patent-intelligence
Patentability assessments, prior art searches, and invention feature mapping for IP law teams.
Open source Repository Open in the app JSON README (API)
About
Patentability assessments, prior art searches, and invention feature mapping for IP law teams.
Details
- Kind
- Plugins
- Topic
- Government & public data
- Publisher
- fluidityiq
- Origin
- marketplace
- Category
- ferramentas
- Last push
- 2026-08-28T18:13:17Z
- Repository state
- ativo
- License
- MIT
- Added
- 2026-08-30 01:48:58
- Updated
- 2026-08-30 01:48:58
- Origin id
fluidityiq/patent-intelligence-plugin/patent-intelligence
README
# Patent Intelligence Plugin
An AI-powered productivity plugin for IP law teams, designed for [Cowork](https://claude.com/product/cowork). Automates patentability assessments, prior art searches, and invention feature mapping — all configurable to your organization's specific workflows and preferences.
> **Disclaimer:** This plugin assists with patent workflows but does not provide legal advice. All analysis should be reviewed by a registered patent attorney or agent before being relied upon. AI-generated analysis should not be used as a substitute for professional legal judgment.
## Target Personas
- **Patent Counsel** — Patentability evaluations, prior art landscape analysis, prosecution support
- **Patent Agents** — Prior art search execution, feature mapping, report preparation
- **Inventors & R&D Teams** — Invention disclosure submission, preliminary patentability feedback
- **IP Portfolio Managers** — Portfolio gap analysis, technology landscape monitoring
## Installation
```
claude plugins add FluidityIQ/patent-intelligence-plugin
```
## Quick Start
### 1. Install the plugin
```
claude plugins add FluidityIQ/patent-intelligence-plugin
```
### 2. Configure your Patent Search MCP
The plugin requires a patent search MCP server that provides semantic search and patent details retrieval. Configure the connection in `.mcp.json` at the plugin root. Authentication is handled automatically via OAuth (auto-discovery); for org-wide or CI deployments using an API key instead, see [CONNECTORS.md](CONNECTORS.md) for the API key configuration.
### 3. Configure your settings (optional)
Create a local settings file to define your organization's preferences. In your project's `.claude/` directory, create a `patent-intelligence.local.md` file:
```markdown
# FluidityIQ Patent Intelligence Plugin Settings
## Report Branding
- Organization: [Your Organization Name]
- Matter numbering convention: [e.g., PAT-2026-001]
- Default privilege marking: PRIVILEGED AND CONFIDENTIAL — ATTORNEY WORK PRODUCT
## Search Preferences
- Default jurisdiction focus: US, EP, WO, CN, JP, KR
- Minimum search rounds: 3
- Result threshold: 20-30 results per round
## Review Workflow
- Reports route to: [Patent Counsel Name/Email]
- Required sign-off: [Senior Counsel / IP Committee]
- Escalation path: [For anticipatory references or complex combinations]
## Technology Focus Areas
- Priority fields: [e.g., semiconductor, biologics, wireless]
- Internal portfolio: [Reference to internal patent database or tracking system]
- Competitive landscape: [Key competitors to monitor]
```
### 4. Connect your tools
The plugin requires the patent search MCP (FluidityIQ Patents MCP). See [CONNECTORS.md](CONNECTORS.md) for configuration.
## Commands
### `/patent-intelligence:assess-patentability` — Patentability Assessment
Conduct a full patentability assessment for an invention or idea. Gathers invention details, searches patent prior art via semantic search, maps features against references, and produces a structured report.
```
/patent-intelligence:assess-patentability
```
Accepts: invention description, Invention Disclosure Form, pasted text, or file upload. The command walks through a structured workflow:
1. **Intake** — Gather invention details and identify the point of novelty
2. **Search** — Multi-round semantic patent search with terminology refinement
3. **Classify** — Categorize references as Anticipatory, Combinable, or Background
4. **Map** — Build feature coverage matrix with specific citations
5. **Analyze** — Evaluate novelty and non-obviousness
6. **Report** — Generate the structured Patentability Assessment Report
7. **Recommend** — Suggest next steps based on findings
## Skills
| Skill | Description |
|-------|-------------|
| `patentability` | Semantic patent search, prior art classification, feature coverage mapping, novelty and non-obviousness analysis, structured report generation |
## Example Workflows
### Patentability Assessment for a New Invention
1. Inventor describes their idea or uploads an Invention Disclosure Form
2. Run `/patent-intelligence:assess-patentability`
3. Provide context: technology field, known prior art, key features believed to be novel
4. The skill runs 3-4 rounds of semantic patent search, harvesting terminology along the way
5. References are classified (Anticipatory / Combinable / Background) and mapped against invention features
6. Receive a structured report with feature coverage matrix, novelty analysis, and non-obviousness assessment
7. Route the report to patent counsel for review and filing decision
### Quick Prior Art Check for an Idea
1. Describe the idea in a few sentences
2. Run `/patent-intelligence:assess-patentability`
3. The skill identifies the point of novelty and runs a focused search
4. Get a rapid assessment of whether similar patents exist and what features overlap
5. Decide whether to invest in a full invention disclosure
### Portfolio Gap Analysis
1. Describe a technology area your team is working in
2. Run `/patent-intelligence:assess-patentability` for each inventive concept
3. Compare feature coverage matrices across assessments
4. Identify unprotected innovations and prioritize filing
## MCP Integration
> If you see unfamiliar placeholders or need to check which tools are connected, see [CONNECTORS.md](CONNECTORS.md).
The plugin connects to tools through MCP (Model Context Protocol) servers. The patent search MCP (FluidityIQ Patents MCP) is **required**:
| Category | Purpose |
|----------|---------|
| Patent search | Semantic prior art search, patent document retrieval |
Semantic search (`search_patents`) takes an optional **`assignee`** for company-scoped landscape search. Put the company in `assignee` and keep `query` as the technical topic (e.g. wireless charging + Apple). Omit `assignee` for a world prior-art / patentability search. Details: [CONNECTORS.md](CONNECTORS.md).
## Customization
### Local Settings
Your settings file (`patent-intelligence.local.md`) controls:
- **Report branding**: Organization name, matter numbering, privilege markings
- **Search preferences**: Jurisdiction focus, search depth, result thresholds
- **Review workflow**: Routing, sign-offs, escalation paths
- **Technology focus**: Priority fields, competitive landscape, internal portfolio references
### Report Output
The Patentability Assessment Report follows a structured template:
1. **Executive Summary** — Key findings in 2-3 sentences
2. **Feature Coverage Matrix** — Visual mapping of features vs. references
3. **Invention Summary** — Technical field, problem, solution, key features
4. **Prior Art References** — Categorized reference table with detailed descriptions
5. **Novelty Analysis** — Anticipation assessment per reference
6. **Non-Obviousness Analysis** — Combination analysis with motivation assessment
7. **Search Methodology** — Search strategy, rounds executed, tools used
8. **Limitations** — Patent-only scope, 18-month window, semantic search constraints
### Reference Classification
References follow a three-tier classification:
| Category | Meaning | Action |
|----------|---------|--------|
| **Anticipatory** | Discloses all key features — threatens novelty | Detailed mapping, immediate flag, design-around analysis |
| **Combinable** | Discloses some features — may threaten non-obviousness | Feature gap analysis, combination assessment |
| **Background** | Provides context, no direct threat | Landscape awareness |
## File Structure
```
patent-intelligence-plugin/
├── .claude-plugin/plugin.json
├── .mcp.json
├── README.md
├── CONNECTORS.md
├── SECURITY.md
├── LICENSE
├── commands/
│ └── assess-patentability.md
└── skills/
└── patentability/
└── SKILL.md
```
## Roadmap
Future skills planned for the Patent Intelligence Plugin:
| Skill | Description | Status |
|-------|-------------|--------|
| `patentability` | Patent prior art search and assessment | Available |
| `npl-search` | Non-patent literature search (journals, conferences, standards) | Planned |
| `fto-search` | Freedom-to-operate / right-to-practice analysis | Planned |
| `invalidity-search` | Patent invalidity and inter partes review support | Planned |
| `portfolio-analysis` | Patent portfolio gap analysis and landscaping | Planned |
| `claim-drafting` | Claim drafting assistance from invention disclosures | Planned |
## Support
- **GitHub Issues:** [Report bugs or request features](https://github.com/FluidityIQ/patent-intelligence-plugin/issues)
- **Email:** support@fluidityiq.com
## Privacy
This plugin sends invention descriptions and semantic search queries to the Patent Search MCP (FluidityIQ Patents MCP at `https://patents.fluidityiq.ai/mcp`) for prior art search and patent retrieval. No other conversation data is transmitted. For full details on data collection, use, and retention, see the [FluidityIQ Privacy Policy](https://www.fluidityiq.com/privacy-policy).
## Terms of Service
For more details, see the [FluidityIQ Terms of Service](https://www.fluidityiq.com/terms-of-service).
## License
MIT — see [LICENSE](LICENSE) for details.