Khub v1.2: Bulk import, version history and coverage score

Khub v1.2 is live. This release includes three features that our early users asked for most: a way to import content in bulk, full version history with rollback, and a coverage score that shows how well your knowledge base answers real queries.

Bulk import

You can now attach files directly to knowledge items and have Khub extract and index the text content automatically. Supported formats include PDF, DOCX, XLSX, TXT, MD, and CSV. Maximum file size is 10 MB per file.

The workflow is simple: create a knowledge item, attach the file in the Attachments section, and publish. Khub's background worker extracts the text, splits it into semantic chunks, generates search embeddings, and indexes everything. The content becomes searchable alongside your manually entered items.

For large imports, you can also trigger indexing immediately via the API:

POST /api/v1/workspaces/{workspaceId}/knowledge/index-files
Authorization: Bearer <token>
Content-Type: application/json

{ "itemId": "your-item-uuid" }

For onboarding an existing FAQ document or product catalogue, create one knowledge item per section/category and attach the relevant portion of the document. This gives you the best of both worlds, structured items for common queries and full-text search across the document content.

Version history

Every published knowledge item now has a full version history. Each time you publish, Khub creates a new version with an auto-incrementing version number. You can view the history for any item by clicking the clock icon, see exactly what changed in each version, and roll back to a previous version with one click.

Rolling back creates a new version (so you never lose anything), and the restored content is immediately re-indexed for search. Version history is available on Pro and Business plans.

IQ Strength indicator

The Khub panel header now shows an IQ Strength indicator that reflects how much knowledge your AI tools have access to. The indicator shows 'Growing' once you have a meaningful number of published items across multiple domain sections.

This isn't a vanity metric, it's a signal to help you understand whether your knowledge base is likely to answer the questions your customers ask. If the strength is low, the biggest improvement is usually adding more FAQs and filling gaps in your product/service listings.

Other improvements

• Hybrid search with re-ranking is now the default for all search queries (API, MCP, and Claiq IQ Guide).
• Public KB tokens (kb_...) now support configurable rate limits per token.
• The GHL integration now defaults to Web Crawler mode with a built-in sitemap endpoint.
• File attachment extraction runs automatically on a background cron cycle, no manual trigger needed for day-to-day use.

This release lays the groundwork for what's coming next: real-time collaboration, automated knowledge gap detection, and deeper connectivity with Claiq. Stay tuned.