Skip to main content
POST
Ingest a document

Body

application/json
namespace
string
required

Target knowledge base namespace.

source_id
string
required

Unique identifier for the source document (e.g. filename).

content
string
required

Text content to ingest (markdown, plain text, etc.).

tags
string[]

Optional tags for filtering during retrieval.

source_type
string

Content type hint (e.g. "markdown", "text").

metadata
object

Arbitrary metadata attached to the source.

Response

Ingest completed or accepted for processing.

ok
boolean
required
job_id
string
required

Job identifier for tracking (poll via /knowledge_base/get_status).

status
enum<string>
required

completed for sub-threshold sync ingests; processing when content length exceeds the async threshold (see gateway knowledge_base route).

Available options:
completed,
processing
async
boolean

True when ingest continues in the background (large content).

chunks_created
integer

Present when status is completed (sync path).

source_id
string

Echo of the ingested source identifier when completed.