reg-citation-normalizer

把 21 CFR、ISO/IEC、GB/YY/T、NMPA 部令公告等医疗器械法规引用解析为规范结构并给出标准写法，可对文档批量检查缺空格、大小写、年份连接符等问题。Python 3.8+ 零依赖离线运行，只做格式识别，不校验条款内容与现行有效性。

zhaoxinghua09-cell

热度 24.20

最近更新 2026-10-17 07:12:21

分类：效率依赖：代码执行 / 读写文件应用场景：校对润色 / 法律文档 / 翻译 / 数据处理

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add zhaoxinghua09-cell/reg-citation-normalizer

Metadata

name: reg-citation-normalizer
title: Reg Citation Normalizer
description: 医疗器械法规引用编号的识别、解析与规范化。将 21 CFR、ISO/IEC、中国国标行标（GB/YY/T）、NMPA 部令公告等各种写法解析为规范结构并给出标准书写建议。零依赖，可离线运行。
description_zh: 医疗器械法规引用编号的识别、解析与规范化：把 21 CFR、ISO/IEC、GB/YY/T、NMPA 部令公告等写法解析为规范结构并给标准书写建议。零依赖离线运行。
description_en: Recognize, parse and normalize medical-device regulatory citations (21 CFR, ISO/IEC, Chinese GB/YY-T, NMPA decrees) into a canonical structure with recommended standard formatting. Zero-dependency, offline.
license: MIT
metadata: {"author":"SynomosAI","category":"other","tags":["medical-device","regulatory","citation","FDA","NMPA","ISO"]}
permalink: internal/vault (neutralized)
version: 1.0.1
category: knowledge-management
classification: {"skillhub":["medical-device","regulatory","productivity"],"iso_25010":["Maintainability","Usability"]}
read_when: ["注册文档审校 / IFU / 技术文件编写中需要规范化法规引用写法","法规翻译中英对照，需保证条款号对齐","批量检查文档中 21 CFR / ISO / GB·YY/T / NMPA 引用书写错误"]
ai_protocol: true
spec: usk/1.0
interface: {"type":"cli","entry_point":"scripts/cite_normalize.py","runtime":"python3","call_pattern":"args"}
input_schema: {"args":{"type":"string","description":"CLI arguments per skill usage","required":false}}
output_schema: {"stdout":"structured result (JSON IR)","exit_code":"rc=0 ok / rc=1 warning / rc=2 bad input"}
capabilities: ["citation","regulatory","normalizer"]
permissions: {"network":false,"filesystem":false,"subprocess":false}
requirements: {"python_packages":[],"min_python":"3.8"}
changelog: 1.0.1: Bilingual rewrite (EN-first) + read_when/classification added; 1.0.0 initial upload (USK v3 injected round2)

凡自治之物 · LGD 治理学 · 创造术家族｜评测协议：skill-quality-gate

English

Reg Citation Normalizer

Recognize and normalize medical-device regulatory citation numbers — turn inconsistent citation styles in your documents into a canonical structure with recommended standard formatting.

When to use

Reviewing registration dossiers, writing IFU/technical files, or checking regulatory-translation consistency (normalizing 21 CFR / ISO·IEC / GB·YY-T / NMPA decree citations).
Triggers: citation normalize, 21 CFR, ISO 13485, YY/T, NMPA.

How to use

python scripts/cite_normalize.py normalize "21 CFR 820.30(a)(6)"
python scripts/cite_normalize.py parse "ISO 13485:2016 clause 7.5.1" --json
python scripts/cite_normalize.py check report.md [--json]

Workflow

Run check on a document to find formatting issues;
Use normalize/parse to get canonical forms;
Batch-fix and re-review.

Supported citation types

Type	Example input	Canonical output
US 21 CFR	`21 CFR 820.30 (a)(6)`	`21 CFR 820.30(a)(6)`
ISO/IEC	`ISO 13485:2016 clause 7.5.1`	`ISO 13485:2016, 7.5.1`
Chinese line standard	`YY/T 0287-2017 7.5.1`	`YY/T 0287-2017, clause 7.5.1`
Decree	`国家市场监督管理总局令第47号`	agency normalized (incl. SFDA→CFDA→NMPA history)
Announcement	`国家药监局公告 2021年第76号`	normalized announcement

Errata (check mode)

Missing space between code and number (ISO13485 → ISO 13485); lowercase code; misplaced §; year separator (: vs dash); undated ISO references → prompt to add year.

Boundary

Format recognition and normalization only; does NOT validate clause content currency (use openstd.samr.gov.cn / ecfr.gov / iso.org).
Rules per ISO/IEC Directives Part 2 §10.4–10.5, GB/T 1.1-2020, GPO/eCFR, NMPA official style.

Requirements

Python 3.8+, zero third-party deps.

Evidence & safety

Evidence: TBD; Version: 1.0.1; Security: no independent scan (TBD).

中文

reg-citation-normalizer

医疗器械法规引用编号识别与规范化工具。把文档里五花八门的法规引用写法，解析成规范结构并给出标准书写建议。

何时用

注册文档审校、IFU/技术文件编写、法规翻译一致性检查时。
触发词：法规引用、21 CFR、ISO 13485、YY/T、NMPA。

用法

python scripts/cite_normalize.py normalize "21 CFR 820.30(a)(6)"
python scripts/cite_normalize.py parse "ISO 13485:2016 clause 7.5.1" --json
python scripts/cite_normalize.py check report.md [--json]

流程

对文档跑 check 发现书写问题；
用 normalize/parse 得到规范写法；
批量回改并复核。

覆盖的引用类型

类型	示例输入	规范化输出
美国 21 CFR	`21 CFR 820.30 (a)(6)`	`21 CFR 820.30(a)(6)`
ISO/IEC	`ISO 13485:2016 clause 7.5.1`	`ISO 13485:2016, 7.5.1`
中国行标	`YY/T 0287-2017 7.5.1`	`YY/T 0287-2017，第7.5.1条`
部令	`国家市场监督管理总局令第47号`	机构归一化（含 SFDA→CFDA→NMPA 沿革）
公告	`国家药监局公告 2021年第76号`	规范公告写法

勘误能力（check 模式）

缺空格、大小写、§ 误用、年份连接符、ISO 未注日期等 9 类常见书写问题。

边界声明

只做格式识别与规范化，不校验条款内容正确性与现行有效性（查现行状态用 openstd.samr.gov.cn / ecfr.gov / iso.org）。

环境要求

Python 3.8+，零第三方依赖。

证据与安全声明

证据：TBD；版本：1.0.1；安全：未做独立扫描（TBD）。