Advertise
MCP.so
Sign In

MCP Server for Snowflake

@dynamike

Repository

Basic information

Category

Databases

License

MIT

Transports

stdio

Author

dynamike

About MCP Server for Snowflake

MCP Server for connecting to Snowflake with read-only questions

#snowflake#modelcontextprotocol

MCP Server for Snowflake config

No standard config provided

This server doesn't expose a parseable MCP config block in its README. See the repository for install instructions.

Repository

MCP Server for Snowflake tools

5

List all accessible Snowflake databases

List all views in a specified database and schema

Get detailed information about a specific view including columns and SQL definition

Query data from a view with an optional row limit

Execute custom read-only SQL queries (SELECT, SHOW, DESCRIBE, EXPLAIN, WITH) with results formatted as markdown tables

Overview

What is MCP Server for Snowflake?

A Model Context Protocol (MCP) server enabling Claude to securely perform read-only queries against Snowflake databases. It supports flexible authentication via service account (private key) or external browser login, and includes connection pooling with automatic background refresh.

How to use MCP Server for Snowflake?

Install Python 3.12+ and the uv package manager. Clone the repository, install with uv pip install -e ., configure a .env file with Snowflake credentials, then run as a stdio-based MCP server with uv run snowflake-mcp. Connect to it from Claude Desktop via Settings → MCP Servers.

Key features of MCP Server for Snowflake

  • Read-only enforcement: SELECT, SHOW, DESCRIBE, EXPLAIN, WITH only
  • Private key or external browser authentication
  • Automatic connection pooling with configurable refresh interval
  • Input validation and automatic LIMIT clauses to prevent large results
  • Stdio transport for easy Claude Desktop integration

Use cases of MCP Server for Snowflake

  • List all accessible databases and schema objects
  • Describe view structure, columns, and SQL definitions
  • Query sample data from views with optional row limits
  • Execute custom read-only SQL and receive markdown table results
  • Compare data across different databases and views in a single session

Frequently asked questions

1

What authentication methods are supported?

Private key authentication for service accounts, or external browser authentication for interactive user login sessions.

2

What are the required dependencies?

Python 3.12+, the uv package manager, and a Snowflake account with either a configured service account or regular user account.

3

Can I modify data with this server?

No. The server strictly enforces read-only operations and blocks INSERT, UPDATE, DELETE, and other modification statements.

4

How can I adjust connection pool refresh timing?

Set the `SNOWFLAKE_CONN_REFRESH_HOURS` environment variable (default 8 hours) to control how often connections are refreshed.

Comments

More Databases MCP servers

mcp-server-qdrant: A Qdrant MCP server

qdrant

An official Qdrant Model Context Protocol (MCP) server implementation

637

ClickHouse MCP Server

ClickHouse

Connect ClickHouse to your AI assistants.

818

Sqlite

modelcontextprotocol

Model Context Protocol Servers

88.1K

PostgreSQL Model Context Protocol (PG-MCP) Server

stuzero

540

Lawstronaut

Lawstronaut-FZCO

Access millions of official legal and regulatory source documents from 155+ jurisdictions via MCP. Lawstronaut supports AI agents with structured legal research, legal data, laws, regulations, cases, guidance, official s

Advertise