motherduck-migrate-to-motherduck

What problem does it solve?

This Skill helps plan a migration onto MotherDuck by turning a messy “move data and rewrite SQL” request into a sequenced, validated, and rollback-aware execution plan.

Core Features & Use Cases

• Migration planning with decision guardrails: choose between native MotherDuck storage, pg_duckdb, and optional DuckLake posture based on the source and target serving pattern.
• Target discovery and compatibility inventory: confirm whether a MotherDuck server/MCP is active, inspect target layout, and identify incompatible SQL, functions, and data type assumptions before cutover.
• Validation and rollback-first workflow: define source-vs-target checks (row counts, metrics with pct variance, uniqueness, new/deleted/changed records) and map rollback/cutover slices to reduce production risk.
• Native DuckDB user-agent watermarking: when producing an md: connection, it specifies a custom user agent using harness/LLM metadata with safe fallbacks.

Quick Start

Ask your agent: "Create a migration plan to MotherDuck from Snowflake for our analytics workloads, including cutover slices, validation checks with pct variance, and a rollback path; we want to use native storage unless DuckLake is explicitly required."