Skip to main content
Glama

verify_citations

Read-onlyIdempotent

Checks whether cited Korean tax rulings, court decisions, and statute articles in a draft actually exist, flagging fabricated or unverifiable references before publication.

Instructions

Check that cited rulings, decisions and statute articles actually exist. 인용 검증 — 초안의 문서번호·조문이 실제로 있는지 확인. 언제: AI나 사람이 쓴 초안을 내보내기 전. 지어낸 문서번호·없는 조문을 걸러낸다. 읽는 형식: 해석 번호(서면-2024-법규소득-2019 등), 판결문 표준 인용('대법원 2002. 9. 24. 선고 2001두489 판결', '88누3123'처럼 두 자리 연도 포함), 조심·국심 번호, 조문 인용의 항·호('제25조 제4항 제2호' — 그 항·호까지 확인). 반환: {문서번호: [{인용, 결과(확인/국세청 DB 미확인/조회 실패), 문서번호, 제목, 일자, 링크, 비슷한 번호}], 조문: [{인용, 결과(확인/조문 없음/항 없음/호 없음/미검증…), 적용 시행일}], 형식 미인식?: [{인용, 결과, 이유}], 요약, 주의}. 번호를 못 읽은 판결 언급·법령명 없는 조문('같은 법 제5조')은 '형식 미인식'으로 세어 요약의 '확인 안 된 것'에 넣는다. '확인'은 존재만 뜻함 — 내용 일치는 get_tax_ruling·law_article 본문으로 확인. 읽기 전용. 조문 확인은 LAW_OC 필요.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "text": "<text>"
    },
    "name": "verify_citations"
  }
}

Input Schema

NameRequiredDescriptionDefault
langNoOutput language. 'en': English keys and labels, official English texts where available (tax treaties, statutes), other Korean texts returned with a host_ai_translate instruction (your AI translates); Solar machine translation only when UPSTAGE_API_KEY or on-prem Solar is configured. 기본 'ko'ko
textYes보고서·의견서·답변 초안 (해석·판례 문서번호와 '법인세법 제52조' 같은 조문 인용이 들어간 글, 최대 2만 자)
as_ofNo조문 존재를 확인할 기준일 YYYYMMDD. 생략하면 오늘

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changedv0.9.1
    • changedInput schema / properties / lang / description
      Previous value: -"Output language. 'en': English keys and labels, official English texts where available (tax treaties, statutes), titles/summaries machine-translated by Upstage Solar when UPSTAGE_API_KEY is set. 기본 'ko'"New value: +"Output language. 'en': English keys and labels, official English texts where available (tax treaties, statutes), other Korean texts returned with a host_ai_translate instruction (your AI translates); Solar machine translation only when UPSTAGE_API_KEY or on-prem Solar is configured. 기본 'ko'"
  2. Addedv0.4.0

TDQS

A4.4/5.0
Behavior3/5

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

Annotations already declare readOnlyHint, idempotentHint, and openWorldHint, so the safety profile is covered. The description still adds real value beyond them: the result buckets (확인/국세청 DB 미확인/조회 실패, 조문 없음/항 없음/호 없음/미검증), the caveat that unparsable citations are counted as '형식 미인식', and the LAW_OC requirement for article checks. It stops short of describing rate limits, latency, or what happens on partial failure.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Labeled sections (언제, 읽는 형식, 반환) front-load purpose first and keep each block doing distinct work. The return-shape enumeration is dense and slightly listy, but nothing is redundant.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With no output schema and a free-form text input, the description supplies the return structure, the meaning of each result value, the '형식 미인식' edge case, the read-only guarantee, and the external-credential prerequisite. An agent has everything needed to call it and interpret the response.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3, but the description goes further by enumerating the citation formats the 'text' parameter must contain (해석 번호, full 판결문 citations, 조심·국심 numbers, 항·호-level article references) and warns that unnumbered case mentions fall into a separate bucket. That is meaning the schema does not carry.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a precise verb and resource — verifying that cited rulings, decisions, and statute articles actually exist — and then narrows the scope further by declaring that '확인' means existence only, with content matching delegated to get_tax_ruling and law_article. An agent can separate this from sibling retrieval tools without reading any schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The '언제' line gives an explicit trigger (before exporting an AI- or human-written draft) and the failure mode it prevents (fabricated document numbers, nonexistent articles). It also names the alternative tools for the adjacent task, so the when-not condition is covered too.

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