Maritime Vessel Data — MCP Server

A Model Context Protocol (MCP) server that exposes maritime vessel data as standard MCP tools and a resource. Any MCP-compatible client — Claude Desktop, an agent framework, an IDE extension — can query live vessel data through the protocol instead of a bespoke integration.

Built with Python and the official mcp SDK (FastMCP). It is the same tool surface used by the companion maritime-vessel-agent project, published here as a reusable, protocol-standard server.

What it demonstrates

Related MCP server: vessel-traffic-mcp

Tools

Resource vessels://all returns the full dataset.

Setup

cd maritime-mcp-server
python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt

Verify it works (offline)

python -m src.smoke_test

This exercises every tool without needing an MCP client.

Run the server

python -m src.server        # serves over stdio (how MCP clients launch it)

Use it from Claude Desktop

Add the server to your claude_desktop_config.json (~/Library/Application Support/Claude/claude_desktop_config.json on macOS):

{
  "mcpServers": {
    "maritime-vessel-data": {
      "command": "/absolute/path/to/maritime-mcp-server/.venv/bin/python",
      "args": ["-m", "src.server"],
      "cwd": "/absolute/path/to/maritime-mcp-server"
    }
  }
}

Restart Claude Desktop, then ask it questions like "Which tankers are at anchor near Port Klang?" — it will call the server's tools directly.

Extending it

• Point data/vessels.json at a live AIS feed.

• Add tools (route ETA, anchorage occupancy) — clients discover them automatically.

• Add authentication and switch to the HTTP/SSE transport for remote clients.