agentsclimarketplace

Patent research

Skill Mattakushi432/Claude-Code-Skills-Custom-DevTools-Pack/plugins/devtools-pack/skills/patent-research

A curated pack of custom Claude Code skills for developers — installable as a Claude Code plugin marketplace.

Install
npx -y skills add Mattakushi432/Claude-Code-Skills-Custom-DevTools-Pack --skill patent-research

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

When to activate: patent research, prior art search, freedom to operate, patent landscape, USPTO, EPO, WIPO, patent claims analysis, IP strategy, patent filing

SKILL.md

6.9 KB, ~1.6k tokens by cl100k_base, as published. Nobody here has run it

Patent Research Patterns

Patent Databases

Primary Databases

DatabaseCoverageBest For
USPTO Patent Full-TextUS patents 1976+US-only prior art
Google PatentsGlobal, 120M+ documentsQuick broad search
Espacenet (EPO)130M+ worldwideEuropean & global
WIPO PatentScopePCT applicationsInternational filings
Lens.orgFree, 120M+, exportableResearch & analytics
Derwent InnovationPaid, curatedDeep analytics
PatSnapPaidPortfolio analytics

Database Selection Guide

Quick prior art check → Google Patents
US freedom-to-operate → USPTO + Google Patents
European clearance → Espacenet
International PCT → WIPO PatentScope
Landscape analysis → Lens.org (free) or Derwent (paid)
Academic + patent combined → Lens.org

Search Strategy

Keyword Development

  1. Core concept — describe the invention in plain language
  2. Synonyms — every term the examiner might use
  3. Technical terms — field-specific nomenclature
  4. Functional terms — what it does (not what it is)
Example: Wireless charging for EVs

Core: "wireless charging" + "electric vehicle"
Synonyms: "inductive charging", "contactless charging", "resonant charging"
Technical: "magnetic resonance", "WPT" (wireless power transfer)
Functional: "charging without cable", "charging while parked"

Classification Codes

Use CPC (Cooperative Patent Classification) to find related patents:

Search process:
1. Go to Google Patents
2. Search keyword
3. Note CPC codes on relevant results
4. Filter search by CPC code to find entire class

Common CPC groups:
G06F — Digital computing
H04W — Wireless communication networks
A61B — Diagnostic/surgical instruments
B60L — Electric propulsion for vehicles

Boolean Search Construction

USPTO Full Text Search:
SPEC/(("machine learning" OR "neural network" OR "deep learning") 
AND ("fraud detection" OR "anomaly detection"))
AND APD/20200101->20240101

Google Patents:
"machine learning" "fraud detection" (neural OR deep) before:2024 after:2019
assignee:Google OR assignee:Meta

Espacenet CQL:
ta="machine learning" AND ta="fraud detection" AND pd=[20200101 TO 20241231]

Patent Anatomy

Key Patent Sections

SectionPurposeRead First?
ClaimsLegal scope of protectionYES — defines what's protected
Abstract150-word summaryFor quick triage
Description/SpecificationHow invention worksAfter claims
DrawingsVisual explanationWith specification
Prior ArtWhat they citedFind related art

Reading Claims

Independent claim (broadest):
"A method comprising:
  receiving, by a processor, input data;
  applying, by the processor, a neural network to the input data;
  generating an output classification."

Dependent claim (narrower):
"The method of claim 1, wherein the neural network comprises
  at least three hidden layers."

Claim elements to note:
- Preamble: "A method" / "A system" / "A device"
- Transition: "comprising" (open, can add elements) vs
              "consisting of" (closed, no additional elements)
- Body: each limitation narrows the claim

Freedom-to-Operate (FTO) Analysis

FTO Process

1. Define the product/process being cleared
   - Document each technical feature
   - Note materials, methods, configurations

2. Identify relevant patents
   - Search by keyword + classification
   - Search by assignee (competitors)
   - Search by inventor

3. Screen for live patents
   - Check legal status (expired, active, abandoned)
   - US patents expire 20 years from filing
   - Maintenance fees required (3.5, 7.5, 11.5 years)

4. Claim chart analysis
   - Map each claim element to your product
   - Literal infringement: element-by-element match
   - Doctrine of equivalents: substantially same function/way/result

5. Determine risk level
   - High: all elements read on your product
   - Medium: some elements read, design-around possible
   - Low: clear non-infringement argument

6. Document opinion
   - Written FTO opinion from patent attorney (privilege)

Claim Chart Template

Patent: US10,XXX,XXX
Claim 1: "A method comprising..."

| Claim Element | Your Product Feature | Reads On? | Notes |
|---------------|---------------------|-----------|-------|
| receiving input data | API endpoint receives JSON | YES | Literal |
| applying neural network | ML model inference | YES | Literal |
| generating classification | Returns label/score | YES | Literal |

Overall: Potential infringement — consult attorney

Patent Landscape Analysis

Landscape Report Components

  1. Filing trends — volume over time, growth areas
  2. Top assignees — who owns the most patents
  3. Key inventors — prolific individual inventors
  4. Geographic coverage — where patents are filed
  5. Technology clustering — CPC code distribution
  6. Citation analysis — foundational/influential patents
  7. White spaces — areas with few patents (opportunities)

Visualization Tools

  • Lens.org → built-in charts, free
  • PatSnap → commercial landscape maps
  • Tableau/Python → custom visualizations from exported data
# Lens.org bulk export → Python analysis
import pandas as pd
import matplotlib.pyplot as plt

df = pd.read_csv('lens_export.csv')
df['year'] = pd.to_datetime(df['Filing Date']).dt.year

# Filing trend
df.groupby('year').size().plot(kind='bar', title='Patent Filings by Year')
plt.savefig('filing_trend.png')

# Top assignees
df['Assignee'].value_counts().head(20).plot(kind='barh')
plt.savefig('top_assignees.png')

IP Strategy Decisions

Build vs. License vs. Design Around

SituationRecommended Action
Core technology, patentableFile provisional → full application
Competitor has blocking patentLicense, design around, or challenge
Patent expires in <3 yearsWait or design around temporarily
Non-practicing entity (troll)Defend with prior art; consider IPR
Open source dependencyVerify patent grant in license

Patent Filing Decision Checklist

  • Is the invention novel (new)?
  • Is it non-obvious to someone skilled in the field?
  • Is there utility / commercial application?
  • Do we have a written description + claims ready?
  • Have we done a patentability search?
  • Do we want US only or PCT international filing?
  • Is a provisional application needed to establish priority date?
  • What is the budget? (US utility: $15-30K+)

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.