Get started with agridata MCP in 15 minutes

Expected outcome

By the end of this guide, your usual LLM client will be connected to the agridata server and you will have run a first query that combines structured data and documentary references. Estimated total time: 15 minutes.

Prerequisites

• An MCP-compatible LLM client: Claude Desktop, Claude Code, VS Code, Cursor, Windsurf, Gemini CLI, or similar.
• An internet connection.
• For Claude Desktop only: Node.js 18 or later installed on the machine. Other clients do not have this dependency.

Step 1 — Identify the server URL

The public server URL is https://mcp.tanitdata.org/mcp. The trailing /mcp path is mandatory — a URL without this suffix returns a 404 error.

Step 2 — Configure the client

Add the configuration snippet for your client. The logic is the same everywhere; only the file format changes.

Claude Desktop

Configuration file:

• Windows: %APPDATA%\Claude\claude_desktop_config.json
• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Linux: ~/.config/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "tanitdata": {
      "command": "npx",
      "args": ["mcp-remote", "https://mcp.tanitdata.org/mcp"]
    }
  }
}

Claude Desktop does not yet support remote MCP servers natively: a small bridge called mcp-remote is used, downloaded automatically on first launch through npx. That is why Node.js 18+ is required. Save the file, then restart Claude Desktop.

VS Code (with MCP support)

Create or edit .vscode/mcp.json at the project root:

{
  "servers": {
    "tanitdata": {
      "url": "https://mcp.tanitdata.org/mcp",
      "type": "http"
    }
  }
}

Cursor, Windsurf, Gemini CLI, other native-MCP clients

{
  "mcpServers": {
    "tanitdata": {
      "url": "https://mcp.tanitdata.org/mcp"
    }
  }
}

Claude Code (CLI)

A single command, no configuration file:

claude mcp add --transport http tanitdata https://mcp.tanitdata.org/mcp

Step 3 — First test call

In your client, start a new conversation and ask:

"What tools are exposed by the tanitdata server?"

The client should display a list of tools provided by the server. If that list appears, the connection works. Otherwise, verify that the URL ends with /mcp and that the configuration was saved before the client was restarted.

Step 4 — Your first real query

Try a question that combines numbers and documentary context:

"What is the production of durum wheat by governorate over the past five years, and how does it compare to the objectives of the national plan?"

The server responds with the matching data and, for each result, points to the source resource on agridata.tn. You can follow the link to verify the source directly on the portal.

Troubleshooting

• The client reports a 404 error or cannot find the server. Check that the URL is https://mcp.tanitdata.org/mcp with the /mcp suffix. The bare https://mcp.tanitdata.org is not enough.
• Claude Desktop fails to connect and shows no clear error. Two checks: the configuration should go through npx mcp-remote rather than a direct url field, and Node.js 18+ must be installed on the machine.
• The first climate query takes about 7 seconds; later queries are instant. This is expected: the server warms a local cache on the first call. Subsequent queries are fast.

What's next

Once the server is running on the client side, browse the research index for more advanced use cases. To share feedback or propose a collaboration, the Get Involved page describes the pathways.

Frequently asked questions

Can I test the server without installing a client?

The server is designed to be called from an LLM client, not from a browser. To check that it is running, open https://mcp.tanitdata.org/health in a browser — a {"status": "ok"} response confirms the server is online.

Which MCP clients are supported?

All clients implementing the standard MCP protocol: Claude Desktop, Claude Code, VS Code with an MCP extension, Cursor, Windsurf, Gemini CLI, and others. The only difference is the configuration file format; the logic is the same everywhere.

Are queries billed?

No. Access to the server is free for research use. Heavy programmatic use should run a local copy — the code is open under an MIT license.

Where do I report a bug or request a dataset?

Open an issue on https://github.com/tarekddoit/tanitdata or write to contact@tanitdata.org.