prior-art-search

Search patents and academic papers for prior art and generate overlap risk reports.

Updated May 29, 2026
One-click install
npx skills add https://github.com/Mang30/myskills --skill prior-art-search-mang30
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: prior-art-search
Source: https://github.com/Mang30/myskills/tree/main/skills/prior-art-search
Command: npx skills add https://github.com/Mang30/myskills --skill prior-art-search-mang30

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It helps users quickly identify prior art in patents and academic literature to support novelty checks, patentability evaluation, and early freedom-to-operate thinking before filing or drafting claims.

Core Features & Use Cases

  • Concept extraction from an invention description: Distills an invention’s core concept, technical problem, key features, and likely IPC/CPC classes to guide searching.
  • Patent searching across major databases: Searches Google Patents and Espacenet, then fetches and extracts key patent details (title, abstract, representative claims, assignee, status, IPC/CPC).
  • Academic literature searching for supporting context: Searches scholarly sources (e.g., Google Scholar, arXiv, Semantic Scholar) to find relevant papers and extract their contributions.
  • Overlap-risk analysis and structured reporting: Rates references as HIGH/MEDIUM/LOW overlap risk and produces a consolidated report output for review.

Quick Start

Use the prior-art-search skill with your invention description to generate a structured report of patent and academic references organized by likely IPC/CPC classes, including preliminary overlap risk and FTO-style flags.

Frequently Asked Questions about prior-art-search

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I find prior art for an invention disclosure before filing a patent?

Prior art search extracts key technical features and predicted IPC/CPC classes from your invention disclosure, then queries Google Patents, Espacenet, and scholarly sources to identify relevant patents and papers for novelty checks.

Can I do a freedom to operate assessment using academic literature and patents?

Yes, preliminary freedom-to-operate assessment is supported by searching patents and academic papers, extracting reference metadata, and rating overlap risk to flag potential FTO concerns without fabricating citations.

What is the best way to analyze patent overlap risk for novelty screening?

The best way to analyze patent overlap risk is to extract representative claims and metadata from searched patents, then rate each reference as HIGH, MEDIUM, or LOW overlap risk against your invention's core technical features.

How does IPC or CPC class prediction work during a patent search?

IPC and CPC class prediction works by distilling the core concept and technical problem from your invention description, then applying those predicted classes to guide targeted querying across major patent databases like Espacenet and Google Patents.

Step by step, how do I run a prior art search from an invention brief?

Provide your invention brief to extract key features and predicted IPC/CPC classes, automatically query Google Patents and Espacenet plus scholarly sources, then generate a consolidated report with reference metadata and overlap risk ratings.

What are the limitations of automated prior art searching for inventive step research?

Automated prior art searching is limited to preliminary novelty and inventive-step research, providing overlap risk ratings and FTO-style flags without legal certainty, meaning results require manual review before filing or drafting claims.