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

Annotations only say readOnly=false, destructive=false, idempotent=false, openWorld=false; the description goes much further by disclosing that nothing is dispatched or emailed, that a local approval plan plus durable receipt is written, that plans expire, and that foreign/stale/unchanged/unapproved/expired plans and revoked grants fail closed. It also names the required OAuth scope, MANAGE_ORDERS permission, and approved owned store, and clarifies that price/stock/import grants do not authorize this action. Nothing contradicts the 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 opening sentence is front-loaded and does the routing work, but the body is an overlong run of clause fragments, and the consent requirement is stated twice with slightly different wording ('Requires OAuth store:orders:tracking:write consent' then 'Requires separate store:orders:tracking:write consent'), which reads as duplicative and could confuse about whether two distinct grants are needed. Several sentences (email acceptance vs delivery, tracking verification vs delivery) are useful but buried in an undifferentiated block.

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 prepare/mutation-style tool with four required parameters, an output schema, and explicit annotations, the description covers authorization, target eligibility, side effects, failure modes, expiry and recovery guidance. An agent has enough to decide eligibility and to know that execution is a separate step, with return-value detail carried by the output schema.

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 description coverage is 100%, so order_id, expected_version, tracking_number and carrier are already well documented, including that expected_version comes from get_order. The description adds only ambient context (exact old/new value review, server-side normalization and native carrier verification), so the baseline of 3 applies rather than credit for new parameter meaning.

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 opening sentences give a specific verb and resource: prepare (not execute) an expiring tracking correction for one exact already-shipped website order. It draws a hard boundary against sibling operations by excluding first shipments, marketplace orders, label purchase, payments, refunds, stock changes, bulk selection and webhook-bearing orders, so an agent can separate it from execute_order_tracking and label/refund tools without opening a 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?

Explicit when-to-use is present (correct a shipped website order's tracking number or carrier), along with a long when-not list of unsupported targets and the instruction to recover the original plan and key after uncertainty rather than resending. The prepare-then-review-then-execute flow is stated, including that native bound review cannot be replaced by conversational approval.

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