AI settings API

The AI settings API allows you to configure and manage LLM configurations for the AI chat .

List AI settings

URI: https://hub.cfengine.com/api/ai-settings

Method: GET

List all configured LLM settings.

Example request (curl):

curl --user <username>:<password> \
  -X GET \
  https://hub.cfengine.com/api/ai-settings

Example response:

HTTP 200 OK
[
  {
    "id": 1,
    "provider": "openai",
    "model": "gpt-4",
    "token": "token is set",
    "base_url": null,
    "name": "OpenAI GPT-4",
    "description": "Production OpenAI configuration",
    "meta": {},
    "created_at": "2025-01-15 10:30:00",
    "updated_at": "2025-01-15 10:30:00"
  }
]

Output:

• id (number) Unique configuration ID.
• provider (string) LLM provider type. Required. Allowed values: openai, gemini, mistral, ollama, anthropic, openai_like.
• model (string) Model name to use. Required. Maximum 255 characters.
• token (string) API token for authentication. Required for all providers except ollama. Maximum 255 characters. Note: in the output it returns “token is set” to mask the actual token value for security purposes.
• base_url (string) Custom base URL for API endpoint. Required for ollama and openai_like providers. Must be a valid URL, maximum 500 characters.
• name (string) Name for this configuration. Optional. Maximum 255 characters.
• description (string) Description of the configuration. Optional. Maximum 4012 characters.
• meta (object) Additional metadata as key-value pairs. Optional. Values must be scalar types (string, integer, float, boolean, or null).
• created_at (string) Timestamp when the configuration was created.
• updated_at (string) Timestamp when the configuration was last updated.

Get AI setting

URI: https://hub.cfengine.com/api/ai-settings/:id

Method: GET

Get a specific LLM configuration by ID.

Parameters:

• id (integer) Configuration ID. Required.

Example request (curl):

curl --user <username>:<password> \
  -X GET \
  https://hub.cfengine.com/api/ai-settings/1

Example response:

HTTP 200 OK
{
  "id": 1,
  "provider": "openai",
  "model": "gpt-4",
  "token": "token is set"
  "base_url": null,
  "name": "OpenAI GPT-4",
  "description": "Production OpenAI configuration",
  "meta": {},
  "created_at": "2025-01-15 10:30:00",
  "updated_at": "2025-01-15 10:30:00"
}

Responses:

Create AI setting

URI: https://hub.cfengine.com/api/ai-settings

Method: POST

Create a new LLM configuration.

Parameters:

• provider (string) LLM provider type. Required. Allowed values: openai, gemini, mistral, ollama, anthropic, openai_like.
• model (string) Model name to use. Required. Maximum 255 characters.
• token (string) API token for authentication. Required for all providers except ollama. Maximum 255 characters.
• base_url (string) Custom base URL for API endpoint. Required for ollama and openai_like providers. Must be a valid URL, maximum 500 characters.
• name (string) Name for this configuration. Optional. Maximum 255 characters.
• description (string) Description of the configuration. Optional. Maximum 4012 characters.
• meta (object) Additional metadata as key-value pairs. Optional. Values must be scalar types (string, integer, float, boolean, or null).

Example request (curl):

curl --user <username>:<password> \
  -X POST \
  https://hub.cfengine.com/api/ai-settings \
  -H 'Content-Type: application/json' \
  -d '{
    "provider": "openai",
    "model": "gpt-4",
    "token": "sk-proj-xxxxxxxxxxxxx",
    "name": "OpenAI GPT-4",
    "description": "Production OpenAI configuration"
  }'

Example with Ollama (no token required):

curl --user <username>:<password> \
  -X POST \
  https://hub.cfengine.com/api/ai-settings \
  -H 'Content-Type: application/json' \
  -d '{
    "provider": "ollama",
    "model": "llama2",
    "base_url": "http://localhost:11434",
    "name": "Local Ollama",
    "description": "Local development LLM"
  }'

Example response:

HTTP 201 Created
{
  "id": 3
}

Responses:

Validation errors response example:

HTTP 400 Bad Request
{
  "success": false,
  "errors": [
    {
      "field": "[token]",
      "message": "API Token is required for all providers except ollama"
    }
  ]
}

Update AI setting

URI: https://hub.cfengine.com/api/ai-settings/:id

Method: PATCH

Update an existing LLM configuration.

Parameters:

• id (integer) Configuration ID. Required (in URL path).
• provider (string) LLM provider type. Required. Allowed values: openai, gemini, mistral, ollama, anthropic, openai_like.
• model (string) Model name to use. Required. Must be 1-255 characters.
• token (string) API token for authentication. Required for all providers except ollama. Maximum 255 characters.
• base_url (string) Custom base URL for API endpoint. Optional. Must be a valid URL, maximum 500 characters.
• name (string) Human-readable name for this configuration. Optional. Must be 1-255 characters.
• description (string) Description of the configuration. Optional. Maximum 4012 characters.
• meta (object) Additional metadata as key-value pairs. Optional. Values must be scalar types (string, integer, float, boolean, or null).

Example request (curl):

curl --user <username>:<password> \
  -X PATCH \
  https://hub.cfengine.com/api/ai-settings/1 \
  -H 'Content-Type: application/json' \
  -d '{
    "provider": "openai",
    "model": "gpt-4-turbo",
    "token": "sk-proj-xxxxxxxxxxxxx",
    "name": "OpenAI GPT-4 Turbo",
    "description": "Updated to use GPT-4 Turbo model"
  }'

Example response:

HTTP 202 Accepted

Responses:

Delete AI setting

URI: https://hub.cfengine.com/api/ai-settings/:id

Method: DELETE

Delete an LLM configuration.

Parameters:

• id (integer) Configuration ID. Required.

Example request (curl):

curl --user <username>:<password> \
  -X DELETE \
  https://hub.cfengine.com/api/ai-settings/1

Example response:

HTTP 202 Accepted

Responses: