Your first report.
Without the setup spiral.

1. Choose a sample report

Start with the SEO example. It describes a supplied documentation page and contains three illustrative findings. The file is public and does not require an API key.

Download the SEO sample JSON, or read it directly from the same origin in JavaScript. The example below assumes this site is served over HTTP or HTTPS, rather than opened as a file on your computer.

// Read a bundled sample — no API key needed.
async function readSample() {
  const response = await fetch(
    "/assets/samples/seo-analysis.json"
  );
  if (!response.ok) {
    throw new Error(`HTTP ${response.status}`);
  }
  const report = await response.json();
  console.table(report.findings);
  return report;
}
readSample().catch(console.error);

# Python 3 standard library; static sample.
import json
from urllib.request import urlopen
from urllib.error import URLError

url = ("https://analyzeapi.com/assets/"
       "samples/seo-analysis.json")
try:
    with urlopen(url, timeout=15) as response:
        report = json.load(response)
    for finding in report["findings"]:
        print(finding["title"])
except (URLError, ValueError, KeyError) as err:
    print(f"Unable to read sample: {err}")

# Download a public example, not a live scan.
curl --fail --show-error --location \
  --max-time 15 \
  https://analyzeapi.com/assets/samples/seo-analysis.json \
  --output sample-report.json

# Inspect the downloaded report.
python3 -m json.tool sample-report.json

2. Check the response

The JavaScript example checks the HTTP result before parsing JSON. Next, check the structure against the supplied schema in your own application. Parsing proves that a response is JSON; schema validation checks its structure. Neither step establishes that the source evidence is correct.

Keep sample: true visible when rendering an illustrative report. Avoid changing the target label to a visitor’s website while leaving the example findings unchanged.

3. Render a useful finding

Use finding.title for the short description, finding.evidence for what was observed, and finding.action for the suggested next step. The outcome field is a display category—pass, review, or note—not a calibrated risk probability.

4. Handle the remaining states

Your production consumer should distinguish complete, partial, and failed reports. Keep successfully collected evidence when a batch is incomplete, and explain unavailable data rather than turning it into a zero measurement.

Read the errors and validation guide before attaching automation. To connect a live service, replace the sample source with a provider or implementation you are authorized to use, and follow its actual interface and authentication requirements.