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

Annotations already declare readOnlyHint, idempotentHint, openWorldHint and destructiveHint=false, so safety is covered. The description's real contribution is the partial-input behaviour: until all needed answers arrive it returns the remaining questions in order with their options, and only on completion does it return each insurer's 12 and 6 month price plus links. That is genuine behavioural disclosure beyond the annotations, though it omits things like failure handling for invalid postcodes or how many round-trips to expect.

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?

Roughly five sentences, each load-bearing: what it compares and from where, when to use it, the ask-loop protocol, and what a completed call returns. The comparison and data provenance are front-loaded, and there is no filler.

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 38-parameter, multi-turn, no-output-schema tool, the description covers the mechanics an agent actually needs: partial invocation, the question loop, and the shape of a completed result. The remaining hole is relational — the existence and role of get_green_slip_result is never addressed, leaving tool selection under-specified.

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 38 parameters at 100% schema description coverage, the schema does the heavy lifting and the baseline is 3. The description earns an extra point by explaining the parameterisation model itself — you need not supply everything up front, and unanswered parameters come back as ordered questions — which is meaning the schema's zero required fields does not convey.

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?

It names a specific verb and resource ('Compare NSW green slip (CTP) prices of all six NSW insurers') and even identifies the data source ('the regulated prices from SIRA's price check'). What it never does is name or distinguish itself from the sibling get_green_slip_result, so an agent has no textual basis for choosing between them.

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 triggers are concrete and scenario-based: rego renewal due, buying a vehicle, bringing one to NSW, or suspecting overpayment. It also gives an explicit multi-turn protocol — call with partial answers, ask the returned questions, never guess, call again — which is unusually actionable. The one gap is that no when-not condition or alternative tool is mentioned.

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