Skip to content

Quick Start

agentomatic logo

API Platform in 60 Seconds

1. Install dependencies

Install Agentomatic with all features:

pip install agentomatic[all]

2. Scaffold a chatbot agent

Generate a functional basic chatbot agent using the CLI:

agentomatic init my_chatbot --template basic

This scaffolds a directory structure under agents/:

agents/my_chatbot/
├── __init__.py      # Manifest declaration + route hook entry
├── graph.py         # LangGraph state graph definition
├── nodes.py         # Node logic functions
├── prompts.json     # Dynamic system and user prompt templates
├── langgraph.json   # LangGraph Studio local developer settings
├── .env.example     # Environment variables blueprint
└── README.md        # Agent documentation markdown

3. Run the API server

# Starts the web server and embeds the graphical Chat UI at /chat
agentomatic run --with-ui --reload
- FastAPI application running at: http://localhost:8000 - OpenAPI Swagger documentation at: http://localhost:8000/docs - Chainlit chat playground at: http://localhost:8000/chat

Create a main.py file to customize the platform stack: ```python

main.py

from agentomatic import AgentPlatform

platform = AgentPlatform.from_folder("agents/") app = platform.build()

``` Run with uvicorn: 

uvicorn main:app --reload

4. Query the endpoints

You can interact with your newly deployed agent using Curl, Python, or the CLI.

curl -X POST http://localhost:8000/api/v1/my_chatbot/invoke \
  -H "Content-Type: application/json" \
  -d '{"query": "Hello!"}'

import httpx

response = httpx.post(
    "http://localhost:8000/api/v1/my_chatbot/invoke",
    json={"query": "Hello!"},
)
print(response.json())

# Opens a chat-like session inside your terminal
agentomatic test my_chatbot

Expected JSON Response

{
  "response": "Hello! How can I assist you today?",
  "agent_type": "agent-my_chatbot",
  "thread_id": "thread_abc123",
  "suggestions": ["Introduce yourself", "What can you do?"],
  "citations": [],
  "steps_taken": ["greeting_node"],
  "metadata": {},
  "duration_ms": 114.2
}

🧭 Explore Further

Now that you have your first API microservice running, here is where to look next: