Cue APIs (1.0)

Download OpenAPI specification:

Handles all ingress and egress of data to/from the QAECY Cue pipelines + calls to the agent.

BLOB Writer

Handles ingress of raw documents

Upload a file for processing by the pipelines

Authorizations:
QAECYAuth
header Parameters
x-project-id
required
string

Project ID

Request Body schema: multipart/form-data
required
file
required
string <binary>

File to upload

filename
required
string

Full file path as represented in host system. If nothing is provided the name of the file is used.

customMetadata
string
Default: ""

Any relevant custom metadata from the host system formatted as stringified JSON

skipLoading
boolean
Default: false

Whether the file should be invisible to the loading events that would normally be triggered by it

skipProcessing
boolean
Default: false

Whether the file should be invisible to the processing events that would normally be triggered by it

providerId
string
Default: ""

Necessary to distinguish duplicate files uploaded from different document systems. Reflected in file location UUIDs. Default value is an empty string.

Responses

Response samples

Content type
application/json
{
  • "message": "report.pdf was successfully uploaded to project ffd5c559-1821-4e8c-a956-08f61f352e3b!",
  • "uuid": "6ad3e777-87e8-56e1-8f37-d7fcf07e3f50"
}

Assistant

Handles interactions with the Cue assistant

Trigger the service to wake up.

Authorizations:
QAECYAuth

Responses

Preload the knowledge base for the requested project.

Authorizations:
QAECYAuth
header Parameters
x-project-id
required
string

Project ID

Request Body schema: application/json
required
projectId
required
string

The project ID (will soon be removed and read from headers)

Responses

Request samples

Content type
application/json
{
  • "projectId": "string"
}

Performs QA against the knowledge base.

Authorizations:
QAECYAuth
header Parameters
x-project-id
required
string

Project ID

Request Body schema: multipart/form-data
required
prompt
required
string

Chat prompt

projectId
required
string

The project id (will soon be removed and read from headers)

target
string
Default: "page_summaries"

The target vector collection, out of the currently supported ones (page_summaries, document_summaries)

limit
number
Default: 20

Number of the vector results included in the retrieval

queryType
string
Default: "vector"

Perform vector-only (vector) or hybrid (hybrid) search which includes fts

identifiers
Array of strings

The identifiers of the documents that the search should be constrained to

Responses

Performs a SPARQL query to the graph database.

Authorizations:
QAECYAuth
header Parameters
x-project-id
required
string

Project ID

Request Body schema: application/json
required
query
required
string

The query

projectId
required
string

The project id (will soon be removed and read from headers)

identifiers
Array of strings

The identifiers of the documents that the query should be constrained to

Responses

Request samples

Content type
application/json
{
  • "query": "string",
  • "projectId": "string",
  • "identifiers": [
    ]
}

Performs search against the vector database.

Authorizations:
QAECYAuth
header Parameters
x-project-id
required
string

Project ID

Request Body schema: application/json
required
question
required
string

Search query

projectId
required
string

The project id (will soon be removed and read from headers)

target
string
Default: "page_summaries"

The target collection, out of the currently supported ones (page_summaries, document_summaries)

limit
number
Default: 20

Number of results included

queryType
string
Default: "vector"

Perform vector-only (vector) or hybrid (hybrid) search which includes fts

identifiers
Array of strings

The identifiers of the documents that the search should be constrained to

Responses

Request samples

Content type
application/json
{
  • "question": "string",
  • "projectId": "string",
  • "target": "page_summaries",
  • "limit": 20,
  • "queryType": "vector",
  • "identifiers": [
    ]
}

Knowledge

Handles writing and deleting of knowledge to the databases

Trigger the service to wake up.

Authorizations:
QAECYAuth

Responses

Remove documents and/or extracted information.

Authorizations:
QAECYAuth
header Parameters
x-project-id
required
string

Project ID

Request Body schema: application/json
required
projectId
required
string

The project id (will soon be removed and read from headers)

fileNames
required
Array of strings
Default: []

Array containing the list of filenames to be deleted, prefixed with their full path

blobNames
Array of strings
Default: []

Array containing the list of blobs to be deleted, prefixed with the projectId

removeGraphs
boolean
Default: true

Remove all the associated (also derived) information from the knowledge graph

removeVectors
boolean
Default: true

Remove all the associated vector embeddings from the vector store

removeRaw
boolean
Default: true

Remove the raw blob files

removeProcessed
boolean
Default: true

Remove all the processed / derivative files that where produced from the files

deleteAll
boolean
Default: false

Bypass other settings and delete all raw and processed files as well as knowledge graph and vector store data

Responses

Request samples

Content type
application/json
{
  • "projectId": "string",
  • "fileNames": [ ],
  • "blobNames": [ ],
  • "removeGraphs": true,
  • "removeVectors": true,
  • "removeRaw": true,
  • "removeProcessed": true,
  • "deleteAll": false
}

Triplestore

Exposes the project triplestore endpoints for querying, SHACL checks etc.

Performs a SPARQL query to the graph database.

Authorizations:
QAECYAuth
header Parameters
x-project-id
required
string

Project ID

Request Body schema: application/x-www-form-urlencoded
required
query
required
string

The query

Responses

Response samples

Content type
{}

Data views

Provides specific views on the data within a Cue project

Exposes a view of the particular resource (file location/content, fragment, entity). The default response mime type is JSON-LD but the endpoint supports content negotiation by setting the Accept header.

Experimental: This API is subject to change or removal

Authorizations:
QAECYAuth
path Parameters
resourceID
required
string

UUID of the resource

query Parameters
includeSchema
boolean

Include schema in the response

header Parameters
x-project-id
required
string

Project ID

Responses

Response samples

Content type
[]

Exposes a view of all facts that are derived from the input document. Id is either the UUID of the file location or content node. The default response mime type is JSON-LD but the endpoint supports content negotiation by setting the Accept header.

Experimental: This API is subject to change or removal

Authorizations:
QAECYAuth
path Parameters
sourceID
required
string

Source file identified by its location content UUID

query Parameters
includeSchema
boolean

Include schema in the response

header Parameters
x-project-id
required
string

Project ID

Responses

Response samples

Content type
[]