Snowflake

You run a Snowflake stored procedure integration as an MCP server to manage and execute Snowflake procedures from an MCP environment. This server exposes stored procedures from your Snowflake schemas so you can invoke them through MCP workflows, enabling streamlined data operations and automation.

How to use

You connect to this MCP server from an MCP client to access and run Snowflake stored procedures. Specify the schemas and procedures you want exposed, then invoke those procedures as part of your MCP workflows. The server handles the translation between MCP calls and Snowflake procedure executions, so you can orchestrate data tasks across Snowflake without writing separate integration code.

How to install

Prerequisites you need before installation:

• Python 3.8 or higher
• uv installed

Step 1: Set up the project locally by obtaining the integration module and preparing the environment.

# Step 1: Get the integration module
# This step assumes you will clone or download the integration package folder

Step 2: Create a Python virtual environment and activate it.

uv venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

Step 3: Prepare configuration with Snowflake credentials. Create a .env file in the root directory and define all required Snowflake values.

SNOWFLAKE_ACCOUNT=<xxx>-<xxx>
SNOWFLAKE_USER=<xxx>
SNOWFLAKE_PASSWORD=<xxx>
SNOWFLAKE_ROLE=<xxx>
SNOWFLAKE_WAREHOUSE=<xxx>
SNOWFLAKE_DATABASE=<xxx>
SNOWFLAKE_SCHEMA=<xxx>
SNOWFLAKE_HOST=<xxx>-<xxx>.snowflakecomputing.com

Step 4: Configure the MCP server entry. Add an MCP configuration that points to the MCP runtime and selects the schemas and procedures you want to expose.

{
  "mcpServers": {
    "snowflake_sp_server": {
      "command": "absolute/path/to/your/uv/uv.exe",
      "args": [
        "--directory",
        "absolute/path/to/snowflake-mcp-sp-integration",
        "run",
        "main.py",
        "--schemas",
        "DB_NAME.SCHEMA_NAME",
        "--procedures",
        "DB_NAME.OTHER_SCHEMA_NAME.PROCEDURE_NAME"
      ]
    }
  }
}

Additional setup notes

Keep your environment variables secure and do not commit the .env file. Ensure the schemas and procedures you expose are properly scoped for your operations.

Additional sections

Security and deployment considerations: Use separate Snowflake credentials for development and production, enforce least privilege on roles, and rotate credentials regularly. For production, consider enabling MCP client authentication and auditing of Snowflake procedure calls.