Skip to content

Query Explorer

The Query Explorer is where you write ad-hoc queries in OQL, Hexcovery's SQL-inspired query language, and see the result as a chart or table instantly. It's the free-form counterpart to the built-in pages.

Writing a query

The editor is a full code editor with OQL syntax highlighting and autocomplete that suggests tables, columns, and functions as you type. Run with Ctrl+Enter, save with Ctrl+S, cancel a running query with Esc.

Don't know the syntax yet? Use "Ask AI to generate your query" — describe what you want in plain language and the AI turns it into OQL, populating the editor for you to run or tweak. See the OQL reference for the language itself.

Visualization

Hexcovery picks the chart type automatically from the shape of your query — time series, table, single value, or bar chart — and you can switch it from the visualization dropdown. Time-series charts support a synced crosshair tooltip and drag-to-zoom.

Toggle "Show SQL" to peek at the ClickHouse SQL your OQL compiles to — useful for understanding exactly what's being run.

Time range

The global time-range picker doesn't just filter — it retargets the query. Changing the picker rewrites the LAST or BETWEEN window in your query to the new range (and adapts an explicit bucket size to match), so the picker drives a query that already carries a window. This keeps the editor and the picker in agreement.

History, saving, and export

  • Query history keeps your recent queries (per user) so you can re-run or recover one.
  • Save a query to reuse it and put it on a dashboard — see Saved queries.
  • Export the result as CSV or JSON.
  • The Explorer accepts a ?q=<oql> deep-link: /explorer?q=... pre-loads the editor, runs the query, and syncs the time-range picker to the window the query carries. These links are shareable.
  • Charts on the built-in pages and on dashboards carry a ⋮ menu with "Open in Explorer", which opens that chart's underlying query here so you can modify it. (It appears only where the chart has a faithful OQL equivalent.)