Add Confluence Content to the Knowledge Base
Use this page after Confluence is already connected. To create or manage the Cloud / On-Premise connection, see Connect Confluence.
Once connected, you can bring Confluence pages and spaces into the project Knowledge Base so agents can search them. Choose URL Upload for a known link, or Explore Confluence spaces to pick from a tree.
Prerequisites
- An active Confluence connection for this project (Connect Confluence)
- Permission to manage Knowledge Base resources
- The Confluence URL or space you want to ingest (and rights to read it in Confluence)
Add pages from a URL
- Go to Knowledge Base → KB Data (or your project’s knowledge browser).
- Click + New → URL Upload.
- Paste the Confluence space or page URL.
- Under Confluence - Additional Options, set Extract subpages from URL:
- On — also ingest nested pages under that URL
- Off — ingest only the page at the URL
- Optionally open Explore Confluence spaces from that dialog to browse instead of pasting a URL.
- Click Add URL.
Pages commonly appear with their page ID as the document name. External links back to Confluence remain on the resource when available.
Already synced / force re-sync
If that URL (or selection) is already in the Knowledge Base, EKB prompts that the resource exists. You can cancel or confirm a force re-sync to refresh content.
Explore Confluence spaces
Use Explore Confluence when you want to pick spaces, folders, or pages from a tree instead of pasting URLs.
- From URL Upload additional options, click Explore Confluence spaces, or open Knowledge Base crawlers / Confluence explore entry points that launch the same modal.
- If Confluence is not connected, use Connect Confluence in the modal (then return here to add content).
- If scopes are insufficient, use Re-authenticate and complete Atlassian consent again.
- Select spaces, folders, or pages. Ticking a space or folder queues its entire contents.
- Click Add selected to KB.
While ingest runs, progress can show success/total counts (and rate-limit notes when Atlassian throttles). Closing the explore modal during an active ingest cancels in-flight sync jobs for the URLs being added.
Large trees stay responsive; if browse/sync is awkward, fall back to a direct URL import (URL Upload above).
After content is in the Knowledge Base
- Documents are searchable by agents that use this project’s Knowledge Base
- Apply Access Tags if only some users should see ingested pages
- Re-sync when Confluence content changes and you need a refresh (force re-sync on conflict, or add again via explore/URL)
Troubleshooting
| Symptom | What to check |
|---|---|
| Explore asks to connect first | Complete Connect Confluence, then retry Explore |
| Explore asks to re-authenticate | Click Re-authenticate — required scopes may have changed |
| No spaces listed | Account permissions in Confluence; re-authenticate; try URL Upload for a known page URL |
| Sync seems stuck on a large space | Wait for progress updates; cancel by closing Explore during ingest; retry with a narrower selection or single URL |
| “Already synced” prompt | Expected on re-add — cancel or force re-sync |
| Agents cannot find new pages | Confirm ingest finished; check Access Tags; confirm the agent’s KB toolkit targets this project |