Catalog
How the Stackless Catalog helps users and the Agent find connectors, models, tables, dashboards, and Monitors, and trace lineage between them.
What is the Catalog?
The Catalog is the searchable inventory of the data assets your role can see: source connectors, dbt and Transformation Models, warehouse tables, dashboards, and Monitors, connected by lineage. Open it from Data > Catalog.
Use it to answer questions such as:
- What data exists in this workspace, and where does it come from?
- Which table or model contains a field?
- Which dashboards depend on this model?
- What feeds this dashboard, and is any of it stale?
Table view
The header shows the total number of assets and a Table / Graph switch. In Table view, the cards at the top filter the list by asset type:
| Card | Lists | Key columns |
|---|---|---|
| Connectors | Source connectors | Service, destination schema, status, last sync, frequency |
| dbt Projects | Sources, snapshots, and models in the dbt project, as a folder tree | Path and model details |
| Tables | Physical warehouse tables | Database, schema, type, rows, last altered, columns |
| Dashboards | Dashboards | Domain filter |
| Monitors | Monitors, where enabled | Domain filter |
Use the search box to filter the current list by name.
Graph view
Graph shows the same assets as a lineage map, from connectors through models to dashboards. Pan and zoom with the controls or minimap, and click any node to open its details.
Asset details
Click an asset to open its detail drawer:
- Its type, metadata, and description.
- Load preview to see up to 10 sample rows, for tables and models.
- Columns or members, with the tests defined on each column.
- Upstream and Downstream tabs showing what the asset depends on and what depends on it.
- Shortcuts such as Open in Transformation Models, Open in Semantic Models, Open dashboard, or Open Monitor.
Lineage
Upstream lineage answers "what does this depend on?" Downstream lineage answers "what will be affected if this changes?"
A typical chain:
Connector -> source tables -> Transformation Models -> Semantic Model -> Dashboard or Monitor
Check downstream lineage before changing or unpublishing a Transformation Model, so you know which Semantic Models, dashboards, and Monitors will need review.
Catalog and the Agent
The Agent searches the Catalog before it queries or builds anything. That lets it find governed models instead of guessing table names, trace how a dashboard KPI is computed, check freshness, and stay within your role's access.
Useful prompts:
- "Search the Catalog for revenue."
- "Find columns named customer_id."
- "Trace upstream from the Northwind Revenue Performance dashboard."
- "What depends on fct_northwind__orders?"
- "Is any source behind this dashboard stale?"
Keeping the Catalog current
The header shows when the Catalog was Last refreshed. Most changes appear automatically: publishing a Transformation Model adds it to the Catalog, and unpublishing removes it.
Administrators can click Full Refresh to re-sync every data source, including the warehouse, which may take a few minutes. A Full Refresh also runs when schema selections change in Settings > Agent Data Access.
If something you just created is missing:
- Confirm it was published, not left as a private draft.
- Confirm its schema is enabled in Agent Data Access and your role can read it.
- Ask an administrator to run Full Refresh, then search again by exact name.
Access limits
The Catalog is access-aware. Assets in schemas your role cannot read, and dashboards your role is not assigned, do not appear for you or for the Agent working on your behalf. See Security and Access.