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

Annotations already declare destructiveHint=true and readOnlyHint=false, so the safety profile is covered, but the description adds real value beyond it: other sections and the theme are preserved, a deleted section returns to its former position, and a divergent draft retains changes from other sections. It does not say what happens on a version conflict (error vs overwrite) despite mentioning 'contrôle de version', which is the one behavioral detail an agent most needs for a destructive call.

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?

Five short sentences, each carrying distinct information: scope, preservation, reinsertion, mode behavior, and draft divergence. The critical scoping distinction ('uniquement une section') is front-loaded, but the last sentence about divergent drafts is a bit dense and could be merged with the mode sentence.

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 destructive, five-parameter mutation with no output schema and zero schema descriptions, the description covers preservation guarantees, reinsertion behavior, and mode semantics well. The gaps are the version-mismatch outcome and whether a returned revision/brouillon is produced, which matter for a tool flagged destructive.

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 parameter meaning; it fully explains the only enum ('auto' restores live if published, 'draft' keeps a draft) and implies expectedVersion via 'contrôle de version'. However pageId, sectionId and revisionId are never explained beyond their self-evident names, and expectedVersion's failure semantics remain undefined.

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 ('Restaurer uniquement une section depuis une révision') and adds the scope qualifier 'uniquement une section', which cleanly separates it from the sibling restore_revision that operates on whole pages. It also spells out the reinsertion case ('Réinsère une section supprimée à son ancien emplacement'), so the agent knows exactly what object is being restored.

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 clear contextual guidance: use it when restoring a single section, and the mode paragraph explains when auto vs draft applies (auto live-restores published pages, draft keeps a draft for divergent edits). It never names the sibling restore_revision as the alternative for full-page restores, so the when-not-to-use routing is left implicit.

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