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

The description asserts "Classification: read-only" while the annotations declare readOnlyHint=false, a direct contradiction that could lead an agent to treat a mutating cancel as a safe read. Although useful details follow (completed_items_not_rolled_back partial-failure policy, deterministic idempotencyKey derivation), they sit alongside mutation-mode boilerplate that further muddies whether this call changes state.

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?

Only the first two words address the tool itself; the rest is a generic response-formatting policy (assistantSummary, never display raw JSON, mention IDs sparingly) plus an execution/metadata preamble that would be identical across the suite. It is front-loaded on paper but wastes most of its length on harness boilerplate rather than selection-relevant content.

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?

An output schema exists, so return values need no explanation, but the description still omits everything an agent needs about the operation itself: what a cancelled task becomes, whether cancellation is reversible, what task states permit it, and how it differs from its sibling retry_task. For a mutation with 18 parameters at 0% schema coverage, this is a significant shortfall.

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 0% across 18 parameters, so the description must carry the load. It adds real semantics for only three: idempotencyKey derivation on omission, and the seatId/seatName requirement on mutations when multiple seats are authorised. The other 15 parameters — reason, accountId, locales, marketContext, languageVariantGroupId, and so on — are neither explained nor mapped to the cancel operation.

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 opens with "cancel task." — a verbatim restatement of the tool name and title — and never explains what cancelling actually does (abort a running job vs. mark it cancelled vs. discard queued work). The remaining text is harness metadata (classification, execution, mutation modes) rather than a statement of purpose, and it does not distinguish this from siblings such as retry_task, delete_flight, or archive_campaign.

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?

There is no guidance on when to cancel a task versus retrying it (retry_task), listing it (get_task/list_tasks), or leaving it alone; no prerequisites such as required task state are stated. The only conditional instruction is parameter-level ("when more than one seat is authorised, seatId or seatName must be supplied"), which is not tool-selection guidance.

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