thingspanel
This MCP server integrates ThingsPanel IoT platform with AI models like Claude, GPT, and others that support the Model Context Protocol.
claude mcp add --transport stdio thingspanel-thingspanel-mcp python -m thingspanel_mcp \ --env THINGSPANEL_API_KEY="Your ThingsPanel API Key" \ --env THINGSPANEL_BASE_URL="Your ThingsPanel base URL (e.g., https://demo.thingspanel.cn/)"
How to use
The ThingsPanel MCP Server provides a natural language interface to your ThingsPanel IoT platform. It exposes device information, telemetry, and control capabilities so AI models or assistants can query device data, monitor status in real time, and issue control commands through natural language prompts. After starting the server, you can interact with it via the CLI or through integrations like Claude Desktop by supplying your API key and base URL. The server translates natural language requests into ThingsPanel API calls, returning structured results that your agent can reason with or present to users.
Key capabilities include: querying device status and telemetry, listing devices, turning devices on/off or adjusting settings, retrieving device information and history, and generating platform-wide insights. By using the CLI or configured integrations, you can feed prompts to the MCP that internally handles authentication and communication with the ThingsPanel API, abstracting away the underlying REST calls for smooth AI-driven workflows.
How to install
Prerequisites:\n- Python 3.8+\n- Internet access to install Python packages from PyPI\n- Pip available in your Python environment\n\nInstallation steps:\n1) Create a virtual environment (recommended):\nbash\npython -m venv venv-thingspanel\nsource venv-thingspanel/bin/activate # macOS/Linux\nvenv-thingspanel\Scripts\activate # Windows"""\n````\n2) Install the MCP package from PyPI:\nbash\npip install thingspanel-mcp\n\n3) Run the MCP server:\n```bash\npython -m thingspanel_mcp # uses the console script from the package if installed\n\n4) (Optional) If you prefer invoking via the CLI directly in environments with a visible entry point, you can also install the executable wrapper and start it with your credentials. For Claude Desktop or other integrations, ensure the config points to the same invocation method.\n\nPrerequisite notes:\n- Ensure your API key is generated in ThingsPanel and has the required permissions\n- Know your ThingsPanel base URL (for cloud or on-prem deployments)\n- If you upgrade, re-check CLI flags, as the project may update argument names.
Additional notes
Environment variables are recommended to avoid exposing credentials in command history. If you deploy behind a proxy or VPN, confirm that the ThingsPanel API endpoint is reachable from the host running the MCP server. Some platforms may require additional network permissions or API whitelisting. The MCP can be wired into AI assistants like Claude Desktop or other tooling by configuring the mcpServers entry with the appropriate command and arguments. If you encounter authentication errors, verify that THINGSPANEL_API_KEY is correct and not expired, and that THINGSPANEL_BASE_URL points to the correct platform instance. For debugging, run the server with verbose logging if available in your installed CLI (e.g., --verbose) to capture request/response details.
Related MCP Servers
mcp-vegalite
MCP server from isaacwasserman/mcp-vegalite-server
github-chat
A Model Context Protocol (MCP) for analyzing and querying GitHub repositories using the GitHub Chat API.
nautex
MCP server for guiding Coding Agents via end-to-end requirements to implementation plan pipeline
pagerduty
PagerDuty's official local MCP (Model Context Protocol) server which provides tools to interact with your PagerDuty account directly from your MCP-enabled client.
futu-stock
mcp server for futuniuniu stock
mcp -boilerplate
Boilerplate using one of the 'better' ways to build MCP Servers. Written using FastMCP