds noise rather than signalling.

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 thin dispatcher whose real contract lives behind decrire_outil, deferring detail is defensible, and no output schema exists to explain. Still, given 0% parameter coverage and a free-form nested 'arguments' object, an agent finishing this description knows the domain and the pre-step but not what either operation does or what arguments to expect — it must make an extra round-trip before it can even judge relevance against siblings.

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 0%, so the description must carry the burden and it largely does not. It gestures at 'operation' by saying to choose one ('rechercher' / 'lire' loosely map to the enum values), but it never mentions the required 'arguments' object at all — a nested free-form bag with additionalProperties:true that is the hardest part of the contract. Worse, it tells the agent to pass 'nom = opération choisie', while the actual schema field is called 'operation', which can actively mislead.

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 pair and resource: 'Recherche et lecture du droit de l'Union européenne', which distinguishes the legal domain from siblings like explorer_droit_africain or explorer_droit_international. It is clear what the tool covers, but it never names the sibling tools or explains the boundary (e.g. EU vs. international law overlap), so the differentiation is left implicit in the domain word alone.

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?

Gives one explicit procedural rule: pick an operation, then call decrire_outil with its name before use. That is genuine workflow guidance an agent could not infer. However, there is no guidance on when to pick this explorer over the thirteen sibling explorers, and no hint about which of the two operations (rechercher vs. lire) to choose for a given need.

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