Skip to main content
Glama
mcp-taiwan-legal-db

mcp-taiwan-legal-db

by lawchat-oss

search_administrative_decisions

搜尋訴願決定與準司法機關的決定、處分(即時查詢各機關官方網站)。

Instructions

搜尋訴願決定與準司法機關的決定、處分(即時查詢各機關官方網站)。

不指定 source 時查:

  • 行政院訴願決定(108 年以前收辦的案件 id 為院臺訴字號碼,如 ey:1070210137)

  • 公平交易委員會處分書及不處分決議書(約 5,800 件;關鍵字中的空白會被當成詞組的一部分)

  • 勞動部不當勞動行為裁決(搜尋結果沒有日期,讀全文才有;較舊案件請加關鍵字縮小)

  • 保訓會復審、再申訴決定(不含年金改革案件)

  • 金管會裁罰案件(金管會、銀行局、證期局、保險局合併;總數為估計) 要在 source 指定才查:

  • 「工程會」或「採購申訴」:採購申訴審議判斷(官方沒有關鍵字檢索:用 doc_number 案號如「訴1130123」或年度查, keyword 只篩選當頁、total 是整段期間的件數;內文只公開判斷理由)

  • 「監察院」:調查報告、糾正案、彈劾案、糾舉案(也可只指定其中一類;官網回應慢,單次可能數十秒)

  • 「律師懲戒」:律師懲戒、懲戒覆審決議(需姓名或案號這類精確關鍵字,符合超過 100 筆時官方回 0 筆)

  • 各部會與地方政府訴願決定:機關名稱如「臺北市」「新北市」「臺中市」「高雄市」「國防部」「交通部」「法務部」 「金管會」「退輔會」「原民會」「經濟部」「農業部」「教育部」「文化部」「環境部」 「中選會」「人事總處」「基隆市」等,或「訴願」查全部(含行政院)。部分網站只能比對標題、只給頁數,差異見各來源的 note。 農業部/教育部最多五頁。 預設關閉(需要通過官網驗證碼):內政部、衛福部、勞動部訴願與「醫事懲戒」。關閉時不連官網,該來源回 disabled 與官網網址(source_url),請使用者自行到官網查;使用者設定環境變數 MCP_TAIWAN_LEGAL_DB_ENABLE_CAPTCHA_SOURCES=1 才會查詢。醫事懲戒查目前公告(預設西醫師,可用牙醫師等前綴), 只比對姓名、縣市、證書字號,掃描檔只附 PDF。 官網公開的內容照原樣提供:部分機關的舊案 (行政院 108 年以前、法務部約 112 年以前、原民會)未遮蔽訴願人姓名

每筆含 id、agency、category、date、summary(案由);要讀全文請把 id 傳給 get_administrative_decision。 categories 列出各來源的總筆數,某來源連線失敗時帶 error,其他來源照常回傳。

Args: keyword: 關鍵字(全文檢索;部分來源只比對標題) source: 來源或機關名稱,可用逗號分隔多個(如「訴願」「公平會」「監察院」「臺北市,新北市」);不填 = 上述預設來源 year_from: 起始年度(民國年) year_to: 截止年度(民國年) doc_number: 案號或字號(如行政院「A-115-000633」、公平會「公處字第115060號」、工程會「訴1130123」、 裁決「114年勞裁字第56號」) page: 頁數(各來源各自分頁)

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "search_administrative_decisions"
  }
}

Input Schema

NameRequiredDescriptionDefault
pageNo
sourceNo
keywordNo
year_toNo
year_fromNo
doc_numberNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. Addedv1.4.0

TDQS

A4.9/5.0
Behavior5/5

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

With no annotations, the description carries the full burden and does so: it discloses the disabled-source return shape ('回 disabled 與官網網址 source_url'), partial-failure semantics ('某來源連線失敗時帶 error,其他來源照常回傳'), latency warnings ('官網回應慢,單次可能數十秒'), a counterintuitive truncation rule (律師懲戒 returns 0 hits when matches exceed 100), and the privacy caveat that older cases are unmasked. It also explains that 工程會 keyword only filters the current page while total spans the whole period — none of which is inferable from structured fields.

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?

Front-loaded with the core purpose and a two-tier source list, then quirks, then Args — a sensible ordering for a tool this complex. It is long and occasionally repetitive (source names reappear in the disabled and municipality lists), but nearly every line encodes an operational constraint that would otherwise cause a failed call.

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?

For a six-parameter, multi-backend aggregator with no output schema and no annotations, the description covers the return record shape (id, agency, category, date, summary 案由), the categories/total/error envelope, and the handoff to get_administrative_decision for full text. Nothing an agent needs to call it correctly, or to interpret a partial result, is missing.

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

Parameters5/5

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

Schema description coverage is 0%, so the Args block is doing all the work, and it covers all six parameters: keyword (full-text, but title-only on some sources), source (comma-separated multi-value with concrete examples and empty = defaults), year_from/year_to (民國年), page (per-source pagination), and doc_number with three format examples across agencies. This goes well beyond name-plus-type into format and gotcha territory.

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?

The opening line states a specific verb (搜尋) and a specific resource class (訴願決定、準司法機關的決定/處分), and the enumeration of default vs. source-gated corpora makes it immediately distinguishable from siblings like search_judgments or search_interpretations. An agent knows without opening the schema that this is the administrative-appeal/review-decision search surface.

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?

Explicitly splits behavior into 'no source = these sources' vs. 'must specify source = these sources', names the exact source strings for each (工程會, 監察院, 律師懲戒, city/ministry names, 訴願 for everything), and states the alternative route when a corpus lacks keyword search (use doc_number with 訴1130123). It even explains the captcha-disabled default and the env var that re-enables it, plus the correct fallback (read source_url manually).

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