API docs by Redocly

101gen.ai API (v1)

Download OpenAPI specification:Download

This is the 101gen.ai API documentation.

Token

Get bearer token

Get a new JWT bearer token for use with API calls for authorization

Request Body schema: application/json

Description of the new bearer token request body

client_id
string

Client ID of the API client

client_secret
string

Pre-shared secret

Responses

Request samples

Content type
application/json
{
  • "client_id": "string",
  • "client_secret": "string"
}

Response samples

Content type
application/json
{
  • "access_token": "string",
  • "expires_in": 3600
}

Models

List models

Lists all models that power available co-pilots

Authorizations:
bearerAuth

Responses

Response samples

Content type
application/json
{
  • "models": [
    ]
}

Chats

Prompt co-pilot

Invoke a multi-modal response from the co-pilot in response to the given query text/prompt.

Authorizations:
bearerAuth
Request Body schema:

Description of the request body to prompt the co-pilot

project_id
string <uuid> = 36 characters

Unique co-pilot ID. Example: <some_uuid4_string>. The copilot_id is a unique identifier for the specific co-pilot being prompted.

end_user_id
string <uuid> = 36 characters

Unique end_user ID. Example: <some_uuid4_string>. The end_user_id is a unique identifier for the end-user interacting with the customer's co-pilot.

interaction_id
string <uuid> = 36 characters

Unique interaction or session ID. Example: <some_uuid4_string>. The interaction_id is a unique identifier for an instance of a multi-turn dialogue between the end-user and the customer's co-pilot.

files
Array of strings <binary> [ items <binary > ]

A list of files to be uploaded and made available to a multi-modal co-pilot.

query
string

Prompt. Example: I am looking for help managing some recurring upper back pain. This is one instance/turn of conversation from the customer to the copilot. All queries in an interaction are appended in received order, along with the co-pilot's generated responses, in the interaction history.

stage_id
integer
Default: -1

The processing stage/agent of the co-pilot (which can be a network of agents) that must process this query.

max_history
integer [ 0 .. 20 ]
Default: 3

The maximum number of previous turns of conversation that must be fed in to the co-pilot's context as conversation history.

model
string

The name of the model to be used for generating the response.

temperature
number [ 0 .. 1 ]
Default: 0

Temperature setting for the co-pilot's text response generator.

max_out_tokens
integer

The approximate maximum number of tokens the co-pilot must generate in its text response. This parameter is clipped to the model's maximum output token limit.

nnq
integer [ 0 .. 3 ]
Default: 0

The number of 'suggested next questions' that the model should propose for the next turn of conversation.

fetch_k
integer [ 0 .. 10000 ]
Default: 20

The number of relevant vectors (embedded chunks of text) that the retriever should fetch from the vector database.

fetch_p
number [ 0 .. 1 ]
Default: 0

The threshold for the hybrid relevance score below which vectors from the vector database should be ignored.

hybrid_alpha
number [ 0 .. 1 ]
Default: 0.8

The relative weightage towards vector similarity vs. keyword search-based matching between the input query and vector objects.

debug
boolean
Default: false

Flag that controls diagnostic-mode output from the API.

Responses

Request samples

Content type
No sample

Response samples

Content type
application/json
{
  • "response_id": "696f09c4-2678-4263-8db2-0a7204e502c2",
  • "timestamp": "2019-08-24T14:15:22Z",
  • "project_id": "405d8375-3514-403b-8c43-83ae74cfe0e9",
  • "end_user_id": "c5ac076e-1e7a-4f0d-9ad4-cae618230def",
  • "interaction_id": "fd463a24-cab4-49e4-b915-3c70a62977c3",
  • "query": "string",
  • "stage_id": -1,
  • "expert_supervised": false,
  • "stage_locked": false,
  • "ai_response_text": "string",
  • "response_text": "string",
  • "generation_config": {
    },
  • "thumbs_up": true,
  • "thumbs_feedback": "string",
  • "star_rating": 1,
  • "star_feedback": "string",
  • "finish_reason": "string",
  • "media_list": [],
  • "next_questions": [
    ],
  • "grounding": [
    ],
  • "file_names": [],
  • "chunk_ids": [],
  • "usage": {
    },
  • "languageModel": "string",
  • "sys_fp": "string",
  • "response_tts_path": "string",
  • "ttsresponse_id": "string",
  • "sttresponse_id": "string",
  • "log_probs": [ ],
  • "scores": [
    ],
  • "llm_full_answer": "string",
  • "project_ids": [
    ]
}

Chat filtering criteria

Get filter criteria metadata for retrieval of multi-modal responses from conversation history.

Authorizations:
bearerAuth
path Parameters
project_id
required
string <uuid> = 36 characters

Unique project ID. Example: <some_uuid4_string>. The copilot_id for the co-pilot whose conversation history filter metadata is being requested.

end_user_id
required
string <uuid> = 36 characters

Unique end-user ID. Example: <some_uuid4_string>. The end_user_id for the end-user whose conversation history filter metadata with this co-pilot is being requested.

interaction_id
required
string <uuid> = 36 characters

Unique interaction or session ID. Example: <some_uuid4_string>. The interaction_id for the interaction or session whose conversation history filter metadata with this co-pilot is being requested.

Responses

Response samples

Content type
application/json
{
  • "users": [
    ],
  • "interactions": [
    ],
  • "from_timestamp": "2019-08-24T14:15:22Z",
  • "to_timestamp": "2019-08-24T14:15:22Z"
}

Get chat history

Get the multi-modal chat history given filtering and pagination criteria.

Authorizations:
bearerAuth
path Parameters
project_id
required
string <uuid> = 36 characters

Unique project ID. Example: <some_uuid4_string>. The copilot_id for the co-pilot whose conversation history is being requested.

end_user_id
required
string <uuid> = 36 characters

Unique end-user ID. Example: <some_uuid4_string>. The end_user_id for the end-user whose conversation history with this co-pilot is being requested.

interaction_id
required
string <uuid> = 36 characters

Unique interaction or session ID. Example: <some_uuid4_string>. The interaction_id for the interaction or session whose conversation history with this co-pilot is being requested.

query Parameters
page
number >= 1

The specific page of results to be returned. This is not a filtering criterion.

per_page
number >= 1

The number of entries to be returned per page. This is not a filtering criterion.

from_timestamp
string <date-time>

Filtering criterion specifying the earliest timestamp (in ISO-8601 format) from which responses need to be retrieved.

to_timestamp
string <date-time>

Filtering criterion specifying the latest timestamp (in ISO-8601 format) till which responses need to be retrieved.

Responses

Response samples

Content type
application/json
{
  • "num_results": 0,
  • "page": 0,
  • "per_page": 0,
  • "has_next": true,
  • "has_prev": true,
  • "results": [
    ]
}

Audio

Speech-to-text transcription

Transcribe a live recording or an uploaded audio file to text.

Authorizations:
bearerAuth
Request Body schema: multipart/form-data

The audio data to be transcribed.

audio_file
string <binary>
project_id
string <uuid> = 36 characters
end_user_id
string <uuid> = 36 characters
interaction_id
string <uuid> = 36 characters
input_language
string

The language of the audio file.

Responses

Response samples

Content type
application/json
{
  • "stt_response_id": "590e1b8c-cec2-401d-a213-2144483eeb2c",
  • "timestamp": "2019-08-24T14:15:22Z",
  • "project_id": "405d8375-3514-403b-8c43-83ae74cfe0e9",
  • "end_user_id": "c5ac076e-1e7a-4f0d-9ad4-cae618230def",
  • "interaction_id": "fd463a24-cab4-49e4-b915-3c70a62977c3",
  • "text_translated": "string",
  • "translated": true,
  • "path": "string",
  • "text": "string",
  • "language": "string",
  • "usage": {
    }
}

Text-to-speech synthesis

Synthesize audio from text.

Authorizations:
bearerAuth
Request Body schema: application/json

The text to be synthesized.

response_id
string <uuid> = 36 characters

The multi-modal response ID (in UUID4 format), the text portion of which is to be synthesized.

Responses

Request samples

Content type
application/json
{
  • "response_id": "696f09c4-2678-4263-8db2-0a7204e502c2"
}

Response samples

Content type
application/json
{
  • "tts_response_id": "2931dfb2-45b4-4854-aacc-d891ea0bf813",
  • "timestamp": "2019-08-24T14:15:22Z",
  • "response_id": "696f09c4-2678-4263-8db2-0a7204e502c2",
  • "project_id": "405d8375-3514-403b-8c43-83ae74cfe0e9",
  • "end_user_id": "c5ac076e-1e7a-4f0d-9ad4-cae618230def",
  • "interaction_id": "fd463a24-cab4-49e4-b915-3c70a62977c3",
  • "path": "string",
  • "usage": {
    }
}

Inferences

Inference filtering criteria

Get filter criteria metadata for retrieval of inference responses.

Authorizations:
bearerAuth
path Parameters
copilot_id
required
string <uuid> = 36 characters

Unique co-pilot ID. Example: <some_uuid4_string>. The ID of the co-pilot whose inference history filter metadata is being requested.

user_id
string <uuid> = 36 characters

Unique end-user ID. Example: <some_uuid4_string>. The ID of the end-user whose inference history filter metadata with this co-pilot is being requested.

Responses

Response samples

Content type
application/json
{
  • "users": [
    ],
  • "interactions": [
    ],
  • "from_timestamp": "2019-08-24T14:15:22Z",
  • "to_timestamp": "2019-08-24T14:15:22Z"
}

Generate an inference

Submit a payload for inference.

Authorizations:
bearerAuth
path Parameters
copilot_id
required
string <uuid> = 36 characters

Unique copilot ID. Example: <some_uuid4_string>. The copilot_id for the co-pilot on which the inference is being requested.

user_id
required
string <uuid> = 36 characters

Unique end-user ID. Example: <some_uuid4_string>. The end_user_id for the end-user making the inference request.

Request Body schema:
required

A list of predictor values (including files, such as images), the inference model, and model-specific inference payload.

model
string

Inference model identifier.

payload
object

Model-specific payload.

Responses

Request samples

Content type
{
  • "model": "ooogenai/xwsegnet",
  • "payload": "{\"key1\": \"value1\", \"key2\": \"value2\"}"
}

Response samples

Content type
application/json
{
  • "response_id": "696f09c4-2678-4263-8db2-0a7204e502c2",
  • "timestamp": "2019-08-24T14:15:22Z",
  • "project_id": "405d8375-3514-403b-8c43-83ae74cfe0e9",
  • "end_user_id": "c5ac076e-1e7a-4f0d-9ad4-cae618230def",
  • "model": "string",
  • "payload": { },
  • "file_list": [
    ],
  • "inferences": [
    ]
}

Get inference history

Get inference history given filtering and pagination criteria.

Authorizations:
bearerAuth
path Parameters
copilot_id
required
string <uuid> = 36 characters

Unique co-pilot ID. Example: <some_uuid4_string>. The ID of the co-pilot whose image inference history is being requested.

user_id
required
string <uuid> = 36 characters

Unique end-user ID. Example: <some_uuid4_string>. The ID of the end-user whose image inference history with this co-pilot is being requested.

query Parameters
page
number >= 1

The specific page of results to be returned. This is not a filtering criterion.

per_page
number >= 1

The number of entries to be returned per page. This is not a filtering criterion.

from_timestamp
string <date-time>

Filtering criterion specifying the earliest timestamp (in ISO-8601 format) from which responses need to be retrieved.

to_timestamp
string <date-time>

Filtering criterion specifying the latest timestamp (in ISO-8601 format) till which responses need to be retrieved.

Responses

Response samples

Content type
application/json
{
  • "num_results": 0,
  • "page": 0,
  • "per_page": 0,
  • "has_next": true,
  • "has_prev": true,
  • "results": [
    ]
}

Co-pilots

Create co-pilot

Create a new co-pilot

Authorizations:
bearerAuth
Request Body schema: application/json
projectId
string <uuid> = 36 characters

The co-pilot's UUID in UUID v4 format.

customerId
string <uuid> = 36 characters

The UUID, in UUID v4 format, of the customer that owns this co-pilot.

useCase
string

The co-pilot use case.

createdAt
string <date-time>

The date and time (UTC, is ISO-8601 format) when the co-pilot was created.

lastUpdatedAt
string <date-time>

The date and time (UTC, in ISO-8601 format) when the co-pilot was last updated.

createdBy
string

The user or entity that created the co-pilot.

agentName
string

Name of the co-pilot.

agentDescription
string

Description of the co-pilot.

companyUrl
string <uri>

The company's website (or any other site that they would like to be fed to the co-pilot for RAG).

Array of objects (Dataset)

List of datasets question associated with the co-pilot.

Array of objects (Section)

List of co-pilot enhancements.

defaultPrompt
string

The "system" prompt for the co-pilot. If it is not set here, a generic default prompt will be assumed, assuming the co-pilot has broad open-domain knowledge.

languageModel
string
Default: "gpt-3.5-turbo-0125"

The language model used by the co-pilot (e.g. 'Llama-2-13B').

Responses

Request samples

Content type
application/json
{
  • "projectId": "5a8591dd-4039-49df-9202-96385ba3eff8",
  • "customerId": "87d8e330-2878-4742-a86f-dbbb3bf522ac",
  • "useCase": "string",
  • "createdAt": "2019-08-24T14:15:22Z",
  • "lastUpdatedAt": "2019-08-24T14:15:22Z",
  • "createdBy": "string",
  • "agentName": "string",
  • "agentDescription": "string",
  • "companyUrl": "http://example.com",
  • "datasets": [
    ],
  • "sections": [
    ],
  • "defaultPrompt": "string",
  • "languageModel": "gpt-3.5-turbo-0125"
}

Response samples

Content type
application/json
{
  • "projectId": "5a8591dd-4039-49df-9202-96385ba3eff8",
  • "customerId": "87d8e330-2878-4742-a86f-dbbb3bf522ac",
  • "useCase": "string",
  • "createdAt": "2019-08-24T14:15:22Z",
  • "lastUpdatedAt": "2019-08-24T14:15:22Z",
  • "createdBy": "string",
  • "agentName": "string",
  • "agentDescription": "string",
  • "companyUrl": "http://example.com",
  • "datasets": [
    ],
  • "sections": [
    ],
  • "defaultPrompt": "string",
  • "languageModel": "gpt-3.5-turbo-0125",
  • "files_job_id": "528f51ef-cfc5-4c00-8241-f5b7124925dd",
  • "website_job_id": "889f1a4a-a108-4b04-92bc-cb8082fef964"
}

Get all co-pilots

Retrieve a list of all co-pilots (owned by this customer).

Authorizations:
bearerAuth

Responses

Response samples

Content type
application/json
{
  • "projects": [
    ]
}

Get co-pilot

Retrieve a specific co-pilot by its ID

Authorizations:
bearerAuth
path Parameters
project_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the co-pilot

Responses

Response samples

Content type
application/json
{
  • "projectId": "5a8591dd-4039-49df-9202-96385ba3eff8",
  • "customerId": "87d8e330-2878-4742-a86f-dbbb3bf522ac",
  • "useCase": "string",
  • "createdAt": "2019-08-24T14:15:22Z",
  • "lastUpdatedAt": "2019-08-24T14:15:22Z",
  • "createdBy": "string",
  • "agentName": "string",
  • "agentDescription": "string",
  • "companyUrl": "http://example.com",
  • "datasets": [
    ],
  • "sections": [
    ],
  • "defaultPrompt": "string",
  • "languageModel": "gpt-3.5-turbo-0125"
}

Update co-pilot

Update a co-pilot given its ID

Authorizations:
bearerAuth
path Parameters
project_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the co-pilot

query Parameters
llm
string

The language model to use to generate the co-pilot's default prompt. Unsupported values are ignored (the system default is used instead).

mode
string
Enum: "sections-only" "prompt-only" "full"

Indicates whether to regenerate only the section texts ('sections-only'), or the default prompt ('prompt-only'), or both ('full', default).

Request Body schema: application/json
projectId
string <uuid> = 36 characters

The co-pilot's UUID in UUID v4 format.

customerId
string <uuid> = 36 characters

The UUID, in UUID v4 format, of the customer that owns this co-pilot.

useCase
string

The co-pilot use case.

createdAt
string <date-time>

The date and time (UTC, is ISO-8601 format) when the co-pilot was created.

lastUpdatedAt
string <date-time>

The date and time (UTC, in ISO-8601 format) when the co-pilot was last updated.

createdBy
string

The user or entity that created the co-pilot.

agentName
string

Name of the co-pilot.

agentDescription
string

Description of the co-pilot.

companyUrl
string <uri>

The company's website (or any other site that they would like to be fed to the co-pilot for RAG).

Array of objects (Dataset)

List of datasets question associated with the co-pilot.

Array of objects (Section)

List of co-pilot enhancements.

defaultPrompt
string

The "system" prompt for the co-pilot. If it is not set here, a generic default prompt will be assumed, assuming the co-pilot has broad open-domain knowledge.

languageModel
string
Default: "gpt-3.5-turbo-0125"

The language model used by the co-pilot (e.g. 'Llama-2-13B').

Responses

Request samples

Content type
application/json
{
  • "projectId": "5a8591dd-4039-49df-9202-96385ba3eff8",
  • "customerId": "87d8e330-2878-4742-a86f-dbbb3bf522ac",
  • "useCase": "string",
  • "createdAt": "2019-08-24T14:15:22Z",
  • "lastUpdatedAt": "2019-08-24T14:15:22Z",
  • "createdBy": "string",
  • "agentName": "string",
  • "agentDescription": "string",
  • "companyUrl": "http://example.com",
  • "datasets": [
    ],
  • "sections": [
    ],
  • "defaultPrompt": "string",
  • "languageModel": "gpt-3.5-turbo-0125"
}

Response samples

Content type
application/json
{
  • "projectId": "5a8591dd-4039-49df-9202-96385ba3eff8",
  • "customerId": "87d8e330-2878-4742-a86f-dbbb3bf522ac",
  • "useCase": "string",
  • "createdAt": "2019-08-24T14:15:22Z",
  • "lastUpdatedAt": "2019-08-24T14:15:22Z",
  • "createdBy": "string",
  • "agentName": "string",
  • "agentDescription": "string",
  • "companyUrl": "http://example.com",
  • "datasets": [
    ],
  • "sections": [
    ],
  • "defaultPrompt": "string",
  • "languageModel": "gpt-3.5-turbo-0125"
}

Delete co-pilot

Delete a co-pilot by its ID

Authorizations:
bearerAuth
path Parameters
project_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the co-pilot

Responses

Response samples

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

(Re-)Generate co-pilot prompt

This method synthesizes the co-pilot's default prompt from the rest of the the co-pilot's settings

Authorizations:
bearerAuth
path Parameters
project_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the co-pilot whose default prompt needs to be (re-)generated

query Parameters
llm
string

The language model to use to generate the co-pilot's default prompt. Unsupported values are ignored (the system default is used instead).

mode
string
Enum: "sections-only" "prompt-only" "full"

Indicates whether to regenerate only the section texts ('sections-only'), or the default prompt ('prompt-only'), or both ('full', default).

Responses

Response samples

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

Datasets

Upload new datasets

Upload a list of datasets to a copilot

Authorizations:
bearerAuth
path Parameters
project_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the co-pilot

Request Body schema: multipart/form-data
files
Array of strings <binary> [ items <binary > ]
website
string <uri>

Responses

Response samples

Content type
application/json
{
  • "message": "string",
  • "files_job_id": "528f51ef-cfc5-4c00-8241-f5b7124925dd",
  • "website_job_id": "889f1a4a-a108-4b04-92bc-cb8082fef964"
}

Get attached datasets

Retrieve the datasets attached to a co-pilot

Authorizations:
bearerAuth
path Parameters
project_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the co-pilot

dataset_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the dataset

Responses

Response samples

Content type
application/json
{
  • "urls": [],
  • "docs": [
    ],
  • "images": [
    ],
  • "tables": [
    ]
}

Update dataset

Update a dataset attached to a co-pilot

Authorizations:
bearerAuth
path Parameters
project_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the co-pilot

dataset_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the dataset

Request Body schema: multipart/form-data
files
Array of strings <binary> [ items <binary > ]
website
string <uri>

Responses

Response samples

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

Delete dataset(s)

Delete dataset(s) from a co-pilot

Authorizations:
bearerAuth
path Parameters
project_id
required
string <uuid> = 36 characters

ID of the co-pilot

dataset_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the dataset

Responses

Response samples

Content type
application/json
{
  • "message": "string",
  • "files": [
    ]
}

Dataset ingestion log

Retrieve dataset ingestion logs

Authorizations:
bearerAuth
path Parameters
job_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the ingestion job

Responses

Response samples

Content type
application/json
{
  • "ingest_log": [
    ]
}

Dataset ingestion job

Retrieve aggregate dataset ingestion job metrics or individual job status

Authorizations:
bearerAuth
path Parameters
copilot_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the co-pilot

job_id
string <uuid> = 36 characters

ID (in UUID v4 format) of the ingestion job

Responses

Response samples

Content type
application/json
{
  • "status_counts": {
    },
  • "recent_failures": [
    ],
  • "avg_processing_time_ms": 0
}

Co-pilot Enhancements

Create co-pilot enhancement

Publishes a new enhancement to this co-pilot

Authorizations:
bearerAuth
path Parameters
project_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) for the co-pilot to which this enhancement is to be added.

query Parameters
llm
string

The language model to use to generate the co-pilot's default prompt. Unsupported values are ignored (the system default is used instead).

mode
string
Enum: "sections-only" "prompt-only" "full"

Indicates whether to regenerate only the section texts ('sections-only'), or the default prompt ('prompt-only'), or both ('full', default).

Request Body schema: application/json

Description of the request body to publish a co-pilot enhancement.

section_id
string <uuid> = 36 characters
cname
string

The canonical name for this enhancement. This should not change even if the creator decides to change the name.

name
string

A name or title for this enhancement that allows a co-pilot creator to understand, apply and configure it correctly.

description
string

A description of this enhancement, to elaborate beyond what the name can mean.

prompt
string

A prompt that can be used to generate content for this enhancement that makes it as a part of the final (configured) co-pilot system prompt template.

template_var_keys
Array of strings

A list of content placeholders for the enhancement-specific prompt template that needs to make it to the final (configured) co-pilot system prompt template.

template_var_vals
Array of strings

A list of values to substitute in the corresponding content placeholders in the enhancement-specific prompt template.

section_text
Array of strings

The contribution of this enhancement to the co-pilot's system prompt as finally decided by the co-pilot creator.

last_section_ts
string <date-time>

The timestamp of the last update to 'section_text'.

recommended_text
Array of strings

The LLM's suggested contribution (through this enhancement) to the co-pilot's system prompt.

last_recommended_ts
string <date-time>

The timestamp of the last update to 'recommended_text'.

display_order
integer

The position at which to show this enhancement in the co-pilot's list of enhancements, in the co-pilot configuration unauthorized.

llm_used
string
Default: "gpt-3.5-turbo-0125"

The language model used to generate the 'section_text'.

Responses

Request samples

Content type
application/json
{
  • "section_id": "f5e5c4dc-7d2e-40f1-a895-9bb2f8703fd8",
  • "cname": "string",
  • "name": "string",
  • "description": "string",
  • "prompt": "string",
  • "template_var_keys": [
    ],
  • "template_var_vals": [
    ],
  • "section_text": [
    ],
  • "last_section_ts": "2019-08-24T14:15:22Z",
  • "recommended_text": [
    ],
  • "last_recommended_ts": "2019-08-24T14:15:22Z",
  • "display_order": 0,
  • "llm_used": "gpt-3.5-turbo-0125"
}

Response samples

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

Get co-pilot enhancement(s)

Lists all co-pilot enhancements

Authorizations:
bearerAuth
path Parameters
project_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) for the co-pilot whose enhancements are to be retrieved.

cname
required
string

Canonical name for this enhancement. Ignore to get all enhancements for this co-pilot.

Responses

Response samples

Content type
application/json
{
  • "sections": [
    ]
}

Update co-pilot enhancement configuration

Update an existing co-pilot enhancement configuration that can apply to all co-pilots.

Authorizations:
bearerAuth
path Parameters
project_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) for the co-pilot whose enhancement is to be retrieved.

cname
required
string

Canonical name for this section.

query Parameters
llm
string

The language model to use to generate the co-pilot's default prompt. Unsupported values are ignored (the system default is used instead).

mode
string
Enum: "sections-only" "prompt-only" "full"

Indicates whether to regenerate only the section texts ('sections-only'), or the default prompt ('prompt-only'), or both ('full', default).

Request Body schema: application/json

Description of the request body to publish a co-pilot enhancement configuration.

section_id
string <uuid> = 36 characters
cname
string

The canonical name for this enhancement. This should not change even if the creator decides to change the name.

name
string

A name or title for this enhancement that allows a co-pilot creator to understand, apply and configure it correctly.

description
string

A description of this enhancement, to elaborate beyond what the name can mean.

prompt
string

A prompt that can be used to generate content for this enhancement that makes it as a part of the final (configured) co-pilot system prompt template.

template_var_keys
Array of strings

A list of content placeholders for the enhancement-specific prompt template that needs to make it to the final (configured) co-pilot system prompt template.

template_var_vals
Array of strings

A list of values to substitute in the corresponding content placeholders in the enhancement-specific prompt template.

section_text
Array of strings

The contribution of this enhancement to the co-pilot's system prompt as finally decided by the co-pilot creator.

last_section_ts
string <date-time>

The timestamp of the last update to 'section_text'.

recommended_text
Array of strings

The LLM's suggested contribution (through this enhancement) to the co-pilot's system prompt.

last_recommended_ts
string <date-time>

The timestamp of the last update to 'recommended_text'.

display_order
integer

The position at which to show this enhancement in the co-pilot's list of enhancements, in the co-pilot configuration unauthorized.

llm_used
string
Default: "gpt-3.5-turbo-0125"

The language model used to generate the 'section_text'.

Responses

Request samples

Content type
application/json
{
  • "section_id": "f5e5c4dc-7d2e-40f1-a895-9bb2f8703fd8",
  • "cname": "string",
  • "name": "string",
  • "description": "string",
  • "prompt": "string",
  • "template_var_keys": [
    ],
  • "template_var_vals": [
    ],
  • "section_text": [
    ],
  • "last_section_ts": "2019-08-24T14:15:22Z",
  • "recommended_text": [
    ],
  • "last_recommended_ts": "2019-08-24T14:15:22Z",
  • "display_order": 0,
  • "llm_used": "gpt-3.5-turbo-0125"
}

Response samples

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

Delete co-pilot enhancement configuration(s)

Deletes one or all co-pilot section configurations

Authorizations:
bearerAuth
path Parameters
project_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) for the co-pilot whose enhancements are to be deleted.

cname
required
string <uuid> = 36 characters

Canonical name for this section. Ignore to delete all of this co-pilot's enhancements.

query Parameters
llm
string

The language model to use to generate the co-pilot's default prompt. Unsupported values are ignored (the system default is used instead).

mode
string
Enum: "sections-only" "prompt-only" "full"

Indicates whether to regenerate only the section texts ('sections-only'), or the default prompt ('prompt-only'), or both ('full', default).

Responses

Response samples

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

Evaluations

Run evaluation

Run an evaluation/benchmark on the specified co-pilot

Authorizations:
bearerAuth
path Parameters
copilot_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the co-pilot for which to fetch the evaluation result

eval_name
required
string

A name for the evaluation

Request Body schema: multipart/form-data
required

Description of the request body to run an evaluation.

file
string <byte>

The input file to evaluate.

model_list
string

A comma-separated list of model names to evaluate.

Responses

Response samples

Content type
application/json
{
  • "eval_id": "b7fe92ca-3c59-4b75-8341-1d28fad399a5",
  • "copilot_id": "bf7c880d-78d8-4455-8571-a18b07e3c79a",
  • "user_id": "a169451c-8525-4352-b8ca-070dd449a1a5",
  • "user_name": "string",
  • "eval_name": "string",
  • "eval_csv_address": "string",
  • "timestamp": "string",
  • "input_filename": "string",
  • "models_used": [
    ],
  • "num_queries": 0,
  • "status": "string"
}

Get evaluations

Gets all evaluations for the specified co-pilot

Authorizations:
bearerAuth
path Parameters
copilot_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the co-pilot for which to fetch evaluation results

query Parameters
all_tests
string

If 'true'/'True'/'t'/'T'/'yes'/'Yes'/'y'/'Y'/'1', returns evaluations conducted by all users. Else, only returns evaluations conducted by the logged-in user.

Responses

Response samples

Content type
application/json
{
  • "evals": [
    ]
}

Get an evaluation

Gets the specified evaluation for the specified co-pilot

Authorizations:
bearerAuth
path Parameters
copilot_id
required
string <uuid> = 36 characters

ID (in UUID v4 format) of the co-pilot for which to fetch the evaluation result

eval_id
string <uuid> = 36 characters

ID (in UUID v4 format) of the evaluation result to fetch

Responses

Response samples

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