trino-dbt-platform

What problem does it solve?

This Skill solves the problem of reliably running dbt projects in production on the Trino query engine while correctly handling Iceberg table layouts, incremental updates, and performance-friendly configuration.

Core Features & Use Cases

• Production-ready dbt profiles.yml for Trino: Supports multiple authentication methods (ldap, kerberos, oauth, jwt, certificate) and includes practical session/query settings for stable execution.
• End-to-end dbt materializations on Iceberg: Covers full-refresh outputs and incremental patterns (append, merge, delete+insert), plus table/view/materialized_view/ephemeral usage with Iceberg-specific properties (format, partitioning, sorted_by, location).
• Incremental correctness and lifecycle management: Provides SCD2 snapshots via timestamp strategy, recommends Iceberg-compatible TIMESTAMP(6) handling, and includes an ANALYZE post-hook to keep cost-based optimization effective.
• Faster CI pipelines: Implements slim dbt CI using state:modified+ with --defer and prior production artifacts to avoid unnecessary work.

Quick Start

Configure your dbt profiles.yml for Trino, set Iceberg properties in your model config blocks, then run dbt in CI using state:modified+ with --defer to build only changed models.