How we put 16 database engines behind one result grid

Description and use

LibreDB Studio unifies 16 database engines in a single browser IDE, normalizing results and execution plans for consistent display.

Problem

Providing a unified, browser‑based IDE that can query and display results from 16 different database engines with varying result shapes and EXPLAIN formats.

Outcome

A single browser‑based IDE that normalizes diverse database outputs into one grid and plan view, simplifying multi‑engine development and reducing credential exposure.

Workflow

• 1. Deploy LibreDB Studio (Docker, Helm, or npm).
• 2. Configure connections to desired engines via drivers.
• 3. Studio’s provider layer flattens each engine’s native response into a list of rows with ordered maps.
• 4. The grid discovers columns by sampling pages for document stores, marking absent fields explicitly.
• 5. Cells carry engine‑native and normalized type tags; grid formats by category and shows native type on hover.
• 6. For EXPLAIN, each engine’s native output is parsed into a common plan node structure and rendered as a collapsible tree.
• 7. Transport security (SSH tunnels, TLS) is handled centrally and mapped to each driver’s config.
• 8. RBAC, SSO, and audit trail are enabled on the shared server to protect credentials and log queries.
• 9. Users interact with the unified UI, schema diff, ER diagrams, and NL2SQL features.
• 10. Feedback is collected via GitHub issues for normalization improvements.

Evidence

LibreDB Studio is an MIT‑licensed, self‑hosted database IDE that runs in the browser.

Source

The internal representation is deliberately dumb: every result is a list of rows, every row is an ordered map from column name to a typed cell.

Source

We defined one plan node: an id, a label, a small set of common metrics…

Source