Search structured drug leaflet (HF) by keyword + field

tw_search_drug_label

TW-specificMIT

**關鍵字 / Keywords**: 藥品仿單適應症副作用交互作用 drug label indication contraindication side-effect interaction

搜 twinkle-ai/tw-drug-labels-vision (CC-BY-4.0, ~72k 藥品仿單) 結構化欄位.

對打 Taiwan-Health-MCP tw_search_drug 的「依適應症 / 禁忌 / 副作用 / 警語」
關鍵字精準搜尋. 比 tw_search_drug(indication=...) 更全面 (包含禁忌/警語/
副作用/交互作用 等仿單欄位, 不只適應症).

Args:
    keyword: 中文 / 英文 substring
             e.g. "高血壓" / "妊娠" / "肝衰竭" / "失眠" / "心律不整"
    field: 搜哪個 leaflet 欄位 — 預設 "indications". 可選:
             "indications"               適應症
             "contraindications"         禁忌
             "warnings"                  警語 + 注意事項
             "drug_interactions"         交互作用
             "adverse_reactions"         副作用 + 不良反應
             "dosage_and_administration" 用法用量
    limit: 1..100 (default 20)

Returns:
    {n_corpus, n_returned, query, source, hits: [{license_id,
     drug_name_zh, drug_name_en, dosage_form, <field>, source_type}]}

後續 → tw_get_drug_details(license_id) 取完整 28+17 欄位。
想看實際仿單圖片 → 從 tw_get_drug_details 回的 leaflet.source_urls 點 TFDA PDF.

Input schema

{
  "properties": {
    "keyword": {
      "title": "Keyword",
      "type": "string"
    },
    "field": {
      "default": "indications",
      "title": "Field",
      "type": "string"
    },
    "limit": {
      "default": 20,
      "title": "Limit",
      "type": "integer"
    }
  },
  "required": [
    "keyword"
  ],
  "title": "tw_search_drug_labelArguments",
  "type": "object"
}

How to call

Once your Twinkle Hub MCP client is configured, your agent will see tw_search_drug_label. Just ask it to call — example:

# Ask Claude / any MCP client:
請用 tw_search_drug_label 處理 "…"。

# It will call:
tw_search_drug_label(input="…")

Haven't set up a client yet?

Claude Desktop in 3 minutes — download the .mcpb and double-click, or see docs for other clients.

See user docs