Scan Website

The scan command fetches a public URL, measures the total page size, and audits the page's i18n and SEO health. It produces a scored report (0–100) covering HTML attributes, canonical links, hreflang tags, robots.txt, sitemap.xml, localized internal links, and JavaScript bundle locale weight.

No extra dependencies are required. When puppeteer is installed the scan can capture lazy-loaded JavaScript chunks for a more precise bundle analysis; otherwise it falls back to inspecting eagerly-loaded scripts declared in the HTML.

Usage

npx intlayer scan <url>

Example

npx intlayer scan https://example.com

Sample output:

🔍 Scanned https://example.com (basic mode)

Score: 90/100
Page size: 10.60 MB (HTML 42.31 KB)
Locales: en, fr, es, de, …

Checks:
  ✓ html lang attribute
  ✓ html dir attribute
  ✓ canonical link
  ✓ hreflang tags
  ✓ x-default hreflang
  ✓ localized internal links
  ⚠ all internal links localized
  ✓ current locale detected
  ✓ robots.txt present
  ✓ robots.txt keeps locale paths crawlable
  ✓ sitemap.xml present
  ✓ sitemap lists every locale
  ✓ sitemap has alternate links
  ✓ sitemap has x-default

Bundle locale weight:
  Translations shipped: 120.50 KB
  Unused (other locales): 45.20 KB (37%)

Options

<url> (required)

The fully-qualified URL to scan (e.g. https://example.com).

--no-deep

Disable the deeper render-based scan.

By default the command attempts to use puppeteer to render the page in a headless browser, capture lazy-loaded JavaScript chunks, and measure the true wire-transfer size. If puppeteer is not installed, the command automatically falls back to basic mode.

Pass --no-deep to force basic mode even when puppeteer is available.

Example: npx intlayer scan https://example.com --no-deep

--json

Output the full scan result as a JSON object instead of a formatted report. Useful for programmatic consumption or CI pipelines.

Example: npx intlayer scan https://example.com --json

Standard configuration options

• --base-dir — Base directory used to locate the intlayer.config.* file.
• -e, --env — Target environment (e.g. development, production).
• --env-file — Path to a custom .env file.
• --no-cache — Disable configuration cache.
• --ci — Run the command in every Intlayer project of the monorepo (or only the current one when run from a project directory). Per-project credentials can be injected via INTLAYER_PROJECT_CREDENTIALS, a JSON map of project path to { "clientId", "clientSecret" }.
• --verbose — Enable verbose logging (default in CLI mode).
• --prefix — Custom log prefix.

What is checked

The final score is the weighted sum of all passing checks expressed as a percentage (0–100).

Using the scan function programmatically

The scan function is also exported from @intlayer/cli so it can be called from your own scripts:

import { scan } from "@intlayer/cli";

await scan("https://example.com", {
  deep: false,
  json: false,
});

For lower-level access, scanWebsite from @intlayer/engine/scan returns a structured ScanResult object:

import { scanWebsite } from "@intlayer/engine/scan";

const result = await scanWebsite("https://example.com", { deep: false });
console.log(result.score, result.totalPageSize, result.events);