This MCP server provides access to Hong Kong finance data through a FastMCP interface. It aggregates a variety of datasets from Hong Kong authorities, including monthly business registrations, mortgage loan trends, credit card lending surveys, ATM locations, HIBOR figures, ATM networks, and more. The server is designed to be consumed by clients that support the MCP protocol, enabling you to query data, fetch statistics, and generate insights or charts from the underlying data sources.

To use the server, you run it using the uvx-based packaging (as configured for Cline integration). The server exposes data endpoints via the MCP protocol. Tools and capabilities you can leverage include: retrieving monthly and quarterly statistics, listing and filtering data by region (e.g., Hong Kong), getting up-to-date figures like HIBOR trends, and producing chart-ready data (the repository notes a chart tool for line plots). If you have an integration layer like Cline, you can connect using stdio and point to the published MCP package.

Prerequisites:

• Python 3.8+ installed
• Git installed
• Internet access to fetch dependencies

Installation steps:

1. Clone the repository: git clone https://github.com/hkopenai/hk-finance-mcp-server.git
2. Navigate to the project directory: cd hk-finance-mcp-server
3. (Optional) Create a virtual environment: python -m venv venv

   on Windows: venv\Scripts\activate

   on macOS/Linux: source venv/bin/activate

4. Install dependencies: pip install -r requirements.txt
5. Run the server (stdio mode): uvx hkopenai.hk-finance-mcp-server

   or if you are not using uvx packaging, you can run via the script directly when installed

6. If you plan to use SSE or other modes, refer to the server's startup options: python server.py python server.py --sse

Notes:

• Ensure network access to data sources listed in the Data Source section of the README is allowed.
• If you deploy in a container or orchestration system, use the mcp_config configuration to connect the server via uvx.

Tips and considerations:

• Data sources come from HKMA and IRD; data freshness depends on source updates.
• If you encounter environment-related issues, ensure the virtual environment is activated and dependencies from requirements.txt are installed.
• When using Cline, the provided example shows how to connect hk-finance via stdio using the uvx-based packaging; adapt the package name if your deployment differs.
• For charting and visualization, the repository references a line chart example; ensure the chart tool is available in your integration layer to render plots.