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

Even though the annotations already cover readOnly, idempotent, destructive, and openWorld, the description adds real behavioral context beyond them: it never guesses when a match is ambiguous (CANDIDATES), it never sends a fax, it never auto-fills the form field ('칸에 자동으로 넣지 않는다'), and it never stores input. It also discloses how uncertain results are flagged (fax_check, confirm_before_submit, 확인방법 tiers) and where the data freshness value comes from (data.checked_at).

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 purpose is front-loaded and every clause carries information, but the whole body is one dense run-on paragraph whose core facts are buried in parentheticals deep inside single sentences. Hyper-specific examples such as '여수지청 0503-8803-0638' and '칠곡군 석적읍' make the text harder to scan than it needs to be, and the status and behavior information would be clearer as separate lines.

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?

There is no output schema, so the description must explain the return surface; it does so thoroughly, covering the returned fields (phone, fax, department address, source, 확인방법), all status branches (CONFIRMED, CANDIDATES, NOT_FOUND), the 지방청 직할 special case with its primary and secondary department contacts, the labor_office_value suggestion, and data.checked_at. An agent can interpret any response without extra lookup.

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 100%: both 'address' (including the tip that starting with 시도 improves accuracy) and 'construction' are already documented in the schema. The description does not add parameter-level meaning, such as address formats or what happens with an ambiguous or missing 시도. Baseline 3 applies because the schema is already doing this work.

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 concrete verb and resource: it returns, for a workplace address, the competent 지방고용노동관서 plus the department that receives 산업재해조사표 and that department's phone, fax, and address. It also preempts the most likely confusion by stating '근로복지공단이 아니다.' It never distinguishes itself from sibling tools such as get_accident_report_guide or check_accident_report, which keeps it just short of a 5.

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 usage context is only implied: give an address when you need the jurisdiction office. The description does state scope limits such as '팩스를 보내지 않는다', '직원 성명은 없다', and '입력을 저장하지 않는다', and it explains what each status value means. However, it gives no explicit guidance on when to pick this tool instead of a sibling, and no prerequisites beyond recommending that the address start with 시도.

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