Create AI Agent

POST

/api/v3/organizations/{organisation}/ai/agents

curl --request POST \
  --url https://dashboard.quantcdn.io/api/v3/organizations/example/ai/agents \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{ "name": "Code Review Assistant", "description": "Reviews code for security vulnerabilities and best practices", "group": "development", "systemPrompt": "You are a senior software engineer specializing in secure code review.", "temperature": 0.3, "modelId": "anthropic.claude-3-5-sonnet-20241022-v2:0", "maxTokens": 4000, "allowedTools": [ "fetch_web_content", "search_web" ], "allowedCollections": [ "example" ], "assignedSkills": [ "example" ], "createdBy": "example", "longContext": true, "guardrailPreset": "official", "filterPolicies": [ "example" ] }'

• QuantCDN Public Cloud
• QuantGov Cloud

Creates a new AI agent with specific configuration, system prompt, and tool permissions. * * Agent Configuration: * - System Prompt: Instructions that guide the agent’s behavior * - Model: Which foundation model to use (e.g., ‘amazon.nova-pro-v1:0’) * - Temperature: Creativity level (0-1) * - Allowed Tools: Which tools the agent can auto-execute * - Allowed Collections: Vector DB collections for RAG * - Group: Optional categorization (e.g., ‘development’, ‘compliance’) * * Auto-Execution: * All tools are automatically executed when an agent requests them (no client confirmation needed).

Authorizations

• BearerAuth

Parameters

Path Parameters

organisation

required

string

The organisation ID

Request Body^required

Media typeapplication/json

object

name

required

string

Example

Code Review Assistant

description

required

string

Example

Reviews code for security vulnerabilities and best practices

group

string

Example

development

systemPrompt

required

string

Example

You are a senior software engineer specializing in secure code review.

temperature

number

<= 1

Example

0.3

modelId

required

string

Example

anthropic.claude-3-5-sonnet-20241022-v2:0

maxTokens

integer

>= 1 <= 200000

Example

4000

allowedTools

Array<string>

Example

[
  "fetch_web_content",
  "search_web"
]

allowedCollections

Array<string>

assignedSkills

Skill IDs to assign to this agent

Array<string>

createdBy

User identifier who created the agent

string

longContext

Enable 1M context window support

boolean

guardrailPreset

Guardrail preset name

string

nullable

Allowed values: official official-sensitive protected

filterPolicies

Filter policy IDs to apply to this agent’s inference requests

Array<string>

Responses

201

Agent created successfully

Media typeapplication/json

object

success

boolean

agent

object

message

string

Example

{
  "success": true,
  "message": "Agent created successfully"
}

400

Invalid request parameters

403

Access denied

500

Failed to create agent