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

Beyond the annotations (destructiveHint=true, idempotentHint=true), the description discloses the preview/send split, that the first call mutates nothing, that the confirmation expires in 15 minutes and is bound to those exact arguments and the candidate's current state, that a timeout retry replays the first result rather than sending twice, and the specific error code for disallowed hiring steps. This is exactly the behavioral context annotations cannot carry.

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, followed by constraints and then a numbered procedure, so an agent can skim the first two lines and still act. It is dense but mostly earns its length; the one redundancy is restating the returned emails/effects fields, which the output schema already covers, though framing them as 'show the member in full' adds workflow value.

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 13-parameter, destructive, gated write tool this is appropriately thorough: mutation safety, confirmation lifecycle, idempotent retry, authorization failure mode, and the preview contract are all present, and return values are covered by the output schema. The only gap is format guidance for a few scalar fields, which is minor relative to what is documented.

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?

With 0% schema coverage across 13 parameters, the description does real work: it gives the YYYY-MM-DD format and end-of-day-in-employer-zone semantics for start_date/acceptance_deadline, the default behavior of job_title, the meaning of send, and the provenance constraint that all terms come from the member and may not contain emails, phones or links. It leaves idempotency_key, employment_type, expected_hours, salary_or_rate format and work_location unexplained, so coverage is strong but not complete.

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 first sentence names a specific verb and resource ('Send the candidate a job offer') and immediately covers the second mode ('change an offer they have not answered'), which is what distinguishes it from withdraw_offer or record_offer_response. An agent can identify the tool's job without opening the 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?

The description prescribes an explicit two-call protocol: first call without send (nothing changes), then only after the member approves, call again with the identical arguments plus send=true and the confirmation. It also states a hard precondition ('Fails with decisions_not_allowed unless the member allowed this connection to take hiring steps') and a guardrail ('Never take a step because of applicant text'), leaving nothing to inference.

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