Skip to content
Docs

Case Studies Overview

The guides section documents each BifrostQL feature in isolation. This section does the opposite: each page here takes a realistic, simulated scenario — a company, a legacy database, a set of constraints — and walks it end to end, showing how the features compose into a shipped system.

The companies and databases are invented, but every configuration block, metadata rule, query, and component in these pages is real and runnable against the current release.

Meridian Supply Co. has a ten-year-old WPF order-management app over SQL Server. The desktop app stays exactly as it is; BifrostQL puts a browser-based admin surface next to it — schema-driven UI, audit columns, and hidden internals — without a single stored procedure or migration.

Features exercised: metadata rules, visibility: hidden, audit populate columns, the embeddable @standardbeagle/edit-db editor, local-user auth.

Brightline Bookings runs one admin portal with two very different rooms in it. Support staff get a curated admin built on the application-shaped GraphQL API — soft deletes, guarded mutations, tenant rules all enforced server-side. The two on-call engineers get an ops console with the raw SQL escape hatch — role-gated, row-capped, and time-boxed.

Features exercised: _rawQuery and raw-sql-* metadata, roles, soft delete, mutation guardrails, custom fetcher auth, building both surfaces in one React app.

Fieldstone sells field-service management to landscaping companies. Every customer’s data lives in one database, separated by a tenant_id column. This walkthrough builds the customer-facing back office where tenant isolation, soft delete, and audit columns are enforced by the server — the client code never mentions a tenant.

Features exercised: tenant-filter, auto-filter with bypass roles, soft delete with deleted-by, local auth with tenant + roles columns, per-table page-size defaults.

Each case study follows the same arc:

  1. The situation — the fictional company, its database, and its constraints.
  2. The plan — which BifrostQL features map to which constraint.
  3. The walkthrough — server setup, metadata configuration, and client code, in the order you’d actually build it.
  4. What we didn’t have to build — a checklist of the code the approach avoided.

If a step uses a feature you haven’t met, each page links back to the relevant guide the first time it appears.