Examples

All examples are interactive chat agents with persistent memory. They share the same integration pattern:

1. Retrieve relevant context from memory
2. Inject context into the agent/LLM
3. Store the exchange after response
4. Processing triggers automatically in background

Prerequisites

Bash

export OPENAI_API_KEY=sk-...

Available Examples

Example	Framework	Run Command
Quickstart	None (core API)	uv run python examples/quickstart.py
Full Demo	None (core API)	uv run python examples/demo.py
PydanticAI	PydanticAI	uv run python examples/pydantic_ai_agent.py
LangGraph	LangGraph + LangChain	uv run python examples/langgraph_agent.py
LlamaIndex	LlamaIndex	uv run python examples/llamaindex_agent.py
CrewAI	CrewAI	uv run python examples/crewai_agent.py
AutoGen	Microsoft AutoGen	uv run python examples/autogen_agent.py

All interactive agents support these commands:

• quit — Flush remaining messages and exit
• flush — Force-process buffered messages
• debug — Show current memory contents