Review a dataset in Web Studio

The CLI is great for scripted runs, but sometimes you want to see the schema before committing to 500,000 rows. Web Studio is a local FastAPI dashboard that gives you an ERD view, a per-column spec editor, live generation progress over SSE, and a fidelity report — all without leaving localhost.

Problem: you want to inspect your schema visually, tweak a column’s generator before generating, and watch progress in real time — without writing config files.

Prerequisites

• dbsprout installed (pip install dbsprout)
• A schema initialised: dbsprout init --db <url> or dbsprout init --file <schema>
• A modern browser

Steps

1. Start Web Studio

dbsprout serve --port 8420

DBSprout starts a FastAPI server and prints the local URL:

Web Studio running → http://localhost:8420

Open that URL in your browser.

2. Walk the ERD

The landing screen shows an interactive entity-relationship diagram of every table and FK edge DBSprout introspected. Foreign keys are drawn as directed arrows; junction tables appear in the centre. Hover a table to see column types.

3. Tune column specs

Click any table row to open the spec panel for that table. Each column shows the generator DBSprout selected (e.g., email, random_int, choice). To override:

1. Click the generator name next to a column.
2. Pick a different generator from the dropdown or type a custom value list.
3. Click Save spec — the change is written to .dbsprout/cache/ immediately.

4. Start generation and watch SSE progress

Click Generate in the toolbar. DBSprout streams progress events via Server-Sent Events — the UI shows a per-table progress bar and running row count updating in real time. No page refresh needed.

authors     ████████████████████  200 / 200  ✓
books       ████████████████████  1000 / 1000 ✓
order_items ████████░░░░░░░░░░░░  412 / 1000 …

5. Export the output

When generation completes, a summary panel shows integrity validation results (FK satisfaction, uniqueness, NOT NULL). Click Download SQL to export the ordered INSERT files, or Copy path to use the ./seeds/ directory directly.

Result

You reviewed the schema visually, caught a column that would have produced generic strings, fixed its generator in two clicks, and watched 1,000 rows build table-by-table in real time — all without touching the CLI again. The spec change persists in cache for future runs.

For full Web Studio reference, see /docs/web.