Load Data Skill

Commands for working with scan data in the Site Scanning Engine.

Prerequisites

• API and database must be running
• Infrastructure services (Docker) must be up

Core Data Operations

Ingest Federal Domain List

Load websites from Federal Website Index into the database:

npm run ingest -- --limit 200

Options:

• --limit N - Limit to N websites (recommended for local dev)
• Omit limit to ingest all domains (production use)

Source: https://github.com/GSA/federal-website-index/blob/main/data/site-scanning-target-url-list.csv

Note: This is a prerequisite for all scanning operations. Must be run before enqueueing scans.

Enqueue All Websites for Scanning

Queue all websites in database for scanning:

npx nest start cli -- enqueue-scans

This adds jobs to the Redis queue for the scan-engine workers to consume.

Scan Single Site (Testing)

Test scan a single site without writing to database:

npx nest start cli -- scan-site --url 18f.gov

Use cases:

• Testing scan logic locally
• Debugging specific site issues
• Validating scan configuration

Note: Results are printed to console but not persisted to database.

Create and Export Snapshot

Generate snapshot and export to S3:

npm run snapshot

What this does:

• Exports current scan results to CSV
• Uploads to S3/Minio for public consumption
• Only includes latest scan per website

Note: Snapshot generation is difficult to test locally because Minio configuration is limited in local dev. If issues occur, address them in the deployed environment.

Data Flow

1. Ingest → Loads URLs into website table
2. Enqueue → Creates scan jobs in Redis queue
3. Scan → Workers process jobs, write to core_result table
4. Snapshot → Exports results to S3

Database Notes

• Database stores only most recent scan per website (no historical data)
• Old scan results are overwritten with new ones
• Historical data is preserved via snapshots in S3