Database MCP Server

Name: Database MCP Server
Author: MCP-Mirror

@MCP-Mirror

代码仓库

关于 Database MCP Server

Mirror of

#database #mcp-server #data-management

基本信息

分类

数据库

运行时

python

传输方式

stdio

发布者

MCP-Mirror

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "georgi-terziyski_database_mcp_server": {
      "command": "python",
      "args": [
        "-m",
        "db_mcp_server"
      ]
    }
  }
}

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is Database MCP Server?

A Model Context Protocol (MCP) server that provides tools for connecting to and interacting with various database systems, including SQLite, PostgreSQL, MySQL/MariaDB, and SQL Server, through a unified interface.

How to use Database MCP Server?

Install from source with pip install -e ., configure via environment variables or a JSON config file, then run python -m db_mcp_server (MCP mode) or python -m db_mcp_server.web_server (standalone web server). Use the provided MCP tools for connection management, query execution, schema management, and transaction control.

Key features of Database MCP Server

Multi-database support: SQLite, PostgreSQL, MySQL/MariaDB, SQL Server
Unified interface for common database operations
Schema management: create, alter, drop tables and indexes
Query execution: raw SQL or structured query tools
Transaction support: begin, commit, rollback
Standalone web server mode for use with any LLM

Use cases of Database MCP Server

Allow an LLM (e.g. Claude) to query and manipulate a database via MCP
Provide a REST API for other LLMs (e.g. Llama 3) to perform database operations
Manage database schemas and run migrations through natural language prompts

FAQ from Database MCP Server

What databases does it support?

It supports SQLite (included in Python), PostgreSQL (via psycopg2-binary), MySQL/MariaDB (via mysql-connector-python), and SQL Server (via pyodbc).

How do I configure database connections?

You can set environment variables (DB_CONFIG_PATH for a JSON file or DB_CONNECTIONS for inline JSON), or provide connection details at runtime using the add_connection tool.

Can I use this server with LLMs other than Claude?

Yes. Run as a standalone web server (python -m db_mcp_server.web_server) and any LLM can call the /list_tools and /call_tool HTTP endpoints.

What are the runtime requirements?

Python 3.8 or higher, SQLAlchemy, and optional database-specific drivers. No external runtime dependencies beyond the Python ecosystem.

How do I run the server?

Use python -m db_mcp_server for MCP mode (default), or python -m db_mcp_server.web_server for web mode. You can specify a config file with --config and set logging level with --log-level.