tdengine-mcp

tdengine-mcp

5
databases

The TDengine Query MCP Server is a tool designed to facilitate read-only data queries from a TDengine database using the Model Context Protocol (MCP), providing AI assistants with database access for data investigation and exploration. It emphasizes security by limiting operations to read-only queries and provides support for various AI tools through easy configuration.

TDengine Query MCP Server

A Model Context Protocol (MCP) server providing read-only TDengine database queries for AI assistants.

Key Features

  • Executes read-only TDengine queries including SELECT, SHOW, and DESCRIBE commands.
  • Provides database information and metadata.
  • Lists available databases and tables.

Limitations

  • Does not support write operations such as INSERT, UPDATE, DELETE, CREATE, and ALTER.
  • Not intended for database administration or schema management.

Supported AI Tools

  • Cursor IDE: Set up in .cursor/mcp.json
  • Anthropic Claude and other MCP-compatible AI assistants.

How to Use

  • Run from Source Code: Clone the repository and run using uv without installation.
  • Install from PyPI: Use pip to install and run.
  • Install from Smithery: Use npx command.

Configuration

  • Supports configuration using .env file, specifying environment variables like LOG_LEVEL, TDENGINE_HOST, etc.

Integration with AI Assistants

  • AI assistants can execute data queries and retrieve information using the TDengine MCP tools: query, info, and databases.

Security

  • Only read-only queries are allowed for security.

Troubleshooting

  • Includes suggestions for connection issues and common errors.

Contributing

  • Contributions are welcome via pull requests.