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

Annotations already declare readOnlyHint=true, idempotentHint=true, and destructiveHint=false, covering safety and idempotency. The description adds substantial behavioral context beyond these: it mandates single-call usage with no retries/refinement/broadening, explains fallback behavior ('stays in-country; filters exact'), specifies response type ('Facts only; web only if asked'), and prohibits asking for location details ('Verbatim request; never ask for ZIP/postal'). No contradiction with annotations.

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 description is a single dense paragraph that packs purpose, exclusions, behavioral rules, and guidance into a concise format. Every sentence adds value, and there is no repetition or filler. However, it could be slightly more structured (e.g., breaking into sections for 'what it does', 'exclusions', 'behavioral rules') to improve scanability. Given the amount of information, it remains efficient.

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 tool with 34 parameters, the description covers the main use cases and constraints well. It explains the core functionality, geographic limitation, single-call rule, and fallback behavior. The presence of an output schema reduces the need to describe return values. However, gaps remain: it does not mention pagination (cursor parameter) or the limit parameter, does not explain the fallback mechanism in detail, and does not address sorting or other optional filters. While strong for a typical MCP tool, it is not exhaustive given the tool's complexity.

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 schema description coverage at only 21%, the description must compensate. It names several filter dimensions (place, budget, mileage, make/model, condition, body, fuel, features, named dealer) that map to parameters like location_text, price_min/max, mileage_max/min, makes, models, conditions, bodies, fuels, features, dealer_ids. It clarifies that 'request' should be verbatim and that location_text should be the shopper's free-text input. However, many parameters (drivelines, transmissions, charge_ports, colors, etc.) are unaddressed, and the description does not provide a structured mapping between use cases and parameter names. While it adds meaning, it does not fully compensate for the low schema coverage.

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 specifies a precise verb-resource pair ('shop vehicles') and clearly scopes to live US/CA listings of common vehicle types (cars, SUVs, trucks, EVs, etc.). It enumerates supported query dimensions (place, budget, mileage, make/model, condition, etc.) and explicitly excludes Quebec. It distinguishes from sibling tools like find_dealerships by stating the prerequisite step for named-dealer stock, and lists what the tool is NOT for (repair, reliability, facts, MSRP, out-of-market, etc.), providing strong differentiation.

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?

The description gives explicit 'when to use' guidance: it covers the supported query types and provides an alternative workflow ('Named-dealer stock: find_dealerships then shop'). It includes clear negative guidance ('Not repair/reliability/facts/MSRP/out-of-market...') and behavioral rules ('Call once; never retry/refine/broaden', 'Fallback stays in-country; filters exact'). It also tells the agent to never ask for ZIP/postal, which prevents improper UX.

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