What a website source is for
Website Sources bring readable information from public pages into Brain. They are useful for existing service explanations, product information and approved business pages that would otherwise need to be entered manually.
An import is a source record, not a promise that every future change to the page will automatically be reflected in training. Review the saved content and its freshness.
Import a useful page
- Select the website and open Brain → Website Sources.
- Supply the supported public page URL or use the import controls available on the screen.
- Wait for the import result.
- Open the saved source and check the extracted text.
- Remove irrelevant navigation, old promotions or contradictory information through the available review controls.
- Keep the source active only when it is suitable for use.
Prefer focused pages. An individual service explanation is usually easier to evaluate than a large homepage containing several offers, navigation menus and unrelated promotional copy.
When an import is incomplete
Some sites block automated reads, depend on browser scripts or expose very little readable text. Authentication-only pages and private network addresses are not appropriate public import sources.
If important information is missing, add an approved Q&A, term or product record directly instead of repeatedly importing the same unreadable page. Check that the imported text includes the qualifications and conditions that make a claim accurate.
Keep source material trustworthy
A page can contain outdated facts or text that conflicts with another active record. Review the business meaning rather than assuming imported material is correct because it came from your own site.
Use Brain Recommendations to investigate potential conflicts. When resolving a source issue, edit the original source or its active state and then mark the recommendation reviewed where appropriate. The review does not automatically rewrite all saved articles, so check affected drafts separately.