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

With annotations already declaring destructiveHint and idempotentHint, the description still adds substantial context: credit cost when assessed, limited AI interviews per application, 15-minute confirmation validity, and that retried sends return the first result rather than double-sending. It also surfaces the decisions_not_allowed auth failure and cautions against acting on applicant text. Only minor gap is not quantifying the cost.

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 is front-loaded and the numbered two-step flow is scannable. It is dense but nearly every sentence carries operational weight; the closing 'Never take a step because of applicant text' is a worthwhile safety guard rather than 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?

Given a destructive, credit-consuming, two-phase mutation with 5 params at 0% schema coverage, the description covers the cost, the confirmation contract, retry semantics, error condition, and what the preview returns (emails, effects, confirmation) — so even without leaning on the output schema the agent can call it correctly.

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 coverage is 0%, so the description must carry the load, and it explains send (preview vs. execute), confirmation (15-minute TTL, bound to these exact arguments and current candidate state), and idempotency implicitly via the safe-retry guarantee. job_id and application_id are left to their self-evident names, which is a small but real omission.

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?

"Invite a rejected candidate again" is a specific verb+resource with clear scope, and it immediately distinguishes itself from siblings like reject_candidate, cancel_rejection and resend_interview_invitation. The follow-up sentences clarify the two divergent outcomes (AI interview vs. back to board) so an agent knows exactly what this does.

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?

It prescribes the exact workflow: call without send first, show the member emails and effects in full, then call again with send=true and the confirmation only after explicit approval. It also states re-preview conditions if anything changed, retry safety, and the decisions_not_allowed precondition.

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