What problem does it solve? Building and deploying Python web apps on Databricks requires navigating framework selection, OAuth authorization models, resource wiring, and deployment configuration. This Skill provides the complete workflow so apps connect correctly to SQL warehouses, Lakebase, and model serving endpoints without hardcoded credentials or misconfigured app.yaml files. ## Core Features & Use Cases - Framework Guidance: Covers Dash, Streamlit, Gradio, Flask, FastAPI, and Reflex with Databricks-specific patterns, correct app.yaml commands, and port configuration. - Authorization Setup: Implements both service principal (app) auth via SDK Config() and on-behalf-of user auth via the x-forwarded-access-token header. - Resource & Data Connectivity: Wires SQL warehouses, Lakebase PostgreSQL, model serving endpoints, secrets, and volumes using the valueFrom pattern. - Deployment: Supports Databricks CLI, Asset Bundles, and MCP tool workflows with log-based verification. - Use Case: A data scientist wants to turn an analysis into an internal Streamlit dashboard backed by a SQL warehouse, deploy it to Databricks Apps, and enforce per-user row-level access through Unity Catalog. ## Quick Start Ask the agent to build a Streamlit dashboard on Databricks Apps that queries a SQL warehouse and deploy it with user authorization enabled.
