utility-patent-reviewer

The Utility Patent Reviewer is a Python-based MCP server that provides Claude-assisted guidance for drafting patent documentation, searching USPTO rules, and performing automated reviews of patent applications. It integrates patent examination standards and large-scale patent data access to help you understand prior art, assess claim language, and generate structured, examination-ready text through natural conversations. You can use it to query MPEP references, run automated checks on your draft, and generate diagrams or structured sections to accelerate patent preparation. The tool emphasizes privacy and local usage, with Claude Code serving as the conversational interface for interactions and drafting assistance.

Prerequisites:

• Python 3.9 or higher installed on your system
• Git to clone the repository
• Internet access for initial setup and optionally USPTO data downloads

Install steps:

1. Clone the repository:

git clone https://github.com/RobThePCGuy/utility-patent-reviewer.git
cd utility-patent-reviewer

2. (Optional) Create a virtual environment:

python3 -m venv venv
# Windows:
venv\Scripts\activate
# macOS/Linux:
source venv/bin/activate

3. Install dependencies (if a requirements file exists or as documented in the repo):

pip install -r requirements.txt

4. Run the server entry point (as configured by the MCP setup):

python -m utility_patent_reviewer

Note: If the project uses a different module name, adjust the -m flag accordingly. The README indicates the install process handles environment setup and optional downloads for USPTO rules and patents; follow any additional on-screen prompts during installation.

Tips and notes:

• The tool emphasizes privacy; data does not need to be uploaded to external services for core functionality.
• Ensure you have Claude Code installed if you plan to use Claude-based interactions for drafting and reviews.
• The installer may download large data packages (USPTO rules ~500MB and patent corpus ~13GB); ensure sufficient disk space and bandwidth.
• If you encounter import errors, verify Python virtual environment activation and compatible package versions listed in Quick Start notes.
• For best results, use the recommended PyTorch/transformer/offline components as specified in the repo documentation and keep your environment isolated from system-wide Python packages.
• When interacting with MPEP and patent content, rely on the automated checks for compliance, but consult legal counsel for final filings.
• If you need to customize startup, you can adjust the MCP command in the configuration to point to a different entry point or script as per your deployment needs.