Skip to main content

Creating Your First Agent

Learn how to build a basic agent using CAP and LangGraph from scratch.

What You'll Build​

A simple agent that can:

  • Process user input
  • Generate contextual responses
  • Maintain conversation history
  • Deploy to AWS Lambda

Prerequisites​

Step-by-Step Guide​

1. Project Setup​

mkdir my-first-agent
cd my-first-agent
python -m venv venv
source venv/bin/activate
pip install cap-engine-langgraph-lite

2. Create Agent Class​

from cap.core import BaseAgent
from cap.utils import Memory

class MyFirstAgent(BaseAgent):
    def __init__(self):
        self.memory = Memory()
        
    async def process(self, input_text: str) -> str:
        # Add to memory
        self.memory.add(input_text)
        
        # Generate response
        response = await self.generate_response(
            context=self.memory.get_context()
        )
        
        return response

3. Configure Deployment​

# main.tf
module "my_agent" {
  source = "github.com/Agaile-com/cap-cloud-ce//modules/agent"
  
  name = "my-first-agent"
  memory_size = 256
  timeout = 30
}

4. Deploy​

terraform init
terraform apply

Next Steps​