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

Annotations already indicate readOnlyHint, idempotentHint, and openWorldHint. The description adds valuable behavioral details: data source (Annuaire santé Ameli, synced weekly), latency (~1 week for new CDS), cache behavior (5 min server-side), and the effect of `include_freshness`. It also explains the output structure (LookupResult discriminated by `found`). This goes well beyond annotations, providing a rich behavioral profile.

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 description is a single paragraph but well-organized with key information front-loaded (purpose, differentiator, output structure). Each sentence adds value—failure conditions, source, alternative tool, aliases. It is slightly dense but avoids fluff. A bit more structuring (e.g., bullet points) could improve, but it is efficient overall, earning a 4.

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?

Given the tool has an output schema and the description provides detailed coverage of inputs, outputs, failure modes, data source, freshness, and relationships to other tools, it is highly complete. The description explains the discriminated union, latency, and when to use alternatives, leaving no obvious gaps. A strong 5.

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% with descriptions for both parameters. The description adds extra context: for `num_finess`, it mentions accepted aliases (numFiness, finess, etab_finess). For `include_freshness`, it explains the opt-in nature, the added field, and the negligible cost. This enhances understanding beyond the schema alone, warranting a 4.

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 clearly states the action: 'Récupère le détail d'un Centre de Santé (CDS) par son numéro FINESS.' It distinguishes from the sibling tool `etablissement_by_finess` by highlighting the specific data fields exposed (carte_vitale, APCV, specialties). This meets the 5-level: specific verb+resource and differentiator from siblings.

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 description explicitly advises when not to use this tool: 'Pour les structures non-CDS, utiliser `etablissement_by_finess`.' It also explains when the lookup may return not found (non-CDS or very recent CDS). While it doesn't compare to all 33 siblings, the guidance provided is clear and actionable, earning a 4.

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