Skip to main content
Vosaire
All docs
Knowledge & Sources

Managing and retraining sources

How to view, edit, delete, and resync your knowledge sources — and what to do when a source shows an error.

Over time you'll add sources, some will go stale, some will error, and some will need updating. Here's how to manage all of that.

Viewing your sources

Go to your agent → Knowledge tab. You'll see a table of all sources with:

  • Source name and type
  • Status (Pending / Processing / Indexed / Error)
  • Number of chunks it created
  • Last sync time (for sources with auto-sync)
  • Action buttons

You can search by name and filter by type using the controls above the table.

Editing a source

Click the edit icon (pencil) next to any source to update its settings:

  • Name — rename it for your own reference
  • Sync frequency — how often it should re-crawl (for URL and integration sources)
  • For CSV sources — update the column mapping

You can't change the source URL or replace an uploaded file. For those changes, delete and re-add.

Deleting a source

Click the actions menu (three dots) next to a source and select Delete. The source and all its indexed chunks are removed immediately. Any questions that relied on that content will no longer be answerable until you replace the source.

Forcing a re-index

For sources that support syncing, you can trigger an immediate re-index by clicking the sync icon next to the source. This is useful when you know the content has changed and don't want to wait for the next scheduled sync.

To re-index all sources at once, use the Re-index all button at the top of the Knowledge tab. Note: this is disabled if any source is currently processing.

Fixing a source that shows Error

When a source shows "Error", hover over the status badge to see the error message. Common causes:

For URL sources:

  • The page returned a 404 or 403
  • The page requires login to access
  • The site blocked the Vosaire crawler (some sites do this)

For sitemaps:

  • The sitemap URL is wrong or the file isn't valid XML

For documents:

  • The file is a scanned PDF (image-only, no extractable text)
  • The file is password protected
  • The file is corrupted

In most cases, the fix is to delete the source and add it again with the correct URL or file. If a URL source keeps erroring and you can access the page normally, try adding it as a text block by copying the page content manually.

Keeping sources fresh

For static content that rarely changes (your pricing page, your refund policy), a manual re-index whenever you update the page is fine.

For content that changes regularly (a blog, a changelog, a help center), set up automatic sync:

  • Sitemaps support daily/weekly sync
  • Integration sources (Notion, Google Drive, GitHub, Zendesk) also support scheduled sync

When content changes frequently and you're not syncing automatically, the agent may give outdated answers. If users are getting stale information, check your sync settings first.

Ready to put this into practice?

Free plan available. Be up and running in under 10 minutes. No credit card required.

Start free — no credit card