Skip to main content
POST
/
agents
Python
from meibel import MeibelClient

client = MeibelClient(api_key="your-api-key")

# Create request body
body = CreateAgentDefinitionRequest(name="Example")

result = client.agents.create(body=body)
print(result)
{
  "id": "<string>",
  "name": "<string>",
  "display_name": "<string>",
  "version": "<string>"
}

Documentation Index

Fetch the complete documentation index at: https://docs.meibel.ai/llms.txt

Use this file to discover all available pages before exploring further.

Authorizations

Meibel-API-Key
string
header
required

Body

application/json

Request model for creating a new agent definition.

display_name
string
required

Human-readable name of the agent (letters, numbers, and spaces only). Converted to kebab-case internally.

instructions
string
required

System prompt/instructions for the agent

type
string | null
default:ReACT

Agent type

description
string | null

Description of the agent

llm_model
string | null
default:claude-sonnet-4-5-20250514

LLM model to use

fallback_models
string[] | null

List of fallback models

datasources
string[] | null

Datasource IDs the agent has access to

tools
AgentToolDefinition · object[] | null

Tools configuration

artifacts
string[] | null

Catalog URNs of artifacts the agent produces

confidence_configs
string[] | null

Confidence scoring module names to apply during execution

temperature
default:0.7

LLM temperature

Required range: 0 <= x <= 2
max_tokens
integer | null

Maximum tokens in response

tags
string[] | null

Tags for categorization

icon
string | null

UI icon identifier

additional_properties
Additional Properties · object

Response

Successful Response

id
string
required
name
string
required
display_name
string
required
version
string
required