The intellectual property landscape has been completely overrun. Invention patent generation skill: https://github.com/handsomestWei/patent-disclosure-skill… Software copyright generation skill: https://github.com/Fokkyp/SoftwareCopyright-Skill… Design patents/utility models are even easier: Claude Design / image2

handsomestWei/patent-disclosure-skill

Source: https://github.com/handsomestWei/patent-disclosure-skill

China Patent Skill

From project documentation to deliverable technical disclosures: Mining patentable points, prioritizing searches on the CNIPA Publication Bulletin, desensitized drafting, and a self-check closed loop.

License: MIT

Python 3.9+ (https://www.python.org/) Node.js (https://nodejs.org/) AgentSkills (https://agentskills.io)

You have design docs and code, but haven’t sorted out the patentable points yet? The disclosure needs system block diagrams, flowcharts, and Word documents that agents can edit directly? After finalizing, you still need multiple rounds of supplementary materials and error corrections, and you want traceability for file modifications? Searching the CNIPA publication site, you expect reliable crawling and precise retrieval?

This Skill orchestrates the entire workflow according to AgentSkills conventions, with SKILL.md + prompts/ being step-by-step, readable, and iterable.

Features · Installation · Usage · Project Structure · Examples · Results · Reference Docs · Detailed Installation Instructions · Skill Entry

Features

Office Extraction: .docx / .pptx are first converted to Markdown using this repo’s docx_to_md.py / pptx_to_md.py before scanning (see SKILL.md).

Python Dependencies (by file):

• Basic (Office / Disclosure Conversion): Root requirements.txt — pip install -r requirements.txt
• Prior Art Search (CNIPA Bulletin, Optional): tools/requirements-cnipa.txt — pip install -r tools/requirements-cnipa.txt, then run python -m playwright install chromium. Installation is optional: Step 5 will fall back to using only WebSearch per prior_art_search.md. See INSTALL.md, tools/README.md for details.

Installation

Claude Code

Please place this directory in the git repository root or global skills path, ensuring SKILL.md is at the root level of the skill folder (consistent with INSTALL.md).

# Example: Install to the current project's skills directory
mkdir -p .claude/skills
git clone <This Repo URL> .claude/skills/patent-disclosure-skill

Cursor

Place the complete contents of this repo into the skills path specified by Cursor (see table in INSTALL.md), restart, and confirm the skill has been discovered in Settings → Rules.

Dependencies

# Basic (Office conversion, disclosure-related Python packages)
pip install -r requirements.txt

# Optional: CNIPA Prior Art Search (epub.cnipa.gov.cn)
pip install -r tools/requirements-cnipa.txt
python -m playwright install chromium

Diagram rendering also requires Node.js; execute npm install under tools/ or use npx mmdc (see tools/README.md for details).

Usage

Describe your requirements in natural language within the Agent, for example:

• Patent mining, patentable points, technical disclosure, prior art search, prior art comparison
• Slash commands (depending on host configuration): e.g., /patent-disclosure-skill, /disclosure

It is recommended to also specify the project path or technical topic (consistent with argument-hint in SKILL.md).

Prior Art Search (Step 5) will first search Chinese patent publication information via the China Patent Publication Bulletin (http://epub.cnipa.gov.cn/), then supplement with other sources as needed; see prompts/prior_art_search.md for the workflow.

When adding materials or correcting errors on an existing disclosure file, there is no need to say “iterate”—the skill will handle it via merger.md / correction_handler.md; details in SKILL.md.

Project Structure

This repo follows AgentSkills (https://agentskills.io), where the root directory is a single skill:

patent-disclosure-skill/
├── SKILL.md # Entry: Trigger conditions, tool table, steps, and prompt references
├── prompts/ # Step-by-step templates (followed after Agent Read)
│   ├── intake.md
│   ├── project_scan.md
│   ├── patent_points_analyzer.md
│   ├── prior_art_search.md
│   ├── disclosure_preview.md
│   ├── disclosure_builder.md
│   ├── disclosure_self_check.md
│   ├── iteration_context.md
│   ├── merger.md
│   ├── correction_handler.md
│   └── template_reference.md
├── tools/ # mermaid_render, md_to_docx, docx_to_md, pptx_to_md; CNIPA cnipa_epub_* (prior art); iteration_dialog_log, etc.
├── docs/ # PRD, repo structure description, result screenshots (result-*.jpg)
├── examples/ # Raw material examples (e.g., example_batch_job_scheduler/knowledge/)
├── outputs/ # User outputs, entire directory is .gitignored
├── requirements.txt
├── LICENSE
├── INSTALL.md
└── .gitignore

Examples

Fictitious scanning raw materials can be found in examples/README.md (e.g., examples/example_batch_job_scheduler/knowledge/).

Complete outputs such as patentable points, prior art notes, and disclosures are generated locally to outputs/{CaseID}/.

Results

Initial Generation (First delivery on disk) Initial generation: Timestamped disclosure in the outputs directory, mermaid diagram directories, etc.

Iterative Updates (Delivery after merge/correction, multiple versions coexist + dialogue log) Iterative updates: New timestamped files and disclosure revision dialogue log

Reference Docs

• Skill Entry and Agent Workflow (Trigger conditions, prompts/ mapping, tool table)
• Detailed Installation Instructions (Claude Code / Cursor paths)
• Diagram and Conversion Scripts (mermaid / mmdc, Word export, CNIPA epub prior art tool)
• Example Case and Raw Material Description
• Product Workflow and Directory Conventions
• Engineering Structure Description
• Disclosure Template Details

MIT License © handsomestWei (https://github.com/handsomestWei/)