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

Annotations already declare destructiveHint=true and idempotentHint=false, but the description adds substantial context beyond them: the mutation is scoped to a single refunded cancellation with "no other change," identity is verified via email plus postcode or an EC-prefixed order number, and the change is gated on explicit customer confirmation. It also discloses conversational state handling, which the annotations do not cover.

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?

Purpose and scope are front-loaded in the first sentence, and every subsequent sentence carries operational information. A few passages are dense or repeated in spirit (the AI Brewer's role and confirmation flow are restated), which costs a little tightness but nothing is truly 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 destructive, open-world tool with no output schema and two undocumented parameters, the description covers what an agent needs: preconditions, refund behavior, identity requirement, confirmation gate, conversation chaining, and what the reply contains. Nothing material is left to inference.

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 0% schema description coverage, the description carries the burden and does so for both parameters: message is defined as the customer's own words (free text, not a structured order reference), and conversationId is explained as absent on the first call and returned by each result to continue the same conversation. It omits the message length ceiling and the conversationId format constraint, which the schema enforces but never explains.

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 ("Cancels one of the customer's IMPOSSIBREW orders") plus the governing precondition ("has not left the warehouse") and the refund semantics. The order-vs-subscription distinction is clear enough that an agent can separate it from cancel_subscription without opening either schema.

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 a concrete applicability condition (order must not have shipped) and an enforced interaction protocol (change is applied only after the customer replies "yes"). It stops short of naming a sibling alternative or stating what to do when the order has already shipped, so the routing guidance is clear but not exhaustive.

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