How to Run an Invention FTO Search with AI
A product’s launch date shouldn’t outrun its patent screening. This tutorial runs a real invention FTO search with AI through Patsnap Patent Research, from product description to a risk-ranked candidate list.
This workflow works in Claude, Claude Code, ChatGPT, or Codex when the selected client supports the required MCP connection. After connecting, the research steps and evidence boundaries are the same.
Product teams preparing a market launch need to know whether their product’s features might infringe existing patents in the jurisdictions they’re entering. That kind of review needs fresh patent data, claim-level analysis, and legal-status filtering — none of which Claude has on its own. MCP lets Claude request that professional retrieval through a connected patent research server while you keep full control over the query, jurisdiction scope, and interpretation. Patsnap Patent Research connects Claude to unified novelty, invention FTO, and design FTO workflows — this tutorial walks through a live invention FTO task for a portable UV-C water-bottle cap launching in the United States.
Define product features, jurisdictions, and launch context
Start by gathering the technical features that define your product’s function and structure, the country or patent authority where you plan to sell, and your expected launch timeline. An invention FTO search needs a clear product description, not an abstract concept — the screening process matches your implementation against granted and published claims in the selected jurisdiction. Generic keyword search can’t answer FTO questions because it doesn’t expose legal-status filters, family grouping, or claim-level risk scoring, and manual screening in a traditional database is possible but slow once hundreds of candidates are involved.
Patsnap Patent Research gives Claude jurisdiction-aware FTO screening, claim-chart review, and legal-status filtering across four tools spanning novelty search, invention FTO, and design FTO — all sharing an asynchronous workflow with structured step-by-step results. The fto_review tool runs in two modes: lite for a faster screening pass, and pro for deeper analysis — this tutorial uses lite.
Distinguish FTO screening from a legal opinion
An FTO search identifies patents that may block your product in a specific jurisdiction. The output is a ranked list of potential blocking rights, not a legal clearance — results need professional review to confirm scope, validity, and enforceability, and zero blocking candidates doesn’t eliminate the need for a signed legal opinion before launch.
Prepare a concise product description covering functional features, structural elements, and any novel implementation details, plus the patent authority relevant to your market (US, EP, CN, and so on). You can optionally set an earliest priority or publication date to exclude art that predates your development timeline. This tutorial’s case: a portable UV-C water-bottle cap with four features — a UV-C LED, a circumferential reflective surface, a tilt-triggered cutoff, and a closure-triggered timed disinfection cycle — targeting the US market, so the jurisdiction filter is set to US.
Connect Patsnap Patent Research
https://connect.Patsnap.com/f176d7/mcp?apikey=yourapikey Replace yourapikey with your own Patsnap Open Platform API key before connecting. Keep the completed URL private. - Claude: add the remote MCP connection from Claude’s connector or integration settings.
- Claude Code: add the server as a remote HTTP MCP connection in your user or project configuration.
- ChatGPT: add the server through the supported connector or developer-mode MCP setup available to your workspace.
- Codex: add the server in Codex’s MCP configuration for the relevant local workspace.
The interface differs by client, but you use the same Patsnap connection URL and the same natural-language research prompt.
- Get a Patsnap Open Platform API key.
- Copy the connection URL or JSON config from the Connect panel on the Patsnap Patent Research page.
- Add the server to your AI assistant or coding agent or another MCP-compatible client, then restart it.
- Confirm the four tools — novelty_search, fto_review, design_fto, get_task — are available before you ask.
Start the invention FTO task
Run an invention FTO review for a portable UV-C water-bottle cap with a UV-C LED, circumferential reflective surface, tilt-triggered cutoff, and closure-triggered timed disinfection cycle. I plan to launch in the United States.Claude calls fto_review with your product description as input text, US as the authority filter, and lite mode. The server returns a task ID and begins the asynchronous workflow; call get_task with that ID to monitor progress through seven steps — feature extraction, keyword expansion, query construction, jurisdiction-aware screening, claim-chart review, risk classification, and report generation. In the recorded run, task fto-78bd805aeb2e completed all seven steps.
Review claims, families, status, and jurisdictions
| Field | Value |
|---|---|
| Candidates screened | 792 |
| Excluded by jurisdiction/status filters | 787 |
| Entered claim-chart review | 5 |
| Risk distribution | 0 high, 0 medium, 5 low |
Each of the five results includes the publication number, title, priority date, legal status, and a summary of the matched claims. The server also generates a Word report and a structured JSON file with the full claim charts and feature mappings, for internal review and for patent counsel.
Identify results requiring specialist review
Ask your AI assistant or coding agent to extract the publication numbers, priority dates, and current legal status for each of the five results, and confirm which are published applications versus granted patents with active status in the US. Request the full independent claims for each low-risk result and compare them to your product features — if a claim element seems to read on your implementation, escalate it to patent counsel for formal claim-chart analysis. If status shows “pending” rather than “granted,” check the prosecution history to understand the scope under examination. Check patent families too, in case related filings in other target jurisdictions were missed by the US-only filter — if Europe or Asia are on your roadmap, widen the authority filter and rerun.
Continue to a qualified FTO assessment
Compile the publication numbers, claim charts, and legal-status summaries into a briefing package for your patent attorney, along with the Word and JSON reports, your original product description, and any prior art you already know about. Schedule a formal FTO consultation to confirm the claim-chart mappings and assess validity and enforceability, and request a signed FTO opinion letter documenting the search scope, reviewed references, and legal conclusion. The search with AI handles the jurisdiction-specific retrieval and claim-level screening a general LLM can’t — the final clearance decision stays a legal judgment only qualified counsel can make.
Run an invention FTO search with AI
Connect Patsnap Patent Research and screen a product description against jurisdiction-specific patent data.
Frequently asked questions
Why can’t I just Ask your AI assistant or coding agent to run an FTO search without connecting Patsnap Patent Research?
Claude has no live connection to patent databases, legal-status filters, or claim-chart tools. An FTO review needs jurisdiction-specific retrieval and claim-level risk scoring that only a connected patent server can provide.
What’s the difference between a novelty search and an invention FTO search?
A novelty search identifies prior art that may affect patentability. An FTO search identifies granted and published patents that may block your product in specific jurisdictions — novelty focuses on disclosure dates and technical similarity, FTO on legal status, claim scope, and geography.
Does a zero high-risk result mean my product is clear to launch?
No. It means automated screening found no blocking claims in the current dataset. Legal status changes, new publications appear, and claim interpretation needs professional judgment — always get a signed FTO opinion before launch.
Can I run an FTO search across multiple jurisdictions at once?
Yes. The fto_review tool accepts authority filters covering multiple patent offices — specify US, EP, CN, or others to screen candidates across your target markets in a single task.
Can I use this workflow in Claude, Claude Code, ChatGPT, or Codex?
Yes, provided the selected client and workspace support the required MCP connection. Connect the same Patsnap server, then use the natural-language steps shown in this guide; interface details may differ by client.
Disclosure & disclaimer
Who published this
Published by Patsnap, which develops and sells the Patsnap Patent Research MCP server described here. This is an editorial tutorial, not an independent review.
How the information was gathered
Server capabilities and connection details are taken from the Patsnap Patent Research product page and Patsnap Open Platform documentation, as accessed on August 20, 2026. They are not the result of independent testing beyond the recorded screening run described in this tutorial. Product capabilities change frequently and may have changed since publication.
Not legal advice
The screening results in this tutorial are a risk-ranked research lead, not a freedom-to-operate opinion or legal conclusion. Obtain a signed FTO opinion from qualified patent counsel before any launch or product decision.
Trademarks
All trademarks, service marks, product names, and company names mentioned — including Claude and Cursor — are the property of their respective owners and are used here solely for identification purposes. Their use does not imply any affiliation with, sponsorship by, or endorsement from their respective owners.
