pilat/mcp-datalink

Overview

pilat/mcp-datalink is an MCP server that provides secure database access to PostgreSQL, MySQL, and SQLite databases. It implements the Model Context Protocol to enable large language models and other MCP clients to interact with these databases through a standardized interface.

Best for

Best for
Developers who need a lightweight, standardized MCP server for secure database access in AI agent or LLM workflows

Use cases

• Connecting AI assistants to relational databases for natural language queries
• Standardizing database access across multiple MCP-compatible applications
• Performing read and write operations on PostgreSQL, MySQL, or SQLite via a single protocol

How to use

Install

npx @pilat/mcp-datalink

Tools exposed

• list_databases
• list_tables
• describe_table

Tested with

Claude Desktop, Claude Code, Cursor, Cline, VS Code

Example client config

{\n  "mcpServers": {\n    "datalink": {\n      "command": "npx",\n      "args": ["-y", "@pilat/mcp-datalink"],\n      "env": {\n        "DATALINK_ANALYTICS_URL": "postgresql://user:password@localhost:5432/analytics",\n        "DATALINK_ANALYTICS_READONLY": "true",\n        "DATALINK_INVENTORY_URL": "mysql://user:password@localhost:3306/inventory",\n        "DATALINK_CACHE_URL": "sqlite:///path/to/cache.db"\n      }\n    }\n  }\n}

Notes

pilat/mcp-datalink is an MCP server that provides secure database access to PostgreSQL, MySQL, and SQLite databases. It implements the Model Context Protocol to enable large language models and other MCP clients to interact with these databases through a standardized interface.

3 stars on GitHub. Last updated 2026-04-08. Licensed MIT.

Use cases

• Connecting AI assistants to relational databases for natural language queries
• Standardizing database access across multiple MCP-compatible applications
• Performing read and write operations on PostgreSQL, MySQL, or SQLite via a single protocol

Pros

• Supports three major relational databases (PostgreSQL, MySQL, SQLite)
• TypeScript codebase offers type safety and broad developer familiarity
• Uses the MCP standard for easy integration with many LLM tools

Cons

• Very low GitHub stars indicate minimal community adoption and limited support
• Only relational databases are supported, no NoSQL or cloud data sources
• Requires an MCP client to be used, adding a dependency on a still-maturing ecosystem

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.