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

Annotations already declare readOnlyHint/openWorldHint=false, and the description adds the two things that are not in structured data: the exact return envelope {_mcp_pdf_base64, _mcp_mime} with the instruction to decode before writing and never display raw, plus the state guarantee 'rien n'est enregistré ni transmis à l'administration'. That is high-value behavior disclosure beyond the annotations.

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?

Front-loads the selection condition, then the differentiator, then the return-shape warning, then the parameter shape — a sensible order with no filler sentences. It is dense and the ordering of the base64 warning ahead of the parameter summary is slightly jumbled, but every sentence carries information an agent needs.

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 no output schema, the description takes on the return contract itself: it names the envelope keys, states the payload is base64 text inside JSON, and warns against displaying it raw — exactly the gap an output schema would normally fill. Combined with the required/optional parameter shape and the preview mode, nothing needed to call this correctly 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?

Schema description coverage is 100%, so the baseline is 3 and the schema carries the parameters. The description still adds real meaning: it decomposes lines[] into (description, quantity, unit_price HT, vat_rate) and clarifies format ('format=preview pour PDF sans XML') in a way that actually resolves the schema's own enum/description mismatch, where the enum is facturx|preview but the field text talks about 'pdf' and 'xml'.

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 precise verb+resource (produire une facture Factur-X, PDF/A-3 + XML EN 16931) and explicitly differentiates itself from every sibling: 'le seul outil de ce serveur qui produise un vrai PDF ; tous les autres documents sortent en HTML'. An agent can pick this over quittance_loyer, devis_travaux, note_frais etc. without opening any 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?

Opens with an explicit trigger condition ('QUAND une facture doit être conforme à la facturation électronique française, réforme e-facture, EN 16931') and quietly excludes the HTML document tools. It does not name a specific alternative sibling to use when compliance is not required, so the when-not side is only implied rather than stated.

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