Create Prompts in Bulk

Endpoint: POST /api/prompts/

Creates or updates prompts in bulk. New prompts are created with full access granted to the caller. Existing ones are updated if the caller has UPDATE permission.

Note

This endpoint processes prompts sequentially. A validation failure on one prompt (e.g. a missing integration) throws immediately and halts processing of remaining prompts in the list. If the caller lacks UPDATE permission on an existing prompt, that prompt is silently skipped.

Request body

Format the request body as a JSON array of objects. Each object supports the following fields:

Parameter	Description	Type	Required/ Optional
name	The name of the prompt.	string	Required.
template	The prompt text body. Can contain ${variable} placeholders.	string	Required.
version	The version of the prompt.	integer	Optional.
description	A description of the prompt.	string	Optional.
integrations	The LLM integrations to associate with this prompt, in the format integrationName:modelName. The integration must already exist in your Conductor cluster.	Array of strings	Optional.
temperature	The temperature for the prompt.	integer	Optional.
topP	Top-p nucleus sampling parameter.	integer	Optional.
responseFormat	The output format of the prompt.	string	Optional.
stopWords	The stop sequences for generation.	Array of strings	Optional.

Response

• Returns 200 OK with no response body on success.
• Returns 401 if authentication is required.
• Returns 403 if the caller does not have READ access to a referenced integration.
• Returns 404 if a referenced integration does not exist.

Examples

Create prompts in bulk

Request

curl -X 'POST' \
  'https://<YOUR-SERVER-URL>/api/prompts/' \
  -H 'accept: */*' \
  -H 'X-Authorization: <TOKEN>' \
  -H 'Content-Type: application/json' \
  -d '[
  {
    "name": "email-summary",
    "template": "Summarize this email in ${language}: ${email_body}",
    "description": "Summarizes an email",
    "integrations": ["openAI:gpt-4o"],
    "version": 1,
    "temperature": 0.1,
    "topP": 1.0,
    "responseFormat": "text",
    "stopWords": []
  },
  {
    "name": "meeting-notes",
    "template": "Summarize the following meeting transcript: ${transcript}",
    "description": "Summarizes a meeting transcript",
    "integrations": ["openAI:gpt-4o"],
    "version": 1,
    "temperature": 0.1,
    "topP": 1.0,
    "responseFormat": "text",
    "stopWords": []
  }
]'

Response

Returns 200 OK, indicating that the prompts have been created successfully.