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

Annotations already carry readOnlyHint=true, idempotentHint=true, destructiveHint=false, so the bar is lower, and the description adds substantive behavior beyond them: data freshness ('refreshed several times a day'), matching semantics ('best match first, one entry per country'), no-match edge behavior ('an empty list means no match'), and the critical click-through rule ('cashback is only credited when the user clicks through it'). No contradiction with annotations — the description is fully consistent with a read-only, idempotent lookup.

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-loaded with the core purpose, then ordered logically: trigger examples, routing rules, return semantics, and presentation requirements. It is a long description, and the six example queries are somewhat redundant (all illustrate the same trigger pattern), while the final presentation paragraph edges into UX policy rather than invocation guidance. Still, every sentence carries decision-relevant information, so it earns a 4 rather than a 3.

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 substring-matching, multi-portal aggregation tool with an output schema present, nothing an agent needs is missing: when to call it, sibling routing, matching and edge-case behavior, data freshness, and how results must be presented (show every portal's rate in a table with a clickable url, not just the best). The output schema relieves the description of explaining return-value fields, and the 100% param coverage relieves it of param semantics.

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 schema already documents store_name (case-insensitive substring, plain name only) and country_code (lowercase ISO alpha-2, 'gb' not 'uk', default null). The description adds value above that: operational retry guidance tied to each parameter ('retry with a shorter name or without country_code') and result-level semantics for country_code ('one entry per country'), plus query examples that map to parameter usage ('Dell cashback in Germany').

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?

Opens with a specific verb+resource+scope: 'Finds a store by name and returns its current cashback rates from every cashback portal that lists it,' including coverage details (online/in-store, percentage/fixed, 'up to' flags). It immediately differentiates from siblings by declaring itself 'the preferred first call for any cashback question that names a store' and explicitly naming calls NOT to make first (get_stores_by_name, get_stores_by_country, get_countries).

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?

Provides explicit when-to-use with six concrete example queries ('best cashback for Walmart', 'Dell cashback in Germany'), explicit when-not-to ('do not call get_stores_by_name...'), and explicit alternative selection conditions: get_cashback_rates_by_store_id only when store_id is known, get_gift_cards_by_store_name for gift card discounts, get_best_deals_by_brand for brand-level purchase questions. Failure recovery is also specified: empty list means no match — retry with a shorter name or without country_code.

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