pggm-mcp-snowflake-server 0.1.1

Custom Model Context Protocol server for Snowflake

PGGM MCP Snowflake Server

A customized Model Context Protocol (MCP) server for Snowflake integration, allowing AI assistants to interact with Snowflake databases securely and efficiently.

Features

• Connect to Snowflake databases and execute queries
• Support for various SQL operations and schema exploration
• Data insights collection and memoization
• SQL write operation detection for enhanced security
• Customizable database, schema, and table filtering
• Support for authentication through environment variables or command-line arguments

Installation

Using pip

pip install pggm-mcp-snowflake-server

Tools Available to AI Assistants

The server provides the following tools for AI assistants:

• list_databases - List all available databases in Snowflake
• list_schemas - List all schemas in a database
• list_tables - List all tables in a specific database and schema
• describe_table - Get the schema information for a specific table
• read_query - Execute a SELECT query
• append_insight - Add a data insight to the memo
• write_query - Execute an INSERT, UPDATE, or DELETE query (if --allow_write is enabled)
• create_table - Create a new table in the Snowflake database (if --allow_write is enabled)

Security

By default, the server runs in read-only mode. To enable write operations, you must explicitly pass the --allow_write flag.

The server uses SQL parsing to detect and prevent write operations in read_query calls, ensuring only approved write operations can be executed.

Development

Setup Development Environment

git clone https://github.com/yourusername/pggm-mcp-snowflake-server.git
cd pggm-mcp-snowflake-server
uv venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate
uv pip install -e ".[dev]"