Library

The MOCA Library — RAG Q&A and search over the museum's knowledge base (powered by Cortex)

Ask the Library

POST

https://api.moca.qwellco.de

/v1/library/ask

RAG Q&A over the MOCA knowledge base: hybrid retrieval, optional knowledge-graph context, cited sources. For token-by-token output use the streaming variant.

Ask the Library › Request Body

question

string · required

The question to ask the Library

collection_id

string

Scope to one Library collection (see GET /v1/library/collections)

top_k

integer · min: 1 · max: 20

Default: 5

use_graph

boolean

Include knowledge-graph context

Default: true

use_agentic

boolean

Deep-research mode: multi-step agent reasoning (slower, more thorough)

Default: false

object[]

Prior turns for follow-up questions (last 20 kept)

Ask the Library › Responses

Answer with citations

LibraryAnswer

POST/v1/library/ask

curl https://api.moca.qwellco.de/v1/library/ask \
  --request POST \
  --header 'Content-Type: application/json' \
  --header 'X-API-Key: <api-key>' \
  --data '{
  "question": "What is the significance of the Genesis Collection?",
  "top_k": 5
}'

Example Request Body

{
  "question": "What is the significance of the Genesis Collection?",
  "top_k": 5
}

json

application/json

Example Responses

{
  "data": {
    "question": "question",
    "answer": "answer",
    "sources": [
      {
        "document_id": "document_id",
        "chunk_id": "chunk_id",
        "content": "content",
        "score": 0,
        "metadata": {
          "filename": "filename"
        }
      }
    ],
    "graph_context": {}
  }
}

json

application/json

Ask the Library (streaming)

POST

https://api.moca.qwellco.de

/v1/library/ask/stream

Same request as /v1/library/ask, answered as Server-Sent Events. Each data: line is a JSON object: {"content": "…"} token chunks, one {"sources": [...]} event, optional {"thinking": "…"} / {"sub_questions": [...]} (agentic mode), {"graph_context": {...}}, and a final {"done": true}. Deep-research streams can run for minutes — disable proxy buffering.

Ask the Library (streaming) › Request Body

question

string · required

The question to ask the Library

collection_id

string

Scope to one Library collection (see GET /v1/library/collections)

top_k

integer · min: 1 · max: 20

Default: 5

use_graph

boolean

Include knowledge-graph context

Default: true

use_agentic

boolean

Deep-research mode: multi-step agent reasoning (slower, more thorough)

Default: false

object[]

Prior turns for follow-up questions (last 20 kept)

Ask the Library (streaming) › Responses

SSE stream (text/event-stream)

string

POST/v1/library/ask/stream

curl https://api.moca.qwellco.de/v1/library/ask/stream \
  --request POST \
  --header 'Content-Type: application/json' \
  --header 'X-API-Key: <api-key>' \
  --data '{
  "question": "question",
  "collection_id": "collection_id",
  "top_k": 5,
  "use_graph": true,
  "use_agentic": false,
  "conversation_history": [
    {
      "role": "user",
      "content": "content"
    }
  ]
}'

Example Request Body

{
  "question": "question",
  "collection_id": "collection_id",
  "top_k": 5,
  "use_graph": true,
  "use_agentic": false,
  "conversation_history": [
    {
      "role": "user",
      "content": "content"
    }
  ]
}

json

Example Responses

data: {"content": "The Genesis"}

data: {"done": true}

text

text/event-stream

Search the Library

POST

https://api.moca.qwellco.de

/v1/library/search

Hybrid retrieval (vector + keyword + graph) without answer generation — returns scored excerpts. Cheaper and faster than ask; ideal for building your own UX on top.

Search the Library › Request Body

query

string · required

top_k

integer · min: 1 · max: 25

Default: 5

collection_id

string

Search the Library › Responses

Scored excerpts

object

List Library collections

GET

https://api.moca.qwellco.de

/v1/library/collections

The knowledge-base collections available to ask/search against (use their id as collection_id).

List Library collections › Responses

Collections

object[]

Search Souls

Ask the Library

POST

https://api.moca.qwellco.de

/v1/library/ask

RAG Q&A over the MOCA knowledge base: hybrid retrieval, optional knowledge-graph context, cited sources. For token-by-token output use the streaming variant.

Ask the Library › Request Body

question

string · required

The question to ask the Library

collection_id

string

Scope to one Library collection (see GET /v1/library/collections)

top_k

integer · min: 1 · max: 20

Default: 5

use_graph

boolean

Include knowledge-graph context

Default: true

use_agentic

boolean

Deep-research mode: multi-step agent reasoning (slower, more thorough)

Default: false

object[]

Prior turns for follow-up questions (last 20 kept)

Ask the Library › Responses

Answer with citations

LibraryAnswer

Ask the Library (streaming)

POST

https://api.moca.qwellco.de

/v1/library/ask/stream

Ask the Library (streaming) › Request Body

question

string · required

The question to ask the Library

collection_id

string

Scope to one Library collection (see GET /v1/library/collections)

top_k

integer · min: 1 · max: 20

Default: 5

use_graph

boolean

Include knowledge-graph context

Default: true

use_agentic

boolean

Deep-research mode: multi-step agent reasoning (slower, more thorough)

Default: false

object[]

Prior turns for follow-up questions (last 20 kept)

Ask the Library (streaming) › Responses

SSE stream (text/event-stream)

string

Search the Library

POST

https://api.moca.qwellco.de

/v1/library/search

Hybrid retrieval (vector + keyword + graph) without answer generation — returns scored excerpts. Cheaper and faster than ask; ideal for building your own UX on top.

Search the Library › Request Body

query

string · required

top_k

integer · min: 1 · max: 25

Default: 5

collection_id

string

Search the Library › Responses

Scored excerpts

object