SignalK MCP Server is an MCP server that provides AI assistants with read-only access to SignalK marine data systems. It connects large language model interfaces to live marine telemetry, allowing boaters, marine technicians, navigators, and maritime software developers to inspect real-time navigation parameters, monitor vessel health, and track surrounding traffic. Through an internal V8 isolate sandbox execution engine, the server allows AI models to execute sandboxed JavaScript code to aggregate, compute, and filter data directly on the host rather than dumping raw JSON payloads into the LLM context window. This architecture enables agents to query vessel position, heading, speed over ground, wind velocity, active system alarms, and nearby AIS targets while reducing token usage by up to ninety-six percent. The server supports customizable network parameters like SignalK host address, port, and transport layer security settings, ensuring communication remains local and read-only. Users can generate structured situation reports, evaluate navigational risks, monitor critical threshold alerts, and discover dynamic SignalK paths without modifying underlying boat instrumentation.
Category: Maps, Weather & Local Data
Tags: ais, marine, navigation, signalk, vessel
npx signalk-mcp-server or install it globally with npm install -g signalk-mcp-server. 3. Open your Claude Desktop configuration file (located at ~/Library/Application Support/Claude/claude_desktop_config.json on macOS). 4. Register the server under the mcpServers object: json { "mcpServers": { "signalk": { "command": "npx", "args": ["signalk-mcp-server"], "env": { "SIGNALK_HOST": "localhost", "SIGNALK_PORT": "3000", "SIGNALK_TLS": "false" } } } } 5. Restart Claude Desktop to enable the marine data connection.Part of MCP Servers
SignalK MCP Server is an open source Model Context Protocol server that bridges AI assistants with SignalK marine data systems. It allows language models to retrieve read-only telemetry from vessel sensors, such as GPS coordinates, boat speed, wind conditions, and AIS targets. By executing JavaScript inside sandboxed V8 isolates, the server filters marine metrics prior to returning responses, preserving context limits.
You can install SignalK MCP Server by running npm install -g signalk-mcp-server or invoke it directly using npx signalk-mcp-server. It requires Node.js 18.0.0 or higher and an accessible SignalK instance. To integrate with MCP clients like Claude Desktop, add the server command and environment variables for your SignalK host and port to your client configuration file.
No, SignalK MCP Server provides strictly read-only access to marine telemetry. The execution sandbox and the underlying SignalK client do not expose write operations, autopilot controls, or steering configuration endpoints. This ensures that connected AI clients can monitor vessel metrics, inspect alarms, and observe AIS targets without interfering with onboard navigation instruments.
SignalK MCP Server works with any client supporting the Model Context Protocol, including Claude Desktop, Cursor, and custom agent architectures. Setup requires defining the npx execution command and passing standard environment variables indicating the SignalK server host address, port, and transport protocol options.
SignalK MCP Server is open source software distributed under the permissive MIT License. You can review the complete source code, inspect security settings in the V8 isolate sandbox, adapt the code execution engine for custom marine setups, and contribute updates through its public GitHub repository without proprietary licensing fees or restrictive commercial terms.