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

Annotations already declare destructiveHint=true and idempotentHint=false, and the description adds substantive behavior beyond that: the exact failure boundary (sending already started), the retention guarantee ('approval and attempt history are kept'), and optimistic-concurrency semantics tied to the read revision with a conflict/retry protocol. Those are precisely the facts an agent needs before invoking a destructive, non-idempotent mutation, and none of them are derivable from the structured fields.

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?

Four short sentences, front-loaded with the action and its scope, then failure condition, retention, and retry protocol in descending order of importance. Nothing is redundant, though 'It is tied to the revision that was read' is terse to the point of being cryptic without the surrounding conflict sentence.

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 non-idempotent, destructive mutation with a nested optional parameter and no output schema, the description covers the critical unknowns: what it refuses to do, what state survives, and how to recover from a concurrency conflict. It stops short of stating what the call returns or what permissions/roles are required, which leaves a small but real gap for a destructive operation.

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 reported at 0% for a two-parameter tool with a nested optional body, so the description carries real weight here. It does clarify intent indirectly — 'an approval still being prepared' maps to the optional preparation binding, and 'tied to the revision that was read' hints at conflict handling — but it never explains the messageId versus preparation_request_id split or when the body should be supplied at all.

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 names a specific verb and resource (cancels one approved message / an approval still being prepared) and bounds the scope temporally ('before it is sent'), which an agent can act on directly. It does not, however, distinguish itself by name from the crowded set of adjacent siblings (denyMessage, skipMessage, returnQueuedMessageToReview, bulkDenyMessages, bulkSkipMessages), so the agent must infer the boundary from the 'queued/sending' framing alone.

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?

Usage context is clear and includes a genuine exclusion: 'Fails once sending has started' tells the agent when this tool will not work. It also warns that a conflict requires re-reading the revision before retrying. What is missing is explicit routing among alternatives — nothing says to use denyMessage or returnQueuedMessageToReview instead when the message is still awaiting approval.

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