Skip to Content

Web crawler

The Web Crawler tab indexes pages from a website into your Content Store, so an agent can retrieve knowledge from a public site alongside your own documents.

By the end of this page you will have set up a crawler, controlled its scope, and know how to run and manage it.

Prerequisites

  • Admin or Developer role.
  • A Content Store open on the Web Crawler tab.

Create a crawler

Click New Crawler to open the four-step wizard (Create Web Crawler - Step N of 4).

Step 1 — Name and URL

Enter a Crawler Name and the Website URL to start from.

Web crawler wizard step 1 with crawler name and website URL

Step 2 — Crawl settings

Set the scope and behavior:

  • Crawl Depth — how many link levels deep to follow.
  • Maximum Pages — a cap on how many pages to index.
  • Respect robots.txt — honor the site’s crawl rules.
  • Render JavaScript — use a headless browser for JavaScript-heavy sites (single-page apps).
Web crawler wizard step 2 with crawl depth, maximum pages, robots.txt, and render JavaScript

Step 3 — URL patterns (optional)

Add Include Patterns to restrict the crawl to matching URLs, and Exclude Patterns to skip sections you don’t want.

Step 4 — Review

Review the configuration and create the crawler.

Run and manage a crawler

Each crawler shows controls and live progress:

  • Start / Pause — run the crawl or pause it. Progress shows the percentage complete, the current URL, and pages crawled and indexed.
  • Expand a crawler to see its crawled pages, each with an indexing status; remove a page if you don’t want it indexed.
  • Use a crawler’s menu to Edit Settings or Delete Crawler.
Web crawler running with live progress and a list of crawled pages
💡

Start with a shallow depth and a modest page cap, run it, then check the Search tab to confirm the store returns useful results before widening the crawl.

Common issues

  • Fewer pages than expected — the crawl stopped at the Maximum Pages cap or Crawl Depth limit, or your Include/Exclude Patterns filtered pages out.
  • A JavaScript-heavy page came back empty — enable Render JavaScript so the crawler loads the page in a headless browser.
  • Pages you didn’t want were indexed — add an Exclude Pattern and reindex.

Next

  • Search — test what the store returns.