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

Annotations only flag write/destructive/idempotent; the description adds what those mean in practice: a record is created at the broker, a proposal may be emailed on the artisan's behalf, the call is refused without consent so nothing is sent, no data may be invented, and `aDireAuVisiteur` must be echoed verbatim to the visitor. It also discloses replay semantics (same inputs plus same numero_demande → no duplicate), which is exactly the operational meaning of idempotentHint.

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?

Length is substantial but front-loaded with the write marker, purpose, and the hard prerequisite before the numbered procedure. The verbatim consent scripts consume real space but are compliance-critical and cannot be paraphrased; minor redundancy exists between the step-3 wording and the preceding step-2 quote.

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 17-parameter, destructive, consent-gated write with no output schema, the description still supplies everything needed to call it correctly: prerequisite ordering, per-situation required fields, the exact consent gate, and the required handling of the `aDireAuVisiteur` return field. Remaining gaps are peripheral optional fields, not call-blocking ones.

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?

With 41% schema coverage across 17 parameters, the description compensates for the most consequential ones: accord and texte_accord (exact wording gate), numero_demande (dedup key), situation_assurance (must match the estimate call), email's conditional requirement, and the cotisation / mois_echeance pairing for in-force contracts. It says nothing about ville, prenom, code_postal, raison_sociale, chiffre_affaires, metier, or precisions, so roughly half the surface is still undocumented.

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 and resource — it transmits an artisan's décennale insurance request to the broker Partner Construction so an advisor can issue a proposal and call back — and explicitly distinguishes itself from the sibling estimate tool. The 'ÉCRITURE' prefix immediately separates it from read-side siblings like estimer_tarif_decennale and lire_fiche_metier.

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 an explicit ordering constraint ('à n'appeler QU'APRÈS estimer_tarif_decennale'), an explicit no-go condition (no explicit 'oui' → the tool refuses and sends nothing), and conditional data requirements (email mandatory for an in-force contract, optional otherwise). Alternatives and exclusions are both named, leaving nothing to inference.

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