How to Run Novelty and FTO Research with AI Using One MCP Server
Run novelty, invention FTO, or design FTO research from one Claude connection — Patsnap Patent Research turns a technical description or product image into an asynchronous, reviewable patent-research workflow.
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.
One recorded novelty task screened 50 candidates and compared five core patents. A separate invention FTO task screened 792 candidates, excluded 787, reviewed five in detail, and classified all five as low risk with no high- or medium-risk findings.
Two observed runs — research snapshots, not legal opinions or performance benchmarks.When you need to assess whether a technical disclosure is novel or whether a product design risks infringing active patents, Claude alone cannot help. The model has no access to current patent databases, legal-status records, or the structured metadata required to screen prior art or map claims against your implementation.
Patsnap Patent Research is a unified MCP server that adds novelty search, invention FTO review, and design FTO analysis directly to your AI assistant or coding agent. Each workflow accepts a natural-language description or image, runs an asynchronous multi-step process across Patsnap’s patent database, and returns structured results—candidate counts, ranked evidence, risk classifications, and signed reports—that patent researchers and product teams can review and continue through follow-up prompts in the same Claude session.
This tutorial demonstrates how to connect the server, prepare the technical or product input, launch a selected research task, track its progress, interpret the returned evidence, and continue to professional review.
Choose between novelty, invention FTO, and design FTO
The server exposes three task types. Novelty search evaluates whether a technical disclosure is anticipated by prior art. Invention FTO review screens a product or technology implementation for potential infringement risk, filtering by jurisdiction, legal status, and expiration date. Design FTO accepts a product or design image plus optional context and runs a similar risk-screening workflow tuned for visual comparison.
All three workflows execute asynchronously and return structured step-by-step results. Choose the task that matches the stage of your work: novelty when you have a disclosure before filing; invention FTO when you are planning to manufacture or launch a product in a specific market; design FTO when visual appearance is the primary differentiator.
What Patsnap Patent Research adds to your AI assistant or coding agent
An AI assistant or coding agent can help you organize technical descriptions, draft claim-like language, and suggest keyword variations, but it cannot search patent databases or retrieve legal-status records. The Patsnap Patent Research MCP fills that gap by connecting Claude to a global patent corpus with structured metadata, legal-event timelines, and multi-dimensional search indexes.
The server unifies three workflows in one connection: novelty search for pre-filing disclosure review, invention FTO for product-launch clearance screening, and design FTO for image-based risk analysis. Each workflow accepts text or image input, runs background candidate retrieval and feature-level comparison, and returns a detailed report with ranked evidence, filter summaries, and risk classifications.
1Prepare the technical description, disclosure, product context, or image
Novelty search requires a technical description of the invention. The server recommends a passage longer than 200 words for best feature extraction and query construction, covering the problem addressed, the technical solution, key structural or functional elements, and any novel combinations or configurations. For example:
Thin aluminum inter-cell cooling plate with serpentine coolant channels, sealed phase-change-material pockets, inlet/outlet temperature sensing and delta-T flow control.
Invention FTO review needs a product implementation or technology summary with enough detail to map against claims in active patents. Include the functional elements, intended market or jurisdiction, and any distinguishing features. For design FTO, supply a clear product or design image and optional contextual text describing the intended use, market, and visual characteristics.
Gather the relevant jurisdiction codes, priority-date ranges, or legal-status filters you want to apply. The server supports jurisdiction filtering for FTO workflows and optional image attachments for novelty search when visual context helps.
Connect the MCP server to your AI workflow
https://connect.Patsnap.com/f176d7/mcp?apikey=yourapikeyReplace yourapikey with your own Patsnap Open Platform API key before connecting.Once your key is included, do not paste the completed link into a public chat, screenshot, shared document, or code repository.
- 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 from your account.
- Copy the connector address above and replace
yourapikeywith your key. - Add the server to your AI assistant or coding agent or another MCP-compatible client, then restart or reconnect it.
Once connected, choose novelty, invention FTO, or design FTO in natural language. An AI assistant or coding agent can start the asynchronous task and use the returned task ID for status checks and result retrieval.
3Start the selected research task
Ask your AI assistant or coding agent to run the appropriate task in natural language. For novelty:
Run a novelty search for this disclosure: [paste your technical description].
For invention FTO:
Run an FTO review for a portable UV-C water-bottle cap: UV-C LED, circumferential reflective surface, tilt-triggered cutoff, and closure-triggered timed disinfection cycle. Filter to US active patents.
For design FTO:
Run a design FTO analysis for this product image [attach image] in the EU market.
Your AI assistant or coding agent will call novelty_search, fto_review, or design_fto with your input, optional filters, and execution settings. The server returns a task ID and confirms that the workflow has started in the background. The novelty and FTO workflows each follow a seven-step process: feature extraction, keyword expansion, query construction, candidate retrieval, screening, detailed comparison, and report generation.
Track the task and read the observed results
The workflow runs in the background and returns a task ID. Ask your AI assistant or coding agent for the task status, then retrieve the completed result with the same ID.
What is the status of task novelty-6ad9b23bec2e?| Observed run | Candidate flow | Recorded outcome | Limitation |
|---|---|---|---|
| Battery-module cooling-plate novelty task | 50 screened; five core patents compared | Novelty and inventiveness appeared present around sealed phase-change-material pockets integrated with the liquid-cooling plate. | A research result, not a patentability opinion. |
| UV-C bottle-cap invention FTO task | 792 screened; 787 excluded; five reviewed in detail | All five reviewed patents were classified as low risk; zero high- or medium-risk findings. | A screening snapshot, not legal clearance. |
The novelty run recorded core comparisons including US20140272496A1 and US11223081B2. Review the underlying claims, family members, legal status, and jurisdiction before relying on the screening result.
Review evidence, filters, and risk signals
Examine the returned patent list, relevancy scores, and feature-overlap summaries. Novelty results highlight which prior-art patents match specific features of your disclosure and where gaps or novel combinations appear. FTO results classify each reviewed patent by infringement risk—high, medium, or low—and provide jurisdiction, legal-status, and expiration data for each.
The server documents applied filters—jurisdiction codes, priority-date ranges, legal-status categories, and any image or keyword refinements—in the task metadata so you can verify that the search scope matches your intended clearance or novelty boundary.
Explore returned patents and prepare for professional review
Use the returned publication numbers to dig deeper into individual records:
Retrieve the full text and legal status for US20140272496A1.
What is the current status and expiration date for the low-risk FTO result US11223081B2?
An AI assistant or coding agent can call additional tools from the same or related MCP servers to pull metadata, legal-event timelines, and citation graphs. When the initial screening identifies risk or prior art that requires closer examination, transfer the evidence and report to a qualified patent attorney or IP counsel for formal clearance, validity assessment, or filing strategy. The MCP workflow supports early-stage research and screening, not legal opinions or final patentability determinations.
Frequently asked questions
Why can’t I just Ask your AI assistant or coding agent directly for novelty or FTO analysis?
Claude has no access to current patent databases, legal-status records, or the structured metadata required to screen prior art or assess infringement risk. The Patsnap Patent Research MCP connects Claude to a global patent corpus with legal-event timelines and multi-dimensional search indexes, enabling evidence-backed novelty and FTO workflows.
What inputs do the three research tasks accept?
Novelty search accepts a technical description (recommended 200+ words) and optional images. Invention FTO review accepts a product or technology summary plus jurisdiction, legal-status, and date filters. Design FTO accepts a product or design image plus optional context describing intended use and market.
How do I track the progress of an asynchronous task?
After starting a task, Your AI assistant or coding agent returns a task ID. Ask your AI assistant or coding agent ‘What is the status of task [ID]?’ to retrieve the current step count, candidate totals, and completion status. When all steps finish, Ask your AI assistant or coding agent to retrieve the completed result using the same task ID.
What does the returned result include?
The result includes candidate counts at each workflow stage, the final set of compared patents with publication numbers and titles, feature-level comparison summaries, risk or novelty classifications, and signed Word and JSON report links. For FTO tasks, each reviewed patent is classified as high, medium, or low risk with jurisdiction, legal-status, and expiration data.
Are the novelty and FTO conclusions final legal opinions?
No. The server outputs support research and screening, not legal conclusions. Signed reports are research snapshots for internal review. Transfer the evidence and report to a qualified patent attorney or IP counsel for formal clearance, validity assessment, or filing strategy.
How much does it cost to run initial novelty or FTO tasks?
Patsnap’s Open Platform Starter plan is free and includes 10,000 credits for initial testing, letting you run early novelty or FTO tasks before committing to a larger plan.
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.
Use Patsnap Patent Research with AI
Connect the MCP server and continue this workflow with approved Patsnap data.
