KitchenGrade

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

Annotations already declare readOnly, idempotent, closed-world and non-destructive, so the bar is lower, yet the description adds real behavior: the catalog covers only four counties and is not a live statewide search. It also explains null handling ('preserves unavailable/historical/expired grades as unknown, without guessing') and provenance claims ('not issued by DBPR', 'not a food-safety guarantee'). What remains unstated is error behavior for a malformed or unknown ID beyond the missing-ID note.

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?

Every sentence carries information: scope, action, caveats, provenance. The structure is slightly back-loaded, since the core action appears in the second sentence after a coverage disclaimer, but no phrasing is redundant.

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 one parameter, rich annotations, and no output schema, the description supplies what an agent needs: the identifier format, whether the data is live or cataloged, how unknowns are handled, and what is returned ('grade and dated inspection basis'). Only invalid-input or not-found error semantics are left uncovered.

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%, so the schema already documents the single establishment_id parameter, its pattern, and its meaning. The description adds only an example value (2010:5816523) and the fact that grades are dated, which is marginal beyond the schema. Baseline 3 applies.

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?

The description names a specific verb and resource: fetching the 'current KitchenGrade grade and dated inspection basis' for an 'exact public DBPR establishment ID'. It also narrowly scopes the data source to four named counties, which sets it apart from a broad search. However, the sibling search_restaurants is never named, so an agent must infer that search is the fallback when no ID is known.

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?

Usage is implied by 'exact public DBPR establishment ID', and the description usefully warns that 'a missing ID is not evidence of closure or no inspections'. It offers no explicit instruction such as 'use search_restaurants first to obtain an ID' and states no conditions under which this tool should not be used.

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

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

Beyond the readOnly/idempotent/openWorld annotations, it discloses that coverage is limited to four counties and is not live or statewide, that zero results do not prove a restaurant does not exist, and that returned grades are independently calculated and dated rather than DBPR grades or safety guarantees. It further explains the semantics of nameSuggestions, filter-recovery options, and the nearRequestedCity proximity qualifier.

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?

The coverage caveat is correctly front-loaded, but the prohibition on placeholder and MCP queries is restated twice in the description and again inside the query schema description, and 'not statewide coverage or a live statewide search' repeats the same idea. Several sentences earn their place, but the definition is noticeably longer than needed.

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-value burden and does so: exact identifiers, dated calculated grades, nameSuggestions versus filter-recovery options, and the proximity flag. For a filtered search tool with a mandatory handoff to get_restaurant_grade, nothing an agent needs to call or interpret results 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 already 80%, and the description adds real meaning on top: city is a strict recorded-city filter, not a nearby radius (Orlando does not match Winter Park), and a city filter cannot expand coverage. The remaining parameters (limit, county, acceptable_only) are left entirely to the schema, which is acceptable given its enum and default documentation.

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 (searches captured Florida DBPR restaurant records) plus a hard scope boundary (Lake, Orange, Osceola and Seminole counties only, not statewide). It also explicitly differentiates from its sibling by telling the agent to use an exact establishment ID with get_restaurant_grade before recommending a restaurant.

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 explicit when-to-use (actual restaurant, address or city searches; 'find Winter Garden restaurants named Spoleto'), explicit when-not (MCP questions, documentation, connection tests, placeholder queries) and names the correct alternative (MCP ping). It also prescribes the preferred calling pattern: restaurant name as query with city as a separate filter.

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