estat

This MCP server provides access to the e-Stat API from the Japanese government statistics portal. It exposes tools that let a language model search for statistics tables by keywords and year, retrieve meta information for a given table ID, fetch data for a specific table or dataset, view dataset filtering criteria, and obtain file and database details for statistical data. The available tools are:

• search_e_stat_tables: search for statistical tables by keyword and year.
• get_e_stat_meta_info: retrieve meta information for a given statistical table ID.
• get_specific_e_stat_data: obtain statistics data for a table ID or dataset ID.
• get_e_stat_ref_dataset: inspect filtering criteria and references for a dataset.
• get_e_stat_data_catalog: obtain information about statistical table files and databases. To use the server, you must supply an e-Stat API application ID. Obtain it by registering at the e-Stat API portal, agreeing to the terms, and generating an App ID. Once the MCP server is running and the API key is configured (via E_STAT_APP_ID), you can issue natural-language requests that map to these tools to search, fetch, and compare statistics data from e-Stat.

Prerequisites:

• Python 3.8+ installed on your system
• Access to the internet to install dependencies
• An e-Stat API App ID (will be provided to the server via environment variable)

1. Install the MCP server dependencies (via uv):

• Ensure you have uv installed (pip install uv-mcp or your preferred MCP framework installation method as described by the repository)

2. Prepare the server directory:

• Place the estat-mcp-server in a local directory. The entry point should be server.py and should expose the MCP tools described in the README.
• Example structure: /path/to/estat-mcp-server/ server.py (any required modules)

3. Configure environment variables:

• Obtain your e-Stat API App ID and set it in the environment variable E_STAT_APP_ID.

4. Run the MCP server (using the uv command as shown in mcp_config):

• uv --directory /path/to/estat-mcp-server run server.py

5. Verify operation:

• Connect a client to the MCP endpoint and perform a simple tool call, e.g., search_e_stat_tables with a keyword like “人口” (population).

Notes:

• The API key (E_STAT_APP_ID) is required for all calls to e-Stat. Keep it secret and do not commit to version control.
• The exact file paths and directory names should be adjusted to your environment when configuring the mcp_config.
• If you switch to the uvx (Python/uv) variant, ensure the environment and command suit the uvx usage (as demonstrated in the README examples).
• If you encounter API rate limits or authentication errors, verify that the App ID is correctly set and active for your account.
• This MCP server mirrors the capabilities described in the README: searching for tables, retrieving metadata, obtaining data, and inspecting data catalogs. Customize error handling and logging in server.py as needed for production deployment.