What is SQL Query (sql-query-mcp) and its primary purpose?

SQL Query (sql-query-mcp) enables AI assistants like ChatGPT, Codex, and OpenCode to safely perform read-only queries on PostgreSQL and MySQL databases. It acts as a secure intermediary, preventing direct DSN exposure and write operations to protect your data.

How does SQL Query ensure database security and data integrity?

It strictly enforces read-only access, prohibiting all DDL/DML operations (like INSERT, UPDATE, DELETE, DROP). It also includes built-in safety guards such as default row limits, query timeouts, and comprehensive audit logging, preventing AI models from accessing raw connection strings.

Which database systems are supported by SQL Query?

SQL Query provides robust and explicit support for both PostgreSQL and MySQL databases. It features clear namespace semantics, using 'schema' for PostgreSQL and 'database' for MySQL, ensuring stable and predictable routing for queries.

SQL Query | Secure Read-Only SQL for AI Assistants

This Python-based Model Context Protocol (MCP) server provides a secure, read-only interface for AI assistants to interact with PostgreSQL and MySQL databases. It abstracts away sensitive connection details and write permissions, preventing accidental data modification or leakage. The tool emphasizes explicit engine routing and clear namespace semantics, ensuring robust and predictable interactions while offering essential safety features like query limits, timeouts, and auditing. Designed for AI models like Codex, OpenCode, and ChatGPT, it empowers them to perform data exploration, schema introspection, and read-only queries with enhanced security and control.

Características Principales

01Explicit Engine Routing: Requires explicit configuration of database engines (PostgreSQL or MySQL) for stable and predictable routing.

02Comprehensive SQL Tools: Offers functionalities for listing connections, schemas/databases, tables, describing table structures, running `SELECT` queries, explaining query plans, and getting data samples.

03Clear Namespace Semantics: Differentiates between `schema` for PostgreSQL and `database` for MySQL, avoiding ambiguous generic namespaces.

040 GitHub stars

05Secure Read-Only Access: Strictly enforces read-only SQL queries, preventing DDL/DML operations and protecting database integrity.

06Built-in Safety Guards: Includes default row limits, query timeouts, and comprehensive audit logging for enhanced security and oversight.

Casos de Uso

01Allowing AI to list and inspect tables within specific PostgreSQL schemas or MySQL databases.

02Providing a secure, unified gateway for AI assistants to perform read-only data exploration across both PostgreSQL and MySQL instances.

03Enabling AI models to view detailed table structures, including columns, primary keys, and indexes, without direct `information_schema` queries.