Quick Start

from collections.abc import Sequence

from agent import Agent, CallableLLM, Message


def generate(messages: Sequence[Message]) -> str:
    user_message = messages[-1].content
    return f"Runtime received: {user_message}"


agent = Agent(
    llm=CallableLLM(generate),
    instructions="Answer clearly and briefly.",
)

print(agent.run("Explain the difference between 은/는 and 이/가."))

Add a Tool

def calculator(expression: str) -> str:
    """Evaluate a small arithmetic expression."""
    return str(eval(expression, {"__builtins__": {}}, {}))


agent.add_tool(calculator)

Tools are exposed to the LLM as schemas. If the LLM returns a structured tool call, the runtime executes the registered function and sends the result back to the LLM.

Add RAG

from agent import Agent, CallableLLM
from rag import RAG

rag = RAG()
rag.add_text(
    "In Korean, 은/는 are topic particles. 이/가 usually mark the subject.",
    source="korean-grammar-notes",
)

agent = Agent(llm=CallableLLM(generate), rag=rag)

The runtime calls rag.retrieve(...) and adds retrieved chunks to the prompt.