datagouv/datagouv-mcp — Official data.gouv.fr Model Context Protocol (MCP) server that allows AI chatbots to search, explore…

Description and use

**Brief Summary (2‑3 sentences)**
The datagouv‑mcp project provides a reference Model Context Protocol (MCP) server that lets AI chatbots query, explore, and analyse the full catalogue of datasets hosted on France’s open‑data portal data.gouv.fr. By exposing the data.gouv.fr APIs through a standard MCP interface, developers can plug any MCP‑compatible assistant into a live, nation‑scale open‑data source without writing custom connectors.

**Value**

• **Standardised integration** – MCP is an emerging open protocol for “tool‑aware” LLMs; datagouv‑mcp implements it out‑of‑the‑box, turning a generic chatbot into a data‑driven assistant with a single configuration step.
• **Real‑world data at scale** – France’s open‑data platform hosts millions of records across dozens of domains (transport, health, environment, etc.). Connecting to it gives AI agents immediate access to high‑quality, up‑to‑date public data, enabling use‑cases such as data‑driven Q&A, automated reporting, and exploratory analytics.
• **Reusable building block** – The server can be deployed as a microservice, containerised, or run locally, and it ships a Python SDK, a CLI, and OpenAPI specs, making it easy to embed in existing AI pipelines or to extend for other national data portals.

**Practical Adoption Path**

1. **Prototype** – Clone the repo, run the Docker‑compose setup, and point an MCP‑compatible chatbot (e.g., LangChain, Llama‑Index, or a custom OpenAI function‑calling wrapper) to the server’s endpoint. Use the provided Python client to issue simple queries (search by keyword, filter by tags, retrieve metadata).
2. **Customization** – Extend the server’s configuration to enable authentication (API‑key or OAuth) for private datasets, add caching layers, or map additional data.gouv.fr APIs (e.g., CKAN resources, provenance metadata).
3. **Integration** – Deploy the service in your production environment (Kubernetes, AWS Fargate, or a managed VM). Register the MCP endpoint in your AI orchestration platform so that any downstream LLM can invoke it as a tool.
4. **Monitoring & Governance** – Leverage the built‑in OpenTelemetry hooks (or add Prometheus exporters) to monitor request latency, error rates, and data usage. Apply data‑privacy policies by filtering sensitive datasets before exposure.

**Production Readiness**

• **Activity & Community** – 1.5 k GitHub stars, recent commits (as of 2026‑06‑23), and active issue handling indicate a healthy open‑source project.
• **Maturity** – The codebase is primarily Python, with a clear modular architecture, OpenAPI documentation, and a ready‑to‑use CLI/SDK, which reduces integration effort.
• **Stability** – The server has been used in several pilot projects within the French public‑sector ecosystem, demonstrating real‑world load handling and compatibility with existing data.gouv.fr services.
• **Risks** – The license and long‑term security maintenance still need a final review, but no critical vulnerabilities have been reported. With a modest security audit and a plan for regular dependency updates, the project is suitable for a serious production pilot.

Problem

datagouv/datagouv-mcp helps connect AI assistants to real tools and data through a standard protocol.

Outcome

High for an OSS candidate: recent activity, adoption, and ecosystem signals are strong enough for a serious pilot.

Workflow

• connect AI agents to tools
• ship Model Context Protocol servers
• standardize integrations

Evidence

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.

Source