← 返回 Skills 市场
jisngzhongling

蒋道理|CNIPA商标驳回复审证据推理引擎(SJ-IRAC)

cross-platform ✓ 安全检测通过
1674
总下载
1
收藏
0
当前安装
3
版本数
在 OpenClaw 中安装
/install jiang-irac-refusal
功能描述
商标驳回复审推理引擎(SJ-IRAC):面向CNIPA驳回通知的要件化论证、证据链工程与A–E风险闸门,输出审查员可读、可直接提交的复审材料结构。
使用说明 (SKILL.md)

蒋道理|商标驳回复审推理引擎(SJ-IRAC)

SJ-IRAC Trademark Refusal Reasoning Engine

Author: Jiang Zhongling(商标蒋道理)
Organization: Nantong Zhongnan Quansheng IP Co., Ltd.
Version: 1.3.0
Last Updated: 2026-02-03


What This Skill Does

本系统面向 CNIPA 商标驳回通知书 / 驳回复审,将“驳回理由 + 引证商标信息 + 事实与证据”转化为:

  • 要件化、条款精准的复审论证(Rule-bound)
  • 可抗反问的证据链工程(SJ-6)
  • 止损优先的风险闸门决策(A–E + Kill-Gates)
  • 可直接落地的官方文书结构(Document Mode)

定位:个人律所级驳回复审智能中台
不生成套话,不堆字,不虚构事实与时间线。只输出可核验、可提交、可复用的论证结构与证据清单。


Typical Refusal Problems This Engine Solves

  1. 近似(Art.30):标识对比写得“像作文”,缺少要件化对比与证据支撑
  2. 缺乏显著性(Art.11):没把“描述性/行业通用/宣传语”与证据路径拆开
  3. 禁用条款(Art.10):忽视政策风险与审查口径,导致“硬碰硬”无效投入
  4. 在先权利(Art.32):时效/权利基础/证据链薄弱,容易被一句话击穿
  5. 引证商标攻防:不会选“打/绕/拆/分流”的最优路线,导致成本倒挂

Scope & Positioning

Primary Scope

  • CNIPA 驳回复审(以 Art.30 / Art.11 / Art.10 为主)
  • 引证商标相关攻防(必要时联动 Art.4 / Art.44 的“秩序/恶意”路径)
  • 在先权利抗辩(Art.32,须严格时效与证据链)

Not a Template Pack

这是推理与证据工程引擎,不是“模板拼装器”。


Legal Basis (Bounded Sources)

仅在以下边界内运行:

  • 《中华人民共和国商标法》(2019)
  • 实施条例
  • 《商标审查审理指南》及相关审查口径
  • 最新《类似商品和服务区分表》(以你维护的版本为准)

禁止:

  • 虚构法条、案例、交易、截图、时间线
  • 用“常识/感觉”代替证据

Core Framework (Mandatory)

1) IRAC(审查员可读的要件化结构)

  1. Issue:锁定驳回点(条款、引证商标、指定商品/服务、争点范围)
  2. Rule:条款目的 + 构成要件 + 举证标准/审查口径
  3. Application:逐要件对应(对比表 + 时间轴 + 证据目的绑定)
  4. Conclusion:明确请求与可执行动作(补证/改路/止损)

2) SJ-6 证据链(每份证据都要过关)

  1. 真实性
  2. 关联性
  3. 完整性
  4. 时间效力
  5. 逻辑一致性
  6. 抗反问能力

输出必须标注:证据编号 → 来源 → 日期 → 证明目的 → 对应要件 → 弱点与补强建议

3) 风险模块(A–E + Kill-Gates)

  • A:路线正确 + 证据强 + 程序风险低
  • B:路线稳健 + 少量缺口可补
  • C:理由可打但证据偏弱,结果敏感
  • D:高度裁量/证据薄弱/成本倒挂
  • E:致命缺陷或多闸门触发 → 建议止损/改路

Kill-Gates(止损闸门)示例

  • 资格/时效硬伤导致主路线不可用
  • 证据不可核验或无法抗反问
  • 近似对比缺乏客观支撑,仅剩主观叙述
  • 仅能依赖审查员自由裁量,客观指标不足
  • 预计收益 \x3C 复审成本(含时间与机会成本)

Supported Scenarios (Refusal Review)

  • Art.30:近似商标 / 近似商品服务 → 对比表 + 混淆风险要件化
  • Art.11:缺乏显著性 → “描述性/通用性/宣传性”分型 + 证据路径
  • Art.10:禁用条款 → 政策风险评估 + 合规替代命名建议(如需)
  • Art.32:在先权利 → 权利基础(字号/著作权/姓名等)+ 五年时效提示 + 证据链
  • 引证商标策略:打引证(撤三/无效)/绕开(限定商品服务)/分案/替换标识

Input Requirements (Minimum Viable Packet)

至少提供:

  1. 驳回通知书(或关键段落截图/文本)
  2. 申请商标号、类别、指定商品/服务
  3. 引证商标号(如有)、类别、核准商品/服务、状态
  4. 时间节点:申请/驳回/复审期限(如能提供)
  5. 现有证据清单:来源、日期、形式、简述(可后补)

输入不完整 → 默认保守输出(先止损评估,再谈写作)。


Output Modes

Quick Mode(快速研判)

  • 条款定位 + 路线选择(1主+1备)
  • 证据最小集清单
  • Go / Cautious / Stop-loss

Pro Mode(IRAC + SJ-6 + 风险)

  • 完整 IRAC(逐要件)
  • SJ-6 证据链诊断(弱点清单 + 补强优先级)
  • 风险等级 A–E + 触发闸门
  • 行动方案(按 ROI 排序)

Document Mode(可直接提交结构)

  • 官方中立语域
  • 去概率措辞
  • 段落结构适配审查员阅读
  • 证据目录 + 证明目的表 + 时间轴(如提供材料)

Compliance (Hard Constraints)

  • 禁止虚构事实、交易、截图、日期
  • 禁止无证据推断
  • 强制识别:最弱环节 + 最小修复集合
  • 成本倒挂 → 必须建议止损并给替代方案(改名/分案/另案攻击引证)

How to Use

  1. 提供驳回通知书 + 引证商标号 + 商品/服务 + 证据清单
  2. 选择输出模式(Quick / Pro / Document)
  3. 获得:路线选择 → 要件化论证 → 证据链与补强 → 风险闸门 → 文书结构

Versioning Rules

  • Patch (x.y.z):文档/一致性修正
  • Minor (x.y.0):新增模块/流程升级
  • Major (x.0.0):架构级重构
安全使用建议
This skill is instruction-only and internally consistent with its stated purpose, but before using it: (1) Do not submit unredacted confidential or privileged documents — redact PII or sensitive business details if you are not certain about sharing. (2) Treat output as draft legal work product: have a qualified lawyer review any submission-ready material before filing. (3) The skill forbids fabricating facts, but you (the user) must supply accurate evidence and timelines; incomplete input will yield conservative recommendations. (4) Check the proprietary license and repository/homepage for usage terms and data-handling policies if you expect to process client data. (5) If you want to prevent autonomous use, consider disabling agent auto-invocation in your environment (the platform default allows invocation).
功能分析
Type: OpenClaw Skill Name: jiang-irac-refusal Version: 1.3.0 The skill bundle is benign. All files (SKILL.md, core.txt, risk-engine.txt, workflow.txt) consistently provide highly detailed and restrictive instructions to the AI agent, emphasizing legal accuracy, evidence-based reasoning, and ethical constraints. Repeated '禁止' (prohibit) and '强制' (mandatory) rules explicitly prevent fabrication of facts, speculation, or unauthorized actions. The instructions are designed to ensure the agent operates strictly within its stated purpose of providing law-firm-grade trademark refusal review, including stop-loss recommendations if costs outweigh expected value. There is no evidence of data exfiltration, malicious execution, persistence, obfuscation, or prompt injection attempts to subvert the agent's intended function.
能力评估
Purpose & Capability
The skill name/description (CNIPA refusal-review IRAC engine) matches the SKILL.md and auxiliary files: all content focuses on legal analysis, evidence mapping, and output modes for filings. It requires no binaries, env vars, or installs — which is proportionate for an instruction-only analysis engine.
Instruction Scope
SKILL.md confines runtime behavior to transforming user-provided refusal notices, trademark identifiers, and evidence into IRAC analyses, SJ-6 evidence audits, and risk decisions. It explicitly forbids fabricating facts. Note: it relies entirely on user-supplied documents (which may be sensitive); the instructions do not ask the agent to read unrelated system files or credentials, but users should avoid submitting unredacted confidential material and should have a human attorney verify any submission-ready output.
Install Mechanism
No install spec or code files are present — lowest-risk model: nothing is downloaded or written by the skill itself.
Credentials
The skill declares no required environment variables, credentials, or config paths. The inputs it asks for (refusal notice, trademark numbers, evidence list) are appropriate and directly related to its legal analysis purpose.
Persistence & Privilege
always:false and no install/actions that modify agent configuration. The skill can be invoked by the agent (normal default) but has no privileged or persistent system presence and requests no cross-skill configuration.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install jiang-irac-refusal
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /jiang-irac-refusal 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.3.0
**Summary: Major rewrite with new bilingual skill metadata and fully restructured documentation.** - Added bilingual YAML frontmatter with detailed skill metadata, tags, and summary. - Completely rewrote and expanded documentation with clearer structure and advanced legal reasoning guides. - Introduced new sections: Typical Refusal Problems, Legal Basis, Core Framework, and Versioning Rules. - Emphasized strict evidence requirements, risk assessment methodology, and stop-loss decision gates. - Shifted to examiner-centric, evidence-driven output principles; clarified non-template positioning.
v1.2.1
- Unified version numbering across all modules. - Corrected internal documentation inconsistencies. - Minor structural and wording optimizations.
v1.2.0
- Integrated SJ-IRAC reasoning framework. - Added evidence-chain and risk assessment modules. - Optimized for CNIPA refusal review practice.
元数据
Slug jiang-irac-refusal
版本 1.3.0
许可证
累计安装 0
当前安装数 0
历史版本数 3
常见问题

蒋道理|CNIPA商标驳回复审证据推理引擎(SJ-IRAC) 是什么?

商标驳回复审推理引擎(SJ-IRAC):面向CNIPA驳回通知的要件化论证、证据链工程与A–E风险闸门,输出审查员可读、可直接提交的复审材料结构。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 1674 次。

如何安装 蒋道理|CNIPA商标驳回复审证据推理引擎(SJ-IRAC)?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install jiang-irac-refusal」即可一键安装,无需额外配置。

蒋道理|CNIPA商标驳回复审证据推理引擎(SJ-IRAC) 是免费的吗?

是的,蒋道理|CNIPA商标驳回复审证据推理引擎(SJ-IRAC) 完全免费(开源免费),可自由下载、安装和使用。

蒋道理|CNIPA商标驳回复审证据推理引擎(SJ-IRAC) 支持哪些平台?

蒋道理|CNIPA商标驳回复审证据推理引擎(SJ-IRAC) 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 蒋道理|CNIPA商标驳回复审证据推理引擎(SJ-IRAC)?

由 Jiang Zhongling|SJ-IRAC Trademark Systems(@jisngzhongling)开发并维护,当前版本 v1.3.0。

💬 留言讨论