Websites
Turn your website into knowledge — discover pages from a sitemap or paste URLs, and keep them fresh with retraining.
Point the agent at your website and it will pull the readable text from each page into its knowledge. Website crawl requires a plan that includes the web-crawl feature.
Adding pages
Open the importer
On Data sources, choose Website and open Add pages.
Discover or paste
Two options:
- From sitemap — enter your domain or sitemap URL and FluentChat discovers pages for you (it tries your sitemap, probes common paths, and follows homepage links).
- Paste URLs — list the exact pages you want, one per line.
Select and import
Pick the pages to add (up to 200 per import). Noisy links like cart, checkout, and login are de-selected for you. Importing starts indexing each page in the background.
The readable article text is extracted from each page — navigation and boilerplate are filtered out — and the source URL is kept so answers can cite it.
Keeping pages current
Website pages are re-fetched when you retrain them. Click Retrain on a page and the agent fetches the live page again and re-indexes it — so updating your site and retraining keeps the agent's answers current.
Retraining is rate-limited to keep things fair; see Training & sync.
Single page vs. many
Adding one URL and adding many both create individual page sources — the only difference is how you add them. Each page can be retrained on its own.