44.1 Opening Knowledge Sources

From the Agent workspace, select Knowledge sources.

The Sources page is the operational area for adding and maintaining content the Agent should learn from.

44.2 Adding a Single Web Page

The Add a page card accepts a website URL.

The interface explains that AIUNIFY Chat will fetch the page, extract its text, and index it for the Agent.

  1. Open the Agent's Sources page.
  2. Paste the complete page URL.
  3. Leave Source type as Single URL.
  4. Select Add & crawl.

44.3 URL Format

Use a complete HTTP or HTTPS URL, for example:

https://your-site.com/about

Use the actual public page that contains the information the Agent should know.

44.4 What Happens After Adding the URL

The Source is created in a pending state and a crawl is queued.

As processing continues, the Source can move through statuses such as Pending, Crawling, Indexed, or Failed.

44.5 Choosing the Right Page

Good individual URL Sources include pages containing:

  • Product or service information.
  • Pricing information.
  • Frequently asked questions.
  • Policies.
  • Documentation.
  • Contact or process information visitors commonly ask about.

44.6 Avoiding Unhelpful Pages

Do not add a page merely because it exists on the website. Prefer pages whose content should genuinely influence Agent answers.

Pages dominated by login forms, account controls, checkout experiences, or other non-informational content are generally poor Knowledge Sources.

44.7 Monitoring the Crawl

After adding the Source, review the Existing Sources list.

Use Refresh to retrieve the latest crawl status when processing finishes in the background.

44.8 Previewing Extracted Content

Select the Preview control to open What the AI sees for this source.

Review the extracted Documents and Chunks to verify that the page produced meaningful text.

44.9 Reindexing a URL

Select Reindex when the page has changed or when the first crawl/index attempt did not produce usable content.

44.10 Single URL vs. Sitemap

Use Single URL for one specific page. Use Sitemap only when the URL points to a real sitemap and you intentionally want the crawler to process multiple URLs from it.

Write Your Comment