Use the CLI

Install and sign in before running these examples.
They use scrappycoco; if setup printed a full launcher path, use that path.
With Node.js 20+ and npm, you can also use npx --yes @scrappycoco/cli@latest.

1. Find a capability

scrappycoco catalog list --available --json
scrappycoco catalog inspect web.extract_content --json

Inspect the input schema and provider options before building your request.
The live catalog also shows availability and pricing evidence.

2. Run it and save the result

scrappycoco run web.extract_content \
  --input '{"url":"https://example.com"}' \
  --provider firecrawl_scrape \
  --provider-options '{"firecrawl_scrape":{"formats":["markdown"],"only_main_content":true}}' \
  --output result.json --json

Open result.json to read records, provider attempts, and usage.
Check that the content meets your needs. See Read results and costs
for errors, missing fields, and truncation.

3. Reuse a request file

For larger requests, save the body as request.json:

{
  "input": {"url": "https://example.com"},
  "providers": ["firecrawl_scrape"],
  "provider_options": {
    "firecrawl_scrape": {"formats": ["markdown"], "only_main_content": true}
  }
}

scrappycoco run web.extract_content --file request.json --output result.json --json

Use the plural providers in a file. On the command line, repeat --provider
to set an ordered fallback list. Run tries the next provider after an error or
empty result and stops when one returns records. It reports every attempt;
you still check whether the records meet your needs. If you omit providers,
Run uses the catalog's deterministic default order.

--limit sets the requested count or collection target, with a default of 10.
Provider limits vary, and complete returned pages are preserved. Put limit
at the top level of a request file, outside input.

Next steps

For connection problems, run scrappycoco doctor --json and follow
installation troubleshooting.