Get knowledge base source

Get one source with its full text (text, FAQ, snippet) and, optionally, the embedded chunks — the only stored text for file and URL sources.

GET https://api.centerfy.ai/webhooks/inbound/knowledge-bases/:id/sources/:sourceId

Headers

Header	Required	Value	Description
`x-api-key`	Yes	your sub-account API key (cfy_…)	Authenticates the request; or use Authorization: Bearer `<key>`.

Path parameters

Name	Type	Required	Description
`id`	`string (uuid)`	Yes	The knowledge base id. A non-uuid returns 400; a knowledge base that does not exist in this sub-account returns 404.
`sourceId`	`string (uuid)`	Yes	The source id (a document or text-entry id, as returned when it was added or listed).

Query parameters

Name	Type	Required	Default	Description
`include_chunks`	`boolean`	No	`false`	Also return the source’s embedded chunks from the vector store, in chunk order.
`chunk_limit`	`integer`	No	`50`	Chunks per page, 1–200.
`chunk_offset`	`string`	No		Cursor from a previous response’s next_chunk_offset to fetch the next page of chunks.

Example

curl -X GET "https://api.centerfy.ai/webhooks/inbound/knowledge-bases/{id}/sources/{sourceId}?include_chunks=false&chunk_limit=50&chunk_offset=" \
  -H "x-api-key: $CENTERFY_API_KEY"

import { CenterfyClient } from "@centerfy/sdk";

const centerfy = new CenterfyClient({ apiKey: process.env.CENTERFY_API_KEY! });

const result = await centerfy.knowledgeBases.getSource("<knowledgeBaseId>", "<sourceId>", {
  "include_chunks": "false",
  "chunk_limit": 50,
  "chunk_offset": 0
});
console.log(result);

centerfy knowledgeBases get-source <knowledgeBaseId> <sourceId> '{"include_chunks":"false","chunk_limit":50,"chunk_offset":0}'

MCP tool: centerfy_knowledge_bases_get_source (see MCP server)

Response

{
  "status": "success",
  "knowledge_base_id": "66666666-7777-8888-9999-000000000000",
  "source": {
    "id": "9d3f1b2a-1111-4222-8333-000000000011",
    "kind": "text_entry",
    "type": "faq",
    "title": "Opening hours FAQ",
    "filename": null,
    "source_url": null,
    "file_type": null,
    "file_size_bytes": null,
    "status": "completed",
    "error_message": null,
    "total_chunks": null,
    "vector_count": 4,
    "content_preview": "Q1: What are your opening hours?\nA1: Monday to Friday, 8am to 6pm.\n\nQ2: Are you open on weekends?\nA2: Saturdays 9am to 1pm; closed Sundays.",
    "content_length": 118,
    "processed_at": "2026-10-01T09:01:12.000Z",
    "created_at": "2026-10-01T09:00:00.000Z",
    "updated_at": "2026-10-01T09:01:12.000Z",
    "content": "Q1: What are your opening hours?\nA1: Monday to Friday, 8am to 6pm.\n\nQ2: Are you open on weekends?\nA2: Saturdays 9am to 1pm; closed Sundays.",
    "faq_pairs": [
      {
        "question": "What are your opening hours?",
        "answer": "Monday to Friday, 8am to 6pm."
      },
      {
        "question": "Are you open on weekends?",
        "answer": "Saturdays 9am to 1pm; closed Sundays."
      }
    ]
  },
  "chunks": [
    {
      "id": "2f6a0c1e-aaaa-4bbb-8ccc-000000000001",
      "chunk_index": 0,
      "text": "Q1: What are your opening hours?\nA1: Monday to Friday, 8am to 6pm.",
      "chunk_type": "faq"
    }
  ],
  "chunk_count": 1,
  "next_chunk_offset": null,
  "chunks_error": null
}

Errors

400 — for an invalid id; 404 if the knowledge base or the source is not in this sub-account.

Notes

Read-only. Text, snippet and FAQ sources return their full content; FAQ sources additionally return the parsed faq_pairs. File and URL sources return metadata only (there is no stored text), so pass include_chunks=true to read what was embedded: chunks has id, chunk_index, text and chunk_type, chunk_count is the page size returned, and next_chunk_offset is the cursor for the next page (null when there are no more). chunks_error is set, and chunks is empty, if the vector store could not be read; the source itself is still returned.