How does Dbmcp ensure database security?

It features default read-only mode, strict SQL validation, single-statement enforcement, dangerous function blocking, credential redaction, and configurable CORS/trusted hosts to protect your databases.

Dbmcp is a single-binary Model Context Protocol (MCP) server designed to securely connect AI assistants and agents to MySQL, MariaDB, PostgreSQL, and SQLite databases without external runtime dependencies.

Which databases can Dbmcp connect to?

Dbmcp offers comprehensive multi-database support, allowing connection to MySQL, MariaDB, PostgreSQL, and SQLite databases, all managed efficiently from one lightweight binary.

What are the key benefits of using Dbmcp for AI-database integration?

Dbmcp is a ~7 MB single-binary with zero external runtime dependencies, providing 9 standard MCP tools and robust security features like default read-only mode and SQL validation, ideal for AI agents.

Does Dbmcp require external dependencies like Docker or Python?

No, Dbmcp is packaged as a single, lightweight (~7 MB) binary with no external runtime dependencies, simplifying installation and deployment across various environments.

Dbmcp: Connect AI Agents to MySQL, PostgreSQL, SQLite

Dbmcp is a lightweight, single-binary Model Context Protocol (MCP) server engineered to seamlessly connect AI assistants with a variety of SQL databases, including MySQL, MariaDB, PostgreSQL, and SQLite. With zero runtime dependencies, this ~7 MB binary offers a robust and secure bridge for AI agents to perform complex database operations, from schema retrieval and read-only queries to write operations (when enabled). It simplifies the integration of AI capabilities with existing database infrastructure, providing essential tools for data interaction in a secure and efficient manner.

Características Principales

01Multi-database support for MySQL, MariaDB, PostgreSQL, and SQLite from one binary

02Offers 9 standard MCP tools including `listDatabases`, `readQuery`, `writeQuery`, and `getTableSchema`

03Packaged as a single, lightweight binary (~7 MB) with no external runtime dependencies

04Supports multiple communication transports: stdio for client-managed processes and HTTP for remote/multi-client scenarios

05Features a robust security suite with default read-only mode, SQL validation, and credential redaction

0610 GitHub stars

Casos de Uso

01Developing AI tools that require zero-dependency integration with local or remote database instances

02Enabling AI-powered applications to perform secure and validated database operations across multiple SQL backends

03Connecting AI assistants and agents to SQL databases for dynamic data interaction and retrieval