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

Despite the destructiveHint annotation already indicating mutation, the description goes far beyond by detailing side effects (daily cap counting, refusals with specific error codes), idempotency mechanics (same-key retries collapse), and deduplication nuances (what deduplicated:true/false actually means, the 10-minute body-hash window, and the possibility of deduplicated:false on a long-delayed repeat). This is rich behavioral disclosure that prevents misuse. No contradiction with 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 description is long but every sentence provides non-redundant information. It front-loads the core purpose, then logically progresses from send consequences, to refusal conditions, to idempotency, to deduplication semantics. There is no fluff or repetition; the density is justified by the complexity of the tool's behavior. Each clause earns its place.

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?

The tool has no output schema, so the description must fully explain return values, which it does: it explains messageId and deduplicated semantics, error codes, and retryability. It also covers prerequisites (mailbox on thread), side effects (cap increments), and edge cases (provider-side dedup). For a tool with this level of subtlety, the description is comprehensive and leaves no critical ambiguity.

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?

The schema covers threadId and idempotencyKey with descriptions, but body lacks one. The description compensates by explicitly listing 'Inputs: threadId, body' and explaining how body is used (it can be varied to avoid deduplication) and how idempotencyKey works in detail (body hash fallback, need for stable key). It adds meaningful context beyond the schema, especially for idempotencyKey, though it doesn't individually describe each parameter. Given the 67% schema coverage, the description fills the gap well.

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 a clear, specific verb+resource statement: 'Send a reply on an existing thread, from the mailbox that sent it.' This immediately distinguishes the tool from siblings like launch_campaign or ack_message, and the scope is well-defined. It also states the return shape, making the purpose unambiguous.

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 provides explicit guidance on when to use the tool (to reply to an existing thread), conditions that cause refusals, and clear advice on retry behavior (e.g., 'stop retrying and don't loop replies to manufacture volume'). It also contrasts with campaign sends ('governed exactly like campaign sends'), which clarifies its place among alternatives, and notes the 404 condition when no sending mailbox exists.

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