home / mcp / e-stat mcp server

e-Stat MCP Server

Provides access to e-Stat statistics via MCP tools for searching, metadata, data retrieval, and data catalog queries.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "koizumikento-e-stats-mcp": {
      "command": "e-stats-mcp",
      "args": [],
      "env": {
        "E_STAT_APP_ID": "your-app-id"
      }
    }
  }
}

The e-Stat MCP Server lets you query and retrieve statistics from the e-Stat API, including table search, metadata, data extraction, and data catalog information. It exposes a set of tools that you can invoke from an MCP client to perform targeted searches, fetch data in JSON or CSV, manage datasets, and explore data catalogs. This server family is designed to integrate with your MCP workflow so you can automate statistical data retrieval in your applications.

How to use

Set up the MCP server in your environment and run it through your MCP client. You will access tools to search statistics, fetch table metadata, obtain actual data, post and reference datasets, and retrieve data catalog information. Each tool accepts structured arguments such as keywords, survey years, data IDs, and dataset parameters. You can combine multiple tools in scripts to build automated data retrieval pipelines for reporting, analysis, or data curation tasks.

Common usage patterns include: performing keyword searches for statistics, retrieving CSV data for offline processing, obtaining detailed metadata for tables, downloading specific statistics datasets, and registering or referencing datasets for later reuse. Your MCP client will expose tools with clearly named endpoints like search_stats_by_keyword, get_stats_list, get_stats_data, post_dataset, and get_data_catalog to cover these tasks.

How to install

Prerequisites you need before installing the MCP server: a runtime environment with the MCP tooling installed (uv or uvx) and Git for cloning repositories when needed.

export PATH="$PATH:/path/to/uv/bin"  # ensure uv is on your PATH

# Install the MCP server package from its source
uv tool install git+https://github.com/koizumikento/e-stats-mcp.git

# Or install locally by cloning the repository and syncing
git clone https://github.com/koizumikento/e-stats-mcp.git
cd e-stats-mcp
uv sync

# Note: You will need to configure your application ID as described in the configuration steps.

Configuration and usage notes

Application ID to use the e-Stat API To access the e-Stat API, you must provide an Application ID. Obtain it by registering, logging in, and generating an application key on the e-Stat API portal.

Set the Application ID in your environment ``export E_STAT_APP_ID="your-app-id"``

MCP server configuration for the Claude Desktop client The MCP server can be added to Claude Desktop for easy management. Use the following configuration snippet to register the server in Claude’s config.

{
  "mcpServers": {
    "e-stats-mcp": {
      "command": "e-stats-mcp",
      "env": {
        "E_STAT_APP_ID": "your-app-id"
      }
    }
  }
}

Alternative direct-from-GitHub and local development configurations are provided to run the MCP server via UVX or UV. Use one of these depending on your setup.

{
  "mcpServers": {
    "e-stats-mcp": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/koizumikento/e-stats-mcp.git",
        "e-stats-mcp"
      ],
      "env": {
        "E_STAT_APP_ID": "your-app-id"
      }
    }
  }
}

{
  "mcpServers": {
    "e-stats-mcp": {
      "command": "uv",
      "args": [
        "run",
        "--directory",
        "/path/to/e-stats-mcp",
        "e-stats-mcp"
      ],
      "env": {
        "E_STAT_APP_ID": "your-app-id"
      }
    }
  }
}

Cursor-based project configuration mirrors the same approach and enables integration at the project level. You add a mcp.json in your project’s cursor folder with the same command structure used above to run the MCP server.

{
  "mcpServers": {
    "e-stats-mcp": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/koizumikento/e-stats-mcp.git",
        "e-stats-mcp"
      ],
      "env": {
        "E_STAT_APP_ID": "your-app-id"
      }
    }
  }
}

Troubleshooting and tips

If you encounter issues connecting to the e-Stat API, verify that your Application ID is correct and that your network can reach the API endpoint. Ensure the environment variable E_STAT_APP_ID is exported in the shell that launches the MCP client.

When running locally, confirm that you are in the correct project directory and that the uv or uvx runtime is properly installed and available in your PATH. If you use a JSON config, verify that the JSON structure follows the exact examples shown to avoid parsing errors.

End-to-end tool surface

The server exposes a comprehensive set of tools to cover common data access tasks. You can search statistics, fetch metadata, retrieve statistics data, manage datasets, and explore data catalogs.

Available tools

get_stats_list

Search for statistics table information and fetch results in JSON or CSV.

get_stats_list_csv

Retrieve statistics table information as CSV format for easy processing.

search_stats_by_keyword

Quick keyword search to find relevant statistics tables.

get_meta_info

Fetch metadata information for a specific statistics table.

get_meta_info_csv

Retrieve metadata information in CSV format.

get_stats_data

Obtain actual statistics data by ID with optional limits.

get_stats_data_csv

Fetch statistics data in CSV format.

get_stats_data_bulk

Retrieve multiple statistics datasets in a single call.

post_dataset

get_dataset

Reference an existing dataset by ID.

get_data_catalog

Obtain data catalog information in JSON or CSV.

get_data_catalog_csv

Retrieve data catalog information as CSV.

get_stats_fields

List of statistic field codes for different domains.