Skip to content
CSA Loom — the Microsoft Fabric experience for Azure tenants where Fabric isn't yet available: lakehouses, warehouses, notebooks, semantic models, Activator rules, Data Agents, across Commercial, GCC, GCC-High, and DoD IL5

Loom Lakehouse Editor — Fabric-parity spec

Reference: lh_bronze Lakehouse in casino-fabric-poc (F64), captured 2026-05-26 by the catalog agent (run a4f93461062e0e80c).

What's there in Fabric

Page chrome

  • Page title shows the Lakehouse name (lh_bronze)
  • Standard Fabric workspace breadcrumb + capacity badges + global action bar (search, account, settings, notifications)

Explorer sidebar (left)

  • Header: lh_bronze with collapse arrow
  • Tables section — expandable, lists every Delta table in the lakehouse:
  • bronze_compliance, bronze_financial, bronze_player_profile, bronze_security, bronze_slot_telemetry, bronze_table_games
  • Each table has an icon + name + (on hover) a kebab menu
  • Files section — ADLS Gen2 file browser
  • "Add lakehouses" search input — lets you add additional lakehouses as data sources to view side-by-side
  • Bottom: ability to add shortcut / Get data

Main pane — Ribbon

Button Behavior
Open notebook Opens an existing notebook attached to this lakehouse OR creates a new one already wired up with the lakehouse as default attachment
Add to data agent ▼ Wires the lakehouse to a Copilot Studio data agent. Dropdown picks which agent.
Manage OneLake security Opens permissions pane: role assignments on the lakehouse (Read, Write, Build reports, etc.)
Update all variables Refreshes variable libraries attached to the lakehouse

Main pane — body

  • View toggle: "Table view" (current) — switchable to other views via dropdown
  • Selected table data grid (e.g. bronze_compliance):
  • Header: column names sortable
  • Row count: "Showing 1000 rows"
  • Up to 22 rows visible, scrollable
  • Columns: filing_id, filing_type, filing_timestamp, amount, player_id, compliance_status, property_code, player_name, sSN_hash, sSN_masked, etc.
  • Info banner at the top: "A SQL analytics endpoint for SQL querying was created with this item." + link → Switch to query with T-SQL (loads the auto-paired SQL endpoint editor)

Auto-pairing pattern

Every Fabric Lakehouse comes with a sibling SQL Analytics Endpoint item automatically created on provisioning. The two items share the same Delta files but expose different access modes: - Lakehouse: Spark-native, table preview, file browser, append/overwrite via notebook - SQL Analytics Endpoint: T-SQL read-only access via TDS, same underlying Delta tables

In the Fabric workspace tree, they appear as two siblings with the same display name but different icons (Lakehouse vs SQL endpoint).

Right-click context menu (observed)

  • Open in notebook
  • New shortcut
  • Open SQL endpoint
  • Properties
  • Refresh
  • Maintain (Delta vacuum/optimize)
  • Delete

What Loom already has

  • ✅ ADLS Gen2 container + path browse (/api/lakehouse/containers, /api/lakehouse/paths)
  • ✅ Synapse Serverless SQL query backend (real T-SQL works)
  • ✅ Sample preview (currently text-based, not full grid)

What's missing for parity

  1. Auto-paired SQL endpoint sibling — when user creates a Lakehouse in Loom, automatically also create a synapse-serverless-sql-pool item pointed at the same ADLS path. Both appear in the workspace tree.
  2. Ribbon command bar — add Open notebook / Add to data agent / Manage OneLake security / Update all variables / Maintenance buttons
  3. Real Delta table list — currently Loom doesn't enumerate Delta tables under a Lakehouse; needs to scan the ADLS path for _delta_log/ and surface them
  4. Per-table data grid with sortable columns + row count (not just text preview)
  5. "Open in notebook" — drops user into the Notebook editor with the Lakehouse pre-attached as default source + a sample read cell
  6. "Switch to query with T-SQL" toggle — switches the editor to the paired SQL endpoint without leaving the page
  7. OneLake security pane — RBAC at the Lakehouse / Table level (maps to ADLS RBAC + Synapse SQL grants)
  8. Right-click context menus on tables: Open in notebook · New shortcut · Properties · Refresh · Maintain · Delete
  9. Add lakehouses search in Explorer — pin multiple lakehouses side-by-side

Build plan

Phase Work
Backend New /api/lakehouse/[id]/tables → scans ADLS for _delta_log/ markers, returns table list with row counts (Synapse Serverless COUNT(*)). New /api/lakehouse/[id]/preview?table=X → SELECT TOP 1000 via Serverless SQL with column types. New /api/lakehouse/[id]/pair-sql-endpoint (POST) → creates the sibling synapse-serverless-sql-pool item in Cosmos pointing at same ADLS path. New /api/lakehouse/[id]/onelake-security (GET/PUT) → ADLS RBAC + Synapse SQL grants.
Frontend Replace flat preview with DataGrid component (sortable, column types). Add ribbon command bar. Wire "Open in notebook" → creates notebook + pre-attaches Lakehouse via the existing attached-sources mechanism. Add right-click context menus on tables. Add Add-lakehouse search box in Explorer.
Auto-pairing Update /api/workspaces/[id]/items POST for itemType=lakehouse to also auto-create a sibling synapse-serverless-sql-pool item in the same workspace, with state.adlsPath matching.

Estimated effort

1-2 focused sessions (Backend first, then Frontend, then E2E test).