What problem does it solve?
This Skill resolves failures setting up or using Lakebase for agent memory by guiding you through the required dependencies, Databricks resource configuration, environment variables, and critical Postgres permission grants.
Core Features & Use Cases
- End-to-end Lakebase configuration: Covers both provisioned and autoscaling Lakebase instance models, including the correct
databricks.yml resource stanza and required env wiring.
- Memory-ready setup for agent templates: Targets agent short-term and long-term memory use cases (LangGraph templates) and the common operational errors users hit when memory is enabled.
- Permission-error troubleshooting: Provides a concrete fix for migration/CREATE TABLE and “permission denied” scenarios by granting the app’s service principal the needed Postgres access.
- Operational runbook: Includes troubleshooting guidance for embedding dimension mismatches, missing store tables, unresolved instance names, and failed connections.
Quick Start
Ask your AI assistant to run through the Lakebase setup for your agent memory by telling it whether you use provisioned or autoscaling Lakebase and what values you have for instance name or project/branch/database id.