Add Databricks Terraform skill for infrastructure-as-code automation#164
Conversation
|
@subhadip18 thank you for the contribution! Would you be willing to trim the PR? We are looking to reduce context bloat as possible So wherever for instance your PR references other services that we support in other skills, we can chop! |
|
@calreynolds: Since this is for Terraform skill, we do not have any existing skill which cover Terraform specific skills yet. |
|
Hey @subhadip18! Thank you for contributing! We chatted about this with the team and decided we're going to separate out anything infra\networking related from the direct ai-dev-kit. Since our personas here are on-platform developers for the most part, we don't necessarily want to encourage them to handle their company's infra! We think this could work its way nicely into a service like vibe or internal vibecoding project 👍 |
2 participants
Summary
Adds a Databricks Terraform skill to the AI Dev Kit so developers can get AI-assisted generation, validation, and management of Databricks infrastructure using the Databricks Terraform Provider.
What's included
This skill guides:
Workspace deployment — AWS, Azure, and GCP (basic and PrivateLink/Private Link/PSC)
Provider configuration & authentication — PAT, OAuth, service principals, and multi-provider (account + workspace) patterns for all clouds
Unity Catalog — metastore, storage credentials, external locations, catalogs, schemas, and grants
Databricks resources — clusters, jobs, SQL warehouses, notebooks, secrets, cluster policies, Databricks Apps, Mosaic AI Vector Search
IAM & permissions — users, groups, service principals, workspace permissions
Lakebase (managed Postgres) — Classic (databricks_database_instance) and Autoscaling (databricks_postgres_project / databricks_postgres_branch / databricks_postgres_endpoint) with HA, PITR, branching, and suspend-on-idle
Design patterns — modular layout, remote state (S3/Azure Blob/GCS), common pitfalls and fixes