국세청 예규 검색

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

Annotations already carry readOnlyHint=true, so the safety profile is covered. The description adds meaningful behavioral context beyond annotations: it names the upstream API (target=ntsCgmExpc), states that only list fields are returned (not full text), and reveals that the result includes ntstDcmId for subsequent full-text lookup. This gives the agent a clear model of the tool's boundaries.

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 a single compact paragraph with the tool name in brackets, a clear verb, and then the essential distinctions. Each sentence adds value: what it searches, what it returns, how to get full text, and how it differs from a related search. It is slightly dense but economical for the information conveyed.

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?

Even though there is no output schema, the description explicitly lists the four return fields, so an agent knows what to expect. It also routes the agent to the sibling tool for the full text, covering the most important follow-up. Pagination and sorting are documented in the schema, so the description is sufficiently complete for a read-only search tool.

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% and each parameter (query, sort, search, explYd, display, page) already has descriptive text. The tool description adds little to parameter semantics themselves, though it contextualizes why ntstDcmId matters. Baseline 3 is appropriate because the schema carries the parameter-documentation burden.

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 uses a specific verb '검색한다' (searches) against a well-defined resource: National Tax Service precedents via the Law Making Information API. It explicitly lists the result fields (안건명, 안건번호, 해석일자, ntstDcmId) and distinguishes itself from the sibling get_tax_interpretation by noting that the full text must be fetched via that tool.

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 states that full text retrieval requires calling get_tax_interpretation with ntstDcmId, which tells the agent when to switch to the sibling. It also compares itself to korean-law's search_decisions(domain='nts'), noting it is the same list but with the ntstDcmId readily available. However, it doesn't explicitly spell out 'use this only for list retrieval' as a when-not condition.

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