- Home
- Skills
- Legal & Compliance
- Patent Novelty Evaluator: Blunt Prior-Art Verdict
Works with the AI Agents you already use
Patent Novelty Evaluator: Blunt Prior-Art Verdict
Checks novelty, inventive step, and eligibility against the closest prior art and gives a blunt verdict.
$10
About this skill
What it does
Stress-tests an invention for patentability before you spend money on filing. It extracts the claim skeleton, searches patents, papers, and products for prior art, compares the claim element by element against the closest reference, and applies the novelty, inventive-step, and eligibility tests for your jurisdiction (US, Europe, or India).
Why it exists
Most inventions feel novel only relative to what the inventor searched. Drafting a patent around something already disclosed wastes time and attorney fees. This skill tries hard to find the prior art that kills the claim, and says so plainly.
How it works
- Writes the claim skeleton first: the problem, the mechanism, and the one element believed to be new.
- Searches within a bounded budget using a bundled patent-search guide (databases, CPC classes, query syntax, priority dates).
- Names the single closest reference instead of a pile of loosely related ones.
- Applies the tests from a bundled jurisdiction table: US (35 U.S.C. 101/102/103, Alice/Mayo, KSR), Europe (EPC Articles 52, 54, 56, problem-solution approach), and India (Patents Act Sections 2(1)(j), 2(1)(ja), 3(k)).
- Suggests the specific narrowing or technical evidence that would make a weak claim stronger.
What you get
A verdict first (Dead on arrival / Weak / Defensible), the claim skeleton, the closest prior art, an element-by-element table, inventive-step and eligibility analysis, ways to strengthen the claim, and a search log.
When not to use it
Drafting the patent itself, freedom-to-operate (infringement) checks, or legal advice. It's a technical assessment to bring to a patent attorney.
Open source
Part of claude-rigor-skills: 30 MIT-licensed skills that make AI show its work. The demo below, with its inputs, is in the examples folder.
See it in action
Request:
Use the patent-novelty-evaluator skill. Invention: a motorcycle helmet with a dry-electrode EEG band in the liner and an IMU. It flags rider drowsiness only when EEG alpha/theta power rises AND the IMU detects head-nodding within the same 10-second window, then vibrates the helmet and sends an alert to a paired phone. I want to file in India.
Result: the complete, unedited output from a fresh Claude Code session is in examples/patent-novelty-evaluator.
Known limitations
- Not legal advice. A patent attorney should review before any filing decision.
- Through general web search it can't query patent databases interactively, so CPC-filtered and national-database searches (such as InPASS) may be incomplete; the search log says exactly what was and wasn't checked.
- Chinese, Japanese, and Korean patent literature is often under-covered by web search.
- Default budget is 15 searches and 6 documents; complex inventions may need the deeper filing-stage budget.
FAQ
Is this legal advice? No. It's a technical prior-art and claim analysis to make your conversation with a patent attorney faster and cheaper.
Which countries does it cover? The bundled reference covers the US, Europe (EPO), and India. For other jurisdictions it applies the general novelty and inventive-step logic and tells you to confirm locally.
Why is the verdict so blunt? Because rounding "weak" up to "defensible" costs you a filing fee. When a claim is weak it tells you what narrowing or evidence would strengthen it.
Compatibility
Works with any agent that supports the SKILL.md standard and has web search. The demo was run with Claude Code 2.1.296 (claude-sonnet-5-5) using its WebSearch and WebFetch tools. Jurisdictions covered in the bundled reference: US, Europe (EPO), India.
Documentation: https://github.com/hemapriyan-rk/claude-rigor-skills/tree/main/skills/patent-novelty-evaluator
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 29 days ago
Permissions
No special permissions declared or detected