中国专利.Skill
v1.8.1通用中国专利挖掘发现与交底书生成全流程：扫描项目文档挖掘专利点、讨论融合、基于脱敏模版生成技术交底书、联网查新、生成后自检含逻辑闭环与公式参数一致性。| Patent mining, disclosure drafting, prior-art search, and consistency self-check.
⭐ 1· 34·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
Capability signals
These labels describe what authority the skill may exercise. They are separate from suspicious or malicious moderation verdicts.
OpenClaw
Benign
high confidencePurpose & Capability
Name/description (patent mining & disclosure drafting) match the declared behavior: scanning project documents, converting Office files, running prior-art web searches, drafting disclosure .md/.docx and performing internal checks. Example code and tools (docx_to_md.py, pptx_to_md.py, md_to_docx.py, mermaid_render.py, iteration_dialog_log.py) are directly relevant to those tasks.
Instruction Scope
SKILL.md precisely instructs the agent to Read prompt templates, scan project files (including converting .docx/.pptx to .md), run local scripts, perform WebSearch for prior art, and write timestamped output files. This scope matches the stated purpose, but it does authorize reading arbitrary files under the provided project path (expected for a scanner). The skill also requires using WebSearch which will send query text to external search endpoints; users should be aware that scanned excerpts used in queries may be transmitted to those services.
Install Mechanism
There is no automatic install spec (instruction-only at registry level). Helper scripts are included in the repo and rely on standard Python packages and optional Node.js tooling (mermaid/mmdc, puppeteer). No downloads from obscure third-party URLs or shorteners are present in the repository metadata. The need to run pip/npm is documented in INSTALL.md and tools/README.md.
Credentials
The skill declares no required environment variables or credentials. It references the common CLAUDE_SKILL_DIR convention (typical for agent environments) but does not request unrelated secrets. The allowed tools (Read, Write, Grep, Glob, WebSearch, Bash) are appropriate for scanning local files and performing prior-art searches.
Persistence & Privilege
The skill is not marked always:true and does not request elevated or persistent platform privileges. It instructs writing files into user-specified output directories and appending a revision log — behavior consistent with its purpose. Autonomous invocation is allowed by default (platform normal) but does not by itself make this skill incoherent.
Assessment
This skill appears to do what it says: it will read your project files (including converting .docx/.pptx to Markdown), perform web searches (Google Scholar / Google Patents via WebSearch) for prior art, render mermaid diagrams (requires Node/mmdc and possibly puppeteer), and write timestamped .md/.docx drafts plus a revision log. Before installing or running:
- Run it only in a workspace that does not contain secrets (API keys, private certificates, internal credentials), because it is designed to scan project documents and may read many files.
- Expect to install optional dependencies if you want .docx generation and mermaid rendering (pip install -r requirements.txt; Node/npm for mmdc/puppeteer). Puppeteer may download a headless browser during npm setup.
- Review outputs and the revision log (交底书修订对话记录.md) in the specified outputs directory; by design the skill writes timestamped files and avoids overwriting unless you explicitly ask it to.
- Be aware that prior-art search uses external web search services; snippets or queries derived from your docs may be sent to those services. If that is a privacy concern, do not include sensitive excerpts or use an isolated environment.
If you want higher assurance, inspect the included tools/ scripts locally (they are plain Python) and run them manually in a controlled environment before letting an agent execute them autonomously.Like a lobster shell, security has layers — review code before you run it.
latestvk970wwewtm7qt6gz5cg70jkmpd84f852
License
MIT-0
Free to use, modify, and redistribute. No attribution required.