Connect websites and files

Website and file sources provide content technically. They do not assess whether this content is suitable. Responsibility for accuracy, timeliness, release and structure stays with the users in charge.

Website connector

The website connector links an external website to the AI Wiki.

Path in enneo: Wiki → Connected sources

When crawling, enneo reads the connected website and takes over suitable content as knowledge entries. These then appear in the knowledge structure.

A connector is controlled by the following settings:

• Website URL: the website to be connected
• Max pages: limits the number of pages taken over
• Re-crawl frequency: how often enneo reads the website again
• Include URL patterns: only these areas of the website are considered. Empty means: all areas.
• Exclude URL patterns: these areas are skipped, for example login pages, privacy policy, imprint or blog.
• Keep only these page elements / Leave out these page elements: control which parts of a page are taken over, for example only the article area without navigation and footer.
• Crawl images: also takes over images from the website.

Examples of URL patterns:

• /help/* includes content in the help area.
• /faq/* includes FAQ pages.
• /login/*, /privacy or /imprint can be excluded.

When editing a connector, the Automatic re-crawling switch pauses regular crawling. Articles and settings are kept. A manual crawl is still possible.

For stable results, connectors should be configured as precisely as possible. A narrowly defined source is easier to review and usually gives better results than a broadly connected website.

Crawling and timeliness

When crawling, content is read from the external source and made available in enneo. The connector shows a status: running, completed, error or idle. If automatic crawling is paused, the label Paused also appears.

"Re-crawl" reads the source again, "Stop crawling" cancels a running crawl. This way, changes on the external website can be taken over into enneo.

If URL patterns or other connector settings are changed via "Edit source settings", "Save and re-crawl" applies the changes and starts a new crawl straight away.

The quality of the result depends on the external source. If URL structures, navigation or content of the website change, the knowledge structure in enneo can change as well.

Files as a knowledge source

Path in enneo: Wiki → Files

Files make existing documents available centrally and usable for AI-supported functions, for example PDF documents, work instructions or specialist material. Uploaded files are processed, turned into knowledge entries and can then be used in answer suggestions, chat contexts or AI agents, provided they are released for this.

Files should not be used as an unchecked dump. What matters is that content, file name and location are clear. A clear folder structure helps to keep larger knowledge bases traceable.

Managing and reprocessing files

Depending on their permissions, users can upload, move, check, replace, edit or archive files.

Files are added via "Upload files" or "Upload folder", or dropped in by drag and drop. "New folder" creates folders. Files can be moved between folders by drag and drop.

Clicking a file opens the preview. It shows the Original, the Converted Markdown generated from it and, if present, the manually edited version.

The menu of a file offers the following actions:

• Refresh: process the file again for search and AI, for example after it has been replaced
• Replace: upload a new version of the file
• Edit: change the content directly in enneo
• Reset to original: discard manual changes and restore the originally uploaded content
• View original
• Rename and Move to folder
• Archive or Unarchive
• Delete: remove the file permanently

Each file shows its processing status: Uploading, Indexing, Indexed, Error or Archived. Only once a file is Indexed is its content available for search and AI functions.

If a file was edited directly in enneo, it is marked "Manually edited". The content in use then no longer exactly matches the originally uploaded document.

Checklist

For website sources:

• Is the external source reliable?
• Are the include and exclude URL patterns set sensibly?
• Is the maximum number of pages appropriate?
• Is the crawl frequency appropriate?
• Does the source need to be crawled again after a change?

For files:

• Is the file name unambiguous?
• Is the file stored in the right folder?
• Is the content up to date and reviewed?
• Was the file reprocessed after changes?
• Is it clear whether the file is used in its original state or in a manually edited version?