KSQL — Konduit SQL

KSQL is Konduit's SQL dialect. It runs against your workspace's DataTables and powers:

Dashboard widgets — drive metrics, charts, and tables
Routines — email a query result on a recurring schedule
The SQL Console — ad-hoc queries with infinite scroll and CSV export
Embedded query blocks — drop a live result into any task description or comment

KSQL is a SELECT-first subset of standard SQL. Use SELECT, JOIN, GROUP BY, WHERE, ORDER BY, LIMIT, subqueries, and most common functions exactly as you would in any SQL database. UPDATE and DELETE are also supported — they run behind a confirmation step (a dry-run preview shows what will change before you commit) and require the right role. INSERT, DROP, ALTER, and other schema/dangerous operations are blocked — see Limitations for the full list.

The visual query builder and the SQL text editor are two interfaces to the same engine — they produce identical results.

Scope: KSQL queries DataTables — your relational tables (customers, inventory, tickets, line items, anything you build). Konduit's Tasks system has its own filtering, views, and dashboards, but isn't queryable via SQL today. If you want a query-driven view over task-like data, model it as a DataTable.

Where to run queries

The Query Console is built into the Konduit app — open it from anywhere with:

Keyboard shortcut: Ctrl + J (Windows / Linux) or Cmd + J (Mac)
Top navigation: click the Query button (terminal icon) in the topnav

It slides up from the bottom as a dock — paste any of the examples on this site, hit Run, and the results appear with infinite scroll + CSV export.

The Query Console in SQL mode: a SELECT / WHERE / ORDER BY query above its live results grid, with CSV export

Quick example

SELECT Name, Status, Revenue
FROM Customers
WHERE Status = 'active' AND Revenue > 50000
ORDER BY Revenue DESC
LIMIT 10

KSQL is read-first. SELECT runs for anyone with query access. UPDATE (member or higher) and DELETE (admin or higher) run in the same Query Console — but each first shows a dry-run preview of exactly which rows will change and waits for your explicit confirmation before committing. INSERT and all schema operations (DROP, ALTER, CREATE, …) are blocked outright — add rows through the table UI, a form, or the API. See Limitations for the full list.

What's covered

Section	Topic
Query Basics	SELECT, FROM, WHERE, ORDER BY, LIMIT, DISTINCT, DESCRIBE
Joins	INNER JOIN, LEFT JOIN, multi-table, aliases
Grouping & Aggregation	GROUP BY, HAVING, COUNT, SUM, AVG, MIN, MAX, GROUP_CONCAT
Operators	Comparison, logical, arithmetic, array (CONTAINS)
Expressions	Arithmetic, string concat (`\|\|`), CASE WHEN
Subqueries	`IN (SELECT …)`, scalar subqueries, derived tables, depth-3 nesting
Functions	String, math, date, conditional, array — full reference
Special Columns	Person columns, Relation columns, Tags/CONTAINS
Identifiers & Escaping	Quoting names, escaping strings, comments
Recipes	Practical examples by category
Limitations & Errors	What's not supported + error message reference

Result limits

Every query returns a maximum of 1000 rows, even if LIMIT is set higher. For larger datasets, paginate with OFFSET or apply filters to narrow results. CSV exports from the SQL Console have a separate higher ceiling — see the in-app docs.

Caching

Query results are cached for 5 seconds. Writes (add, edit, or delete rows) invalidate the cache immediately so subsequent queries see fresh data.

Navigation