This MCP server provides access to patent data intelligence capabilities, including fuzzy company searches, patent information lookups, and patent analytics. You can query for company-related information using fuzzy searches to identify potential patent holders or relevant entities, perform precise or broad patent searches by name, application number, applicant, or代理机构, and analyze patent statistics such as application and grant trends by enterprise. The available tools enable patent landscape analysis, competitive intelligence, and IP risk assessment by returning structured results that include patent titles, numbers, dates, IPC classifications, and status. To connect, deploy the server (via the provided STDIO/uv method) and configure Cursor/Cherry Studio MCP to point at the local or remote endpoint (e.g., streamable HTTP endpoint or the STDIO interface) as described in the configuration guides. The server exposes the following tools: patent_bigdata_fuzzy_search for enterprise keyword fuzzy matching; patent_bigdata_patent_search for patent information search; and patent_bigdata_patent_stats for enterprise patent statistics, including status distributions and time-based trends. Use these tools to build IP intelligence workflows, monitor competitors, or support R&D and legal teams with up-to-date patent data.

Prerequisites:

• Python 3.10+
• Git
• Basic shell/terminal access
• Optional: Virtual environment tools (venv)

1. Clone the repository git clone https://github.com/handaas/patent-mcp-server cd patent-mcp-server

2. Create and activate a Python virtual environment python -m venv mcp_env && source mcp_env/bin/activate

3. Install dependencies pip install -r requirements.txt

4. Configure environment variables

   • Copy the template and edit cp .env.example .env

     Edit .env and set INTEGRATOR_ID, SECRET_ID, SECRET_KEY as provided by Handaas

5. Start the server (streamable-http) python server/mcp_server.py streamable-http

   The server will typically be available at http://localhost:8000

6. (Optional) STDIO/Cursor deployment Use the provided STDIO configuration to run the Python script via uv (stdio mode) by adding the following to your Cursor/Cherry Studio MCP config:

   { "mcpServers": { "patent-mcp-server": { "type": "stdio", "command": "uv", "args": ["run", "mcp", "run", "{workdir}/server/mcp_server.py"], "env": { "PATH": "{workdir}/mcp_env/bin:$PATH", "PYTHONPATH": "{workdir}/mcp_env", "INTEGRATOR_ID": "your_integrator_id", "SECRET_ID": "your_secret_id", "SECRET_KEY": "your_secret_key" } } } }

7. For remote/streamable-http usage, configure Cursor/Cherry Studio to point to the streamable HTTP endpoint as described in the usage section.

Tips and caveats:

• Ensure Python 3.10+ is installed and the virtual environment is active when installing dependencies.
• In streamable-http mode, the service starts on http://localhost:8000 by default; adjust firewall or endpoint settings if you run it on a server.
• When using STDIO/UV deployment, the mcpEnv virtualenv path must be correctly set in PATH and PYTHONPATH.
• The environment variables INTEGRATOR_ID, SECRET_ID, and SECRET_KEY are required for authenticated access to the patent data services.
• For the Cursor/Cherry Studio integration, one must follow the guide to fetch enterprise full names via patent_bigdata_fuzzy_search if only partial company names are available.
• API results enforce a maximum of 50 items per page; use pageIndex/pageSize to paginate results.
• Supported patent types include invention, utility model, design, and combinations thereof; filter using patentType where provided.