Snowflake Migration Plugin

An AI-powered plugin for Cortex Code that guides you through end-to-end database migration to Snowflake — from connecting to your source database through deployment, data migration, and validation.

Supported sources: SQL Server, Amazon Redshift

Download

Download link: https://snowconvert.snowflake.com/storage/linux/beta/plugins/migration-plugin-pr.zip (https://snowconvert.snowflake.com/storage/linux/beta/plugins/migration-plugin-pr.zip)

Prerequisites

Before starting a migration, ensure you have:

  • Snowflake account with a connection configured in ~/.snowflake/connections.toml

  • Source database accessible from your machine (SQL Server or Redshift)

  • Cortex Code installed. Installation guide here

All other dependencies (uv, scai, Python packages) are installed automatically by the plugin on first launch.

Example Workload

You can use AdventureWorksDW (https://learn.microsoft.com/en-us/sql/samples/adventureworks-install-configure?view=sql-server-ver17&tabs=ssms) as the example source database. You can substitute any SQL Server or Redshift database you have access to — the migration skill adapts to whatever source you connect.

What It Can Do

Quick Start

cortex --plugin-dir /path/to/plugin/

You can get updated information about the skill by saying: “Produce a README.html about what this plugin can help me with. “

Then say: “Migrate my database to Snowflake”
Dependencies (uv, scai) are installed automatically on first launch. The agent detects where you are in the migration and picks up from there.

Skill’s Migration Workflow

Stage

What happens

1. Connect

Set up a connection to your source database (SQL Server or Redshift)

2. Init

Create a migration project

3. Register

Extract DDL/code from the source database, or import local SQL files

4. Convert

Translate source SQL into Snowflake SQL via SnowConvert

5. Assess

Generate assessment reports (deployment waves, object exclusions, dynamic SQL, ETL analysis)

6. Migrate

Deploy objects, migrate data, validate, and fix — wave by wave

What You Can Ask

You don’t need to follow the prescribed order — ask for any capability at any time:

  • “What is the current state?” — progress checklist

  • “Connect to my SQL Server database”

  • “Extract objects from the source”

  • “Convert my source code”

  • “Run a full assessment”

  • “Generate deployment waves”

  • “Deploy tables” / “Migrate data”

  • “Deploy and test the next function”

  • “Search rules for this file” / “Apply all matched rules”

  • “Continue” — picks up where you left off

Troubleshooting

  • Plugin fails on startup — Check plugin/logs/install-dependencies.log.

  • scai not found — Run plugin/hooks/install-dependencies manually, or brew install --cask snowflakedb/snowconvert-ai/snowconvert-ai.

  • Snowflake connection errors — Verify your connection in ~/.snowflake/connections.toml.

  • Agent seems lost — Say “What is the current state?” to reset context.