Tenali AI Docs
Sources & Knowledge Base

Add URL Sources

Point Tenali at a docs site, help center, or PDF on the web and let it read the content — including JavaScript-rendered pages.

URL sources are the fastest way to give Tenali a body of content you don't want to maintain by hand — your public docs, a help center, release notes, or a partner's specification.

Add a Site

  1. 1

    Open Sources, then the Sites tab

    Click Add Site to open the Add New URL Source dialog.

  2. 2

    Fill in the four fields

    • Name — how the source appears in your list.
    • URL — the page to start from.
    • Crawl the whole site — tick this to follow links from that page; leave it unticked to read only the single page you gave.
    • Visibility — whether this is personal to you or shared with the workspace (admins only).
  3. 3

    Watch the status pill

    The source shows PENDING, then IN PROGRESS, then COMPLETED. Its content becomes answerable at COMPLETED.

What Tenali Can Read

  • Regular web pages — HTML, read as text.
  • PDFs on the web — point the URL straight at a PDF and Tenali parses it as a document, so tables and page structure survive. If you crawl a site that links to PDFs, those get parsed too. This means a docs portal whose real content is in linked PDFs now works.
  • JavaScript-rendered pages — documentation sites that build their content in the browser are rendered before reading. These used to come back empty without saying why.

Anything else — a video, an image, an archive, a download — is skipped, and the source tells you which URLs were skipped and why rather than failing silently.

When a URL Source Fails

A FAILED source shows the reason directly under its row in the Sources list. Common ones:

  • The page couldn't be reached — it's behind a login, blocked to automated readers, or the URL is wrong. Tenali can only read pages that are publicly reachable.
  • Nothing readable on the page — the URL resolved but held no extractable text.
  • Unsupported content type — the URL pointed at a file format Tenali doesn't parse.

Fix the cause and refresh the source from its row menu.

Best Practices

  • Use stable, canonical URLs. A URL that redirects around or changes structure will drift out of date.
  • Start narrow. Crawl a docs section rather than a whole marketing site; low-value pages (logins, legal boilerplate, blog indexes) dilute answers.
  • Refresh after a docs overhaul. Tenali reads a site when you add it — it doesn't re-crawl on its own.
  • Prefer the integration where one exists. For Notion, SharePoint, OneDrive or Google Drive, connect the app rather than crawling a public URL — you get authenticated access and better structure.

Was this page helpful?