Enables LLM agents to perform CRUD (Create, Read, Update, Delete) operations on MySQL databases.
Enables LLM agents to perform CRUD (Create, Read, Update, Delete) operations on PostgreSQL databases.
Integrates with SQLAlchemy to provide a unified interface for LLM agents to perform CRUD operations across multiple supported database systems.
Enables LLM agents to perform CRUD (Create, Read, Update, Delete) operations on SQLite databases.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@MCP Database ManagerShow me the schema for the users table"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
MCP Database Manager
An MCP server that enables LLM agents to perform CRUD operations across multiple databases.
Features
Multi-Database Support
Permission Management (Read-only by default)
SQLAlchemy Integration
list_connectionsredacts passwords from connection URLs in its output
Database Support
This MCP server supports multiple databases. You can enable specific database drivers using installation extras:
Extra | Database | Drivers Installed |
| PostgreSQL |
|
| MySQL |
|
| SQL Server |
|
| All of the above | All drivers |
Configuration
The configuration file is located at:
~/.mcp-database-manager/config.yaml(on all platforms)
Example config.yaml:
Cursor Configuration
To use this MCP server in Cursor, add the following to your MCP settings (Settings > Features > MCP):
Option 1: Local Development (Recommended for debugging)
Use this if you have the source code locally and want to test changes immediately.
Note: Replace
/absolute/path/to/mcp-database-managerwith the actual path to your project directory.
Option 2: Using PyPI (Published version)
Use this to run the stable version published on PyPI without cloning the repository.
Troubleshooting
If you see "Module not found" errors, ensure you are using uv run which handles the virtual environment automatically.