Get the FREE Ultimate OpenClaw Setup Guide →

datagouv

Official data.gouv.fr Model Context Protocol (MCP) server that allows AI chatbots to search, explore, and analyze datasets from the French national Open Data platform, directly through conversation.

Installation
Run this command in your terminal to add the MCP server to Claude Code.
Run in terminal:
Command
claude mcp add --transport stdio datagouv-datagouv-mcp npx -y mcp-remote https://mcp.data.gouv.fr/mcp

How to use

The data.gouv.fr MCP server provides a Model Context Protocol (MCP) endpoint that lets AI chatbots query and interact with datasets from data.gouv.fr directly within conversations. By connecting to the MCP endpoint at https://mcp.data.gouv.fr/mcp, clients can search for datasets, browse metadata, and request data highlights or summaries without leaving the chat interface. This makes it easier to ask natural language questions like about dataset availability, recent updates, or specific data fields, and receive structured responses powered by the MCP backend. The server is designed to be compatible with a wide range of MCP clients, including AnythingLLM, ChatGPT, Claude, Cursor, Gemini CLI, HuggingChat, IBM Bob, Kiro, Le Chat (Mistral), Mistral Vibe, VS Code, Windsurf, and more, by providing standard MCP server configuration options.

To use it, you typically connect your chatbot to the hosted endpoint https://mcp.data.gouv.fr/mcp (or your self-hosted URL if you run it locally). For clients that require a local command to launch the MCP proxy, you can execute the recommended setup using npx mcp-remote https://mcp.data.gouv.fr/mcp (the example shown here includes the -y flag to streamline the prompt handling). Once connected, queries are passed to the MCP server, which returns indexed results, dataset descriptions, and analytic insights from data.gouv.fr.

How to install

Prerequisites:

  • Node.js (LTS version) and npm installed on your machine
  • Basic familiarity with running commands in your terminal

Option 1: Quick start using npx (no local installation required)

  1. Ensure Node.js and npm are installed on your system.
  2. Run the MCP client with the remote endpoint:
npx -y mcp-remote https://mcp.data.gouv.fr/mcp
  1. Follow any prompts and connect the generated MCP client to your chatbot of choice.

Option 2: Local installation for repeated usage

  1. Install Node.js (and npm) from https://nodejs.org/.
  2. Install the mcp-remote package globally (optional):
npm install -g mcp-remote
  1. Start the MCP client pointing at the data.gouv MCP endpoint:
mcp-remote -y https://mcp.data.gouv.fr/mcp
  1. Use the endpoint in your chat plugin or integration as documented for your chatbot.

Notes:

  • If you use a specific chatbot integration, copy the provided MCP server configuration snippet from the integration’s docs and replace the URL with https://mcp.data.gouv.fr/mcp.
  • The hosted endpoint at mcp.data.gouv.fr/mcp is public and requires no authentication for standard MCP tooling.

Additional notes

  • The hosted endpoint is https://mcp.data.gouv.fr/mcp. If you run the server yourself, point clients to your own URL.
  • This MCP server is designed to be compatible with a wide range of clients by providing standard MCP server configurations (e.g., http transport).
  • Some clients (like certain premium ChatGPT plans) may have additional connector steps; follow the individual tool’s MCP integration docs for exact UI steps.
  • For local testing, using npx mcp-remote https://mcp.data.gouv.fr/mcp is typically sufficient to establish a connection.
  • When testing with sensitive data or restricted datasets, ensure you comply with data.gouv.fr usage terms and any applicable access controls.

Related MCP Servers

Sponsor this space

Reach thousands of developers