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

The annotations expose only readOnlyHint=false, destructiveHint=false and openWorldHint=false; the description explains why the call is not pure read-only by disclosing that it consumes one token per call, creates an XBOSS task on the provider side, and that the institution's original data is never mutated. It also spells out the full idempotency contract (generate before the first request, reuse only to recover a lost response, use a new key for a new execution, and keep any provider-specific key separate), which goes well beyond anything the annotations convey. The internal tension between '조회만 하고' and '순수 읽기 전용이 아닙니다' is resolvable in context, namely that it does not modify the report but still has side effects, and it does not contradict readOnlyHint=false.

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 prerequisite is front-loaded, which is good, but the body mixes languages and layers together the purpose, the idempotency procedure, the token cost, and the side-effect disclosure, with the idempotency block occupying more space than the core operation. The closing sentence about provider data changes following per-tool execution and existing confirmation procedures is generic boilerplate that does not earn 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?

For a single-required-parameter lookup with 100% schema coverage and no output schema, the description covers the essentials an agent needs to call it: authentication prerequisite, per-call token cost, side effects, and idempotency contract. It stops short of describing the shape of the returned candidate list and never situates this call within the candidate→prepare→submit deletion-request flow.

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 100%, so the baseline is 3; however, the description adds real semantic value for request_idempotency_key, whose schema text only says 'see the tool description', by explaining when to reuse it and when to generate a new one. It does not extend or clarify attrYm, taxCode or businessNo, so the parameter explanation is 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 description names a specific verb and resource: it looks up the list of already-filed withholding tax reports that are eligible for a deletion request ('삭제를 신청할 수 있는 접수된 원천세 신고 목록을 조회합니다'). It is distinguishable in scope from write-oriented siblings, but it never names siblings such as delete_request__prepare or __requirements, so the agent must infer where this sits in the workflow.

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 only explicit usage rule is that it requires a prior hometax__session__login call with the same account_link_id, which is genuinely useful for routing and sequencing. There is no guidance on when to choose this call instead of the prepare/requirements/submit siblings, nor are there any exclusions or preconditions beyond the login requirement.

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