Marée à La Réunion

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

Annotations already declare readOnly/idempotent/non-destructive/openWorld, and the description adds substantial context beyond them: data source and cadence (SHOM gauges, one measurement per minute), computation provenance (rundata, not official prediction), accuracy bounds (~15-20 min and 5 cm), geographic validity radius, and degraded behavior beyond that radius. It even instructs the agent to disclose the precision caveat, which is exactly the behavioral disclosure a downstream agent needs.

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?

Information is front-loaded — data source first, computed-times caveat second, geographic limits third — and every clause carries operational meaning. It is nonetheless long and dense for a one-parameter tool, with some redundancy between the French and English halves and embedded response-writing instructions rather than pure tool description.

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?

There is no output schema, so the description carries the return-shape burden, and it does: current level, rising/falling state, last observed high/low, and next highs/lows over two days, or a reduced payload (nearest gauge height plus SHOM link) for distant places. Combined with the stated precision and coverage limits, nothing an agent needs to call or interpret this tool 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 coverage is 100% and the parameter's own description already explains that 'lieu' selects the nearest gauge and that empty means both. The tool description adds meaning the schema does not: any coastal place is accepted, but places beyond 15 km yield only height data without computed times, which changes how an agent should interpret and phrase results for a given 'lieu'.

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 resource and scope: sea level measured each minute by SHOM gauges at Le Port and Sainte-Marie, current level, tide direction, last observed high/low, plus next highs/lows computed over two days. It goes further by distinguishing itself from the SHOM's official prediction ("CALCULÉES par rundata ... pas la prédiction officielle du SHOM"), which tells an agent exactly what class of data it is getting versus other tide sources.

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 concrete trigger phrases ("la marée monte ?", "c'est quand la marée basse demain ?", "niveau de la mer au Port ?") and explicit applicability conditions: computed times are valid only within ~15 km of a gauge, with the qualifying communes listed. It also states the alternative for distant locations — return the nearest gauge height without times plus a link to official SHOM times — so the agent knows both when to use it and what to do outside its validity envelope.

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