The knowledge base feature lets ops upload documents that bots can retrieve from at runtime (RAG). It spans the knowledge base list and create page, the knowledge base detail page (documents), and the knowledge step in the bot wizard.

List and create

At /dashboard/knowledge-bases the page lists KBs as cards showing name, a status badge (active → success), description, and document_count / chunk_count. An inline form creates a KB (with name, optional description, default_language: 'en'). Cards link to /dashboard/knowledge-bases/:id.

Document detail and status

The detail page (/dashboard/knowledge-bases/:id) loads the KB and its documents in parallel, and supports:
  • Editable description — inline textarea (max 2000 chars).
  • Upload — accepts .docx, .pdf, .txt, .md, .csv (max 20MB).
  • Delete — per document.
The document table shows File Name, Status, Chunks, Characters, Version, and a delete action.

Parse status

Each document carries a parse_status rendered with a status class:
StatusMeaning
uploadedreceived, not yet processed
indexingextracting text, chunking, embedding
indexedready for retrieval
failedshows parse_error inline
While any document is indexing, the page polls the document list every 5 seconds and stops once nothing is indexing. The empty-state drop zone shows a spinner with “Indexing document…” during the first upload.

Attaching a KB to a bot

In the bot wizard, the Knowledge Base RAG section controls retrieval per bot:
ControlField
Enable knowledge retrievalknowledge_enabled
When should the bot use knowledge?knowledge_trigger_instructions (passed to the search_knowledge tool)
Attached Knowledge Basesknowledge_kb_ids (checkboxes; only active KBs listed)
Top Kknowledge_top_k (1–10)
Score Thresholdknowledge_score_threshold (0–1)
Strict Modeknowledge_strict (answer only from KB when retrieval is used)
Fallback Messageknowledge_fallback_message
Each attachable KB shows its name, document/chunk counts, and description — which is why adding a clear description on the KB detail page matters: it is what the bot sees about that KB’s contents.

Knowledge pipeline

How the CXB API chunks, embeds, and indexes documents.

Create a bot

Where the knowledge step fits in bot building.