Join beta

Scout Examples

Beta-safe workflows you can run today.

Start with small, inspectable runs. Scout is strongest when every record can point back to source evidence and every failure leaves a useful artifact trail.

Workflows

Pick one concrete job, then inspect the folder Scout writes.

Page to markdown

Use this when you need one public page captured as readable text with source metadata.

mkdir -p ./scout-runs/example-page
scout scrape https://example.com \
  > ./scout-runs/example-page/scrape.json

Product category to records

Use this for a category or listing page when you want product records plus blocked-page evidence.

scout products books \
  --site books.toscrape.com \
  --start-url https://books.toscrape.com/ \
  --output-dir ./scout-runs/books-products

Company intelligence packet

Use this for a small company packet with overview, key URLs, and citable source pages.

scout run company \
  --query Adobe \
  --mode auto \
  --workdir ./scout-runs/adobe-company

Captured HTML to records

Use saved mode when you already have HTML or browser-captured evidence and want reproducible extraction.

scout run research \
  --mode saved \
  --query ./captures/page.html \
  --workdir ./scout-runs/captured-html

Artifact contract

Every useful run should leave evidence behind.

records.json

Structured records created from the run. Product records can be prepared for search ingestion.

source_pages.json

Source registry with URLs, providers, fetched timestamps, content hashes, and evidence pointers.

blocked_pages.json

Blocked, sparse, or failed pages with provider attempts and user-facing reasons where available.

extraction_report.md

A human-readable summary of what worked, what failed, and what should be reviewed next.

Exports

Move records into the tool that needs them.

Product records to JSONL, CSV, SQLite, Google Sheets, and Algolia is the intended handoff path. Start with preview and validation before pushing any hosted or production index. 

scout product-export \
  ./scout-runs/books-products/records.json \
  --format jsonl \
  --output-dir ./scout-runs/books-products/exports

Boundaries

What these examples do not claim.

No guaranteed hard-site bypass

Browser and saved-evidence modes improve visibility, but Scout does not promise access to every protected site.

No certified legacy /app UI claim

The beta website, CLI, HTTP API, artifacts, and hosted limits are the current product surface for launch claims.