← Back to Skills Marketplace
jisngzhongling

蒋道理|撤三证据链与风险审核引擎(SJ-IRAC)

cross-platform ⚠ suspicious
1925
Downloads
0
Stars
2
Active Installs
4
Versions
Install in OpenClaw
/install jiang-irac-nonuse-evidence
Description
CNIPA撤三(连续三年不使用)双轨证据引擎:答辩证据链构建 + 质证审计(SJ-6 + IRAC + 风险A–E)。
README (SKILL.md)

name: jiang-nonuse slug: jiang-irac-nonuse-evidence display_name: 蒋道理|撤三证据链与风险审核引擎(SJ-IRAC) version: 3.0.1 description: CNIPA撤三(连续三年不使用)双轨证据引擎:答辩证据链构建 + 质证审计(SJ-6 + IRAC + 风险A–E)。 homepage: https://github.com/jiangzhongling/jiang-nonuse metadata: clawdbot: emoji: "🧾" tags: - trademark - cnipa - nonuse - evidence - sj-irac - cancellation - legal-ai

SJ-IRAC Non-Use Engine(撤三证据推理系统)

Professional Release v2.0.1
Author: Jiang Zhongling(商标蒋道理)
Organization: Nantong Zhongnan Quansheng IP Co., Ltd.
Last Updated: 2026-02-04


1. System Positioning

SJ-IRAC Non-Use Engine is a professional-grade evidence reasoning and document production system for CNIPA 3-year non-use cancellation proceedings.

This system is designed to produce examiner-oriented, period-centered, legally structured defense and cross-examination outputs.

Core Objective:

Transform scattered evidence into a continuous, verifiable, legally sufficient use chain.


2. Problems Addressed

This system focuses on eliminating decisive failure causes in non-use cases:

  • Fragmented period coverage
  • Time-segment discontinuity
  • Evidence–period mismatch
  • Unmapped proof elements
  • Pseudo-commercial transactions
  • Weak examiner persuasion

3. System Architecture

Core Principle:

Period → Evidence → Elements → Logic → Risk → Decision → Document


4. Legal & Evidence Constitution

4.1 Applicable Law

  • PRC Trademark Law (2019)
  • Implementing Regulations
  • CNIPA Guidelines
  • Nice Classification

Trademark use must satisfy:

  • Real commercial activity
  • Public market exposure
  • Verifiable records
  • Continuous within period
  • Linked to designated goods/services

4.2 No-Fabrication Rule

Prohibited:

  • Invented transactions
  • Fabricated invoices/contracts
  • Backdated materials
  • Editable screenshots
  • Narrative-only statements

Each exhibit must show:

Source / Date / Carrier / Proof Purpose


4.3 SJ-6 Evidence Chain

All exhibits are evaluated by:

  1. Authenticity
  2. Relevance
  3. Completeness
  4. Temporal Validity
  5. Logical Consistency
  6. Resistance

Non-qualified items are excluded.


4.4 Core Proof Targets (T1–T6)

Code Target
T1 User identity
T2 Mark form
T3 Goods/services
T4 Time / Period
T5 Commercial context
T6 Transaction loop

Missing targets → mandatory risk escalation.


4.5 Designated Period Control Rule(指定期间控制规则)

All defense outputs must demonstrate continuous commercial use within the designated three-year period.

Mandatory segmentation:

P1: First 12 months
P2: Middle 12 months
P3: Last 12 months

Rules:

  • Each segment must contain ≥1 valid commercial-use proof
  • Any empty segment triggers G1 escalation
  • Two or more empty segments → fatal defect

Zero-gap continuity is the primary examination anchor.


4.6 Period–Element Binding Rule

Each exhibit must be mapped as:

Date → Period Segment → T4 → Goods → Mark → Transaction

Unmapped evidence = invalid.


5. Workflow Architecture (SJ-REVOKE OS)

Phase 0 — Admission

Feasibility + Cost + Period Risk

Phase 1 — Inventory

Registration + Period Segmentation + SJ-6

Phase 2 — Chaining

Period → Transaction → Mark → Goods

Phase 3 — Risk Audit

Period / Subject / Mapping / Authenticity

Phase 4 — Strategy

Reinforce / Supplement / Stop-loss

Phase 5 — Drafting

Period-driven automation

Phase 6 — QC

Continuity + Readability

Phase 7 — Archive

Review + Reuse

No skipping permitted.


6. Risk & Stop-Loss Engine

6.1 Risk Levels

Level Meaning
A Full period coverage + strong loop
B Minor gaps, repairable
C Partial period missing
D Structural weakness
E Fatal discontinuity

6.2 Kill Gates

  • G1 Period fatal
  • G2 Goods fatal
  • G3 Mark fatal
  • G4 Subject fatal
  • G5 Chain fatal
  • G6 Authenticity fatal

6.3 Output

Each case outputs:

Decision + Level + Period Map + Risk Points + Actions


7. Dual-Track System

7.1 Defensive Track(被申请人端)

Goal:

Period continuity + Closed commercial loop

Minimum Chain:

Mark + Goods + Period + Transaction


7.2 Offensive Track(申请人端)

Focus:

Break period continuity and element linkage

Priority:

Time → Subject → Goods → Loop → Authenticity


8. Evidence Index Standard (SJ-6 Mandatory)

8.1 Binding Rule

All submissions must include a Period-Mapped SJ-6 Index.


8.2 Mandatory Fields

  1. ID
  2. Name
  3. Source
  4. Date
  5. Period Segment
  6. Goods
  7. Target
  8. SJ-6
  9. Pages
  10. Risk

8.3 Template

ID Name Date Period Goods Target SJ6 Pages Risk

8.4 Grouping

  • P1 Evidence
  • P2 Evidence
  • P3 Evidence

All mandatory.


9. Standard Deliverables

Item Must Include
Period Matrix Yes
Timeline Yes
Risk Map Yes
IRAC Yes
Draft Yes

No partial submission.


10. Input & Invocation

10.1 Inputs

  • Reg No
  • Class
  • Goods
  • Period
  • Evidence
  • Opponent (opt)

Missing period info → reject.


10.2 Modes

  • build (defense)
  • audit (attack)

11. Non-Suitable Cases

  • Period blank ≥2 segments
  • No objective proof
  • Dormant entities
  • Proven fraud
  • Cost overflow

→ Mandatory stop-loss.


12. Version Governance

12.1 Format

Major.Minor.Patch

12.2 Principles

  • Period first
  • Evidence first
  • Examiner logic first
  • No speculation

12.3 Change Log

v2.0.0

  • Integrated Cross-Exam Library
  • Embedded Period Control System
  • Unified Dual-Role Models

13. Professional Ethics

  • Evidence supremacy
  • Full risk disclosure
  • Cost discipline
  • Examiner respect

No manipulation.


14. Cross-Examination Argument Library(质证模型)

All offensive outputs must reference this library and period defects.

(以下内容保持原版,略)


15. Respondent Defense & Self-Audit Model(答辩+自查)

15.1 Pre-Defense Self-Audit

  • Period Coverage Test (P1–P3)
  • Kill-Gate Test
  • T1–T6 Mapping
  • SJ-6 Scorecard

Any missing segment → Stop-loss review.


15.2 Defense Workflow

  1. Build Period Chain
  2. Repair Weak Segments
  3. Rank Evidence
  4. Risk Label
  5. IRAC Draft

15.3 Defense Output

  • Period-Based Opinion
  • Evidence Index
  • Risk Sheet
  • Supplement Plan

16. Applicant Cross-Examination Model(申请人质证)

16.1 Deconstruction

  • Period gaps
  • Mapping failure
  • Loop rupture
  • Subject defects

16.2 Strategy Matrix

Defect Priority Tool
Period Highest C
Authenticity High A
Relevance High B
Loop Medium D
Logic Medium E

16.3 Challenge Output

  • Period Defect Table
  • Cross-Opinion
  • Question List
  • Rebuttal Plan

End of Document

Usage Guidance
This package is coherent with its stated purpose, but take these precautions before installing/running: (1) Run it in a dedicated environment (virtualenv or isolated VM). (2) Ensure required native binaries are installed (soffice/LibreOffice, pdftoppm, tesseract) and set appropriate timeouts. (3) Provide only evidence directories you intend to process — the tool will copy/link many files into its working dirs and create output artifacts. (4) If you run the web UI, verify it binds to 127.0.0.1 (local only) or enable WEBUI_REQUIRE_AUTH and set WEBUI_AUTH_TOKEN before exposing to any network. (5) Review LICENSE.md and DISCLAIMER.md and inspect utils/logger.py (audit implementation) if you need to confirm that audit/logging is local only. (6) If you have low tolerance for unknown code, consider reviewing the full repo locally or running it on an air‑gapped machine first.
Capability Analysis
Type: OpenClaw Skill Name: jiang-irac-nonuse-evidence Version: 3.0.1 The skill bundle is classified as suspicious due to its extensive use of `subprocess.run` to execute external binaries (`pdftoppm`, `tesseract`, `soffice`) and other Python scripts, as well as the presence of local web and desktop UIs that expose HTTP servers and interact with the file system. While these capabilities are central to the stated purpose of document processing and report generation, they introduce a significant attack surface for potential vulnerabilities (e.g., RCE risks, local privilege escalation, or misconfiguration leading to unauthorized access). There is no clear evidence of intentional malicious behavior such as data exfiltration or backdoors; in fact, the `AGENTS.md` documentation and `utils/logger.py` indicate a focus on security best practices and sensitive data sanitization. However, the inherent risks associated with these powerful system interaction features, even when plausibly justified, warrant a 'suspicious' classification rather than 'benign'.
Capability Assessment
Purpose & Capability
Name/description (撤三证据链/风险审核) aligns with included Python scripts and YAML rules: OCR, PDF/Office parsing, date extraction, rule-based scoring, document generation and a local web/desktop UI. Required packages in requirements.txt match the code usage (pandas, openpyxl, python-docx, pypdf, PyMuPDF, PyYAML). No unrelated cloud credentials or extraneous binaries are requested.
Instruction Scope
SKILL.md describes inputs (reg no, class, goods, period, evidence) and modes (build/audit); code reads user-supplied evidence directories, parses/normalizes files, runs OCR and converters, and outputs documents and validations. The instructions do not ask for unrelated system files or secrets. Note: the tool copies/links user files into a working directory and may read many local files as part of normal operation.
Install Mechanism
There is no automated install spec in the registry (instruction-only), but the package includes full source and an INSTALL.md that expects creating a virtualenv and pip installing requirements.txt. Dependencies are from common Python packages; conversion relies on external binaries (soffice/pdftoppm/tesseract) that the user must install. This is proportionate to the stated functionality but requires native tools.
Credentials
The skill does not request environment variables, credentials, or external tokens in registry metadata. Code does consult an optional NONUSE_DOCX2PDF_TIMEOUT env var (for timeout control), which is reasonable. No evidence of requesting unrelated secrets.
Persistence & Privilege
Registry flags: always=false and model invocation enabled (default). The tool runs as a local process and optionally starts a local HTTP UI. Be aware: the web UI by default appears to run without authentication (WEBUI_REQUIRE_AUTH defaults to False and WEBUI_AUTH_TOKEN empty) — the desktop launcher uses localhost by default, but if the server is started binding a non-local address it could be reachable from the network. No indication it modifies other skills or system agent settings.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install jiang-irac-nonuse-evidence
  3. After installation, invoke the skill by name or use /jiang-irac-nonuse-evidence
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v3.0.1
- Major codebase update: introduced new utilities, agents, and workflow scripts. - Added multiple new files for automation, rule configuration, and supporting utilities. - Removed legacy documentation and core engine data files. - Improved modularity by separating risk, score, and time rules into dedicated YAML files.
v2.0.0
## v2.0.0 — Structural Upgrade ### Added - Integrated SJ-IRAC automation workflow - Linked risk engine (G1–G6) with document generation - Embedded cross-examination A–F model - Introduced Excel-driven single source of truth ### Improved - Enhanced use-period recognition logic - Strengthened T1–T6 mapping in defense drafts - Improved examiner-oriented structure - Upgraded evidence admissibility checks ### Fixed - Reduced misclassification of cross-period evidence - Fixed date parsing inconsistencies - Improved placeholder replacement stability ### Notes - This version focuses on judicial-grade reasoning consistency. - All outputs require professional verification before submission.
v1.0.1
Version 1.0.1 Changelog - Updated metadata fields: added `name`, `description`, `homepage`, and structured tags for improved discoverability. - Streamlined and condensed documentation for greater clarity and readability. - No functional or file-level changes detected; content refactored without altering logic or features. - Examples and instructions reformulated in both English and Chinese for user accessibility. - Maintained strict compliance and original evidence-driven logic.
v1.0.0
v1.0.0 - Initial release of the SJ-IRAC Trademark Non-Use (3-Year) Evidence Engine. - Supports both defense (evidence chain construction) and attack (cross-examination auditing) for CNIPA non-use cancellation proceedings. - Implements mandatory IRAC structure and SJ-6 evidence chain logic for rigorous evidence assessment. - Provides risk ratings (A–E) and identifies weakest points and stop-loss gates. - Ensures fully compliant, fact-based outputs with no fabricated or speculative content. - Offers multiple output modes: Quick, Pro, and Document.
Metadata
Slug jiang-irac-nonuse-evidence
Version 3.0.1
License
All-time Installs 2
Active Installs 2
Total Versions 4
Frequently Asked Questions

What is 蒋道理|撤三证据链与风险审核引擎(SJ-IRAC)?

CNIPA撤三(连续三年不使用)双轨证据引擎:答辩证据链构建 + 质证审计(SJ-6 + IRAC + 风险A–E)。 It is an AI Agent Skill for Claude Code / OpenClaw, with 1925 downloads so far.

How do I install 蒋道理|撤三证据链与风险审核引擎(SJ-IRAC)?

Run "/install jiang-irac-nonuse-evidence" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is 蒋道理|撤三证据链与风险审核引擎(SJ-IRAC) free?

Yes, 蒋道理|撤三证据链与风险审核引擎(SJ-IRAC) is completely free (open-source). You can download, install and use it at no cost.

Which platforms does 蒋道理|撤三证据链与风险审核引擎(SJ-IRAC) support?

蒋道理|撤三证据链与风险审核引擎(SJ-IRAC) is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created 蒋道理|撤三证据链与风险审核引擎(SJ-IRAC)?

It is built and maintained by Jiang Zhongling|SJ-IRAC Trademark Systems (@jisngzhongling); the current version is v3.0.1.

💬 Comments