SQL editor

The SQL editor is a full query workspace for when you want to talk to your data directly. It’s read-only by design: SELECT and WITH statements only.

Tabs and sources

Work in multiple query tabs — each tab keeps its own SQL buffer and its own data source, chosen from the toolbar’s source selector. Tabs restore when you switch back, and the editor warns before closing the browser with unsaved work.

Writing and running

The editor is Monaco-based (the VS Code editor) with SQL keyword completion and word wrap; Format (Ctrl+Shift+F) tidies the statement.
Run (Ctrl+Enter) executes the buffer; long queries are cancellable mid-flight.
Execution info reports rows and time (N rows in N ms) and flags when results were capped by the server’s row limit.

The schema browser

The toggleable left panel walks source → schema → table → column. Click any name to insert it at the cursor; search to jump; re-run schema discovery if the source’s metadata is stale.

Results

A paginated grid (100 rows per page) with a row-count badge.
CSV export downloads the result set.
The Visualize tab auto-builds a chart from the result’s categorical and numeric columns — a quick sanity look before you commit the query to a dashboard.

AI assistance

Explain turns the current SQL into a plain-English description.
AI Generate goes the other way: describe what you want in natural language and preview the generated SQL, then Accept or Reject it.

Saving queries

Save stores the query — name, description, tags, source binding — into the workspace’s saved-query library, from which it can be re-run, shared with the team, or dropped onto a dashboard as a live SQL card.