Scan database tables directly — no CSV export needed.

Install 

pip install goldencheck[db]  # adds connectorx

Usage 

# Postgres
goldencheck scan-db "postgresql://user:pass@host/db" --table orders

# With domain pack
goldencheck scan-db "postgresql://..." --table patients --domain healthcare

# Custom SQL query
goldencheck scan-db "snowflake://..." --query "SELECT * FROM orders WHERE date > '2024-01-01'"

# JSON output
goldencheck scan-db "bigquery://project" --table events --json

# HTML report
goldencheck scan-db "postgresql://..." --table users --html report.html

Supported Databases

Any database supported by connectorx:

  • PostgreSQL
  • MySQL
  • SQLite
  • SQL Server
  • Oracle
  • Snowflake
  • BigQuery
  • Redshift
  • Trino/Presto

Falls back to SQLAlchemy + pandas if connectorx is not available.

How It Works

  1. Fetches rows from the database (respects --sample-size, default 100K)
  2. Writes to a temp CSV
  3. Runs the full GoldenCheck profiler pipeline
  4. Returns findings + profile (same as file scanning)

Python API 

from goldencheck.engine.db_scanner import scan_database

findings, profile = scan_database(
    "postgresql://user:pass@host/db",
    table="orders",
    sample_size=50000,
    domain="ecommerce",
)