Starter AI Agents: Quick-Start Frameworks

Explore 19 starter agent projects built with popular AI frameworks including Agno, LangChain, CrewAI, OpenAI, LlamaIndex, and more.

The Starter Agents category contains 19 quick-start projects designed to help you explore different AI frameworks and patterns. Each project is self-contained and demonstrates how to build, configure, and run a basic agent or assistant with a specific framework. These are ideal if you are comparing tools or learning the fundamentals before building production pipelines.

Starter Projects Overview

Project	Framework	Description
Agno HackerNews	Agno	Trend analysis agent for HackerNews
OpenAI SDK Starter	OpenAI Agents SDK	Email helper and haiku writer examples
LlamaIndex Task Manager	LlamaIndex	Task assistant powered by LlamaIndex
CrewAI Research Crew	CrewAI	Multi-agent research team example
Letta Starter	Letta	Stateful agent with persistent long-term memory across sessions
PydanticAI Weather Bot	PydanticAI	Real-time weather information agent
LangChain Starter	LangChain	Tool-calling agent with AgentExecutor, powered by Nebius
LangGraph Starter	LangGraph	Prebuilt ReAct agent with custom tools, powered by Nebius
AWS Strands Starter	AWS Strands	Weather report agent using AWS Strands SDK
Mastra Starter	Mastra	TypeScript-first agent with custom tools powered by Nebius
Camel AI Starter	Camel AI	Benchmarking tool comparing various AI models
DSPy Starter	DSPy	Building and optimizing AI systems with the DSPy framework
Google ADK Starter	Google ADK	Google Agent Development Kit starter template
Semantic Kernel Starter	Semantic Kernel	ChatCompletionAgent with plugin-based tool calling
smolagents Starter	smolagents	Hugging Face code-first web-search agent
AutoGen Starter	AutoGen	AssistantAgent with custom tools, powered by Nebius
cagent Starter	cagent	Open-source customizable multi-agent runtime by Docker
Sayna Voice Agent	Sayna	Real-time voice infrastructure with multi-provider STT/TTS and WebSocket streaming
KAOS Starter	KAOS	Kubernetes-native multi-agent system with MCP tools and in-cluster LLM

How to Use a Starter

1. Navigate into a project directory

   cd starter_ai_agents/<project_name>
   
2. Set up environment variables

   cp .env.example .env
   

   Add the API keys listed in the project README.

3. Install and run

   Install dependencies with pip or uv, then run the script:

   pip install -r requirements.txt
   python main.py
   

If you are new to agent frameworks, Agno, LangChain, and OpenAI SDK starters are excellent first projects because the README documentation is detailed and the run commands are simple.