skillZs
★ LIVE SKILL TAGS ★
>>> LIVE SKILLS INDEX <<<
* OPEN SOURCE *
NO LOGIN, NO TRACKING
※ REAL INSTALL DATA ※
← back to all skills
xiaojiongqian/skills-hub200 installs

patent-search-cn-us

Patent prior-art search and novelty check guidance focused on China and the United States (CNIPA, USPTO, Google Patents, WIPO). Use when asked to plan or run CN/US patent searches, draft search strategies, or produce a preliminary novelty search report.

How do I install this agent skill?

npx skills add https://github.com/xiaojiongqian/skills-hub --skill patent-search-cn-us
view source ↗

Is this agent skill safe to install?

  • Gen Agent Trust Hubpass

    This skill provides structural guidance, workflows, and list resources to assist an AI agent in performing a preliminary patent prior-art and novelty search across US and Chinese jurisdictions using official public databases. The skill contains only standard Markdown files and references to trusted, official tech and governmental web portals without any executable commands, system alterations, package installation requirements, or remote code endpoints.

  • Socketpass

    No alerts

  • Snykwarn

    Risk: MEDIUM · 1 issue

  • Runlayerpass

    3 files scanned · No issues

What does this agent skill do?

Patent Search CN/US

Overview

Plan and execute a preliminary patent search for CN/US, then summarize novelty risks and gaps with a traceable search log.

Workflow

1) Scope the invention

  • Extract the core technical problem, steps, and outputs.
  • List essential features vs. optional features.
  • Capture synonyms in EN + CN (translate key terms both ways).
  • Define the jurisdiction scope (CN, US, and optionally WIPO/EPO) and time range.

2) Build search queries

  • Start with 3-5 keyword clusters (problem, method, system, output, domain).
  • Add technical terms, abbreviations, and synonyms; include Chinese equivalents when searching CN.
  • If available, map to IPC/CPC classes and include them as filters.
  • Add assignee/inventor filters only when narrowing is needed.

3) Run searches in public databases

  • Use at least one CN database and one US database.
  • Include Google Patents as a cross-jurisdiction sweep.
  • Optionally use WIPO PATENTSCOPE or Espacenet for broader coverage.
  • Reference site list: references/patent-search-sites.md.

4) Triage and record results

  • Scan titles/abstracts; keep candidates that overlap with the essential features.
  • Open claims for the top hits; record claim elements that match or differ.
  • Keep a log with: database, query, filters, date, and top hits.

5) Assess novelty risk

  • For each candidate, mark: full match / partial match / non-match.
  • Identify distinguishing features (potential novelty points).
  • Call out risky overlaps and missing evidence (gaps).

6) Produce a preliminary novelty search report

  • Include: scope, queries, databases, key hits, and novelty assessment.
  • State limitations (public sources only, no paid databases, time constraints).

Output checklist

  • Search scope and assumptions
  • Keyword clusters (EN/CN)
  • Databases used and date
  • Query strings and filters
  • Top results summary (title, year, link)
  • Novelty assessment and risk notes

Resources

  • references/patent-search-sites.md
  • references/open-paper-sites.md

Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.

<a href="https://skillzs.dev/skills/xiaojiongqian/skills-hub/patent-search-cn-us">View patent-search-cn-us on skillZs</a>