feat(skills): add databricks-lakebase-migration skill

Captures the Provisioned → Autoscaling migration mechanics that aren't in
the public docs as of 2026-05. Covers:

- pg_dump/pg_restore workflow with the --role flag for SP ownership
- Why raw CREATE DATABASE breaks app SP OAuth (missing databricks_auth
  + neon extensions) and how to recover
- databricks_create_role(<sp-uuid>, 'SERVICE_PRINCIPAL') as the proper
  way to register an app SP for OAuth-token resolution
- The Apps API update-mask limitation when re-pointing database
  resources via bundle deploy, and the direct apps update --json
  workaround
- Synced-table snapshot semantics (UC sync pipelines do not auto-follow)
- Step-by-step runbook plus a common-issues table covering every
  failure mode hit during a real lakemeter migration

Also updates databricks-skills/README.md to list the existing
databricks-lakebase-autoscale skill (was missing) and the new migration
skill, and adds them to install_skills.sh.

Co-authored-by: Isaac

Author: David O'Keeffe

databricks-skills/README.md

@@ -105,7 +105,9 @@ cp -r ai-dev-kit/databricks-skills/databricks-agent-bricks .claude/skills/
 - **databricks-app-python** - Python web apps (Dash, Streamlit, Flask) with foundation model integration
 - **databricks-python-sdk** - Python SDK, Connect, CLI, REST API
 - **databricks-config** - Profile authentication setup
-- **databricks-lakebase-provisioned** - Managed PostgreSQL for OLTP workloads
+- **databricks-lakebase-provisioned** - Managed PostgreSQL for OLTP workloads (legacy fixed-capacity model)
+- **databricks-lakebase-autoscale** - Next-gen managed PostgreSQL with autoscaling, scale-to-zero, branching
+- **databricks-lakebase-migration** - Migrate from Lakebase Provisioned → Autoscaling via pg_dump/pg_restore
 
 ### 📚 Reference
 - **databricks-docs** - Documentation index via llms.txt