🤖 AgentOS — The Operating System for AI Workers

Build, deploy, and manage AI agents with 5 lines of code.

AgentOS is an open-source framework that makes it dead simple to create AI agents that can use tools, remember context, and work together in swarms. Think of it as the operating system for your AI workforce.

from agentos import Agent, tool

@tool
def search_web(query: str) -> str:
    """Search the web."""
    return f"Results for: {query}"

agent = Agent(
    name="Researcher",
    instructions="You are a helpful research assistant.",
    tools=[search_web],
)

result = agent.run("What's the latest in AI?")
print(result)

✨ Why AgentOS?

🚀 Quick Start

Installation

pip install agentos

Set your API key:

export OPENAI_API_KEY="sk-..."

Your First Agent

from agentos import Agent

agent = Agent(
    name="Helper",
    instructions="You are a helpful assistant. Be concise.",
)

result = agent.run("Explain AI agents in one sentence.")
print(result)

Agent with Tools

from agentos import Agent, tool

@tool
def get_weather(city: str) -> str:
    """Get current weather for a city."""
    return f"Sunny, 72°F in {city}"

agent = Agent(
    name="WeatherBot",
    instructions="You help people check the weather.",
    tools=[get_weather],
)

result = agent.run("What's the weather in Tokyo?")

Multi-Agent Swarm

from agentos import Agent, Swarm

researcher = Agent(name="Researcher", instructions="Research topics thoroughly.",
                   tools=[web_search])
writer = Agent(name="Writer", instructions="Write clear reports.")

swarm = Swarm(agents=[researcher, writer])
outputs = swarm.run_sequential([
    ("Researcher", "Research AI trends in 2026"),
    ("Writer", "Write a report based on: {Researcher_output}"),
])

CLI

# Run a one-shot task
agentos run "What is 2+2?"

# Start interactive chat
agentos chat

# Run a multi-agent swarm
agentos swarm "Research and write a report" -c swarm_config.json

# Start the API server
agentos serve

API Server

agentos serve -p 8520

Then call the API:

curl -X POST http://localhost:8520/api/agents/run \
  -H "Content-Type: application/json" \
  -d '{"task": "Explain quantum computing in 2 sentences"}'

🛠️ Built-in Tools

AgentOS comes with 7 production-ready tools:

🏗️ Architecture

┌─────────────────────────────────────┐
│            AgentOS Platform          │
├─────────────────────────────────────┤
│  CLI  │  Python SDK  │  REST API    │
├─────────────────────────────────────┤
│          Core Engine                 │
│  ┌─────────┐  ┌──────────────────┐  │
│  │  Agent  │  │   Orchestrator   │  │
│  │  (LLM)  │  │   (Swarm)        │  │
│  └─────────┘  └──────────────────┘  │
│  ┌─────────┐  ┌──────────────────┐  │
│  │  Tools  │  │    Memory        │  │
│  └─────────┘  └──────────────────┘  │
├─────────────────────────────────────┤
│  OpenAI │ Anthropic │ Local Models  │
└─────────────────────────────────────┘

🗺️ Roadmap

• Core Agent engine with tool calling
• Multi-agent orchestration (Swarm)
• CLI + REST API server
• Built-in tools library
• Web Dashboard (agent monitoring)
• Memory persistence (SQLite/Postgres)
• Streaming responses
• Anthropic Claude support
• Local model support (Ollama)
• Agent marketplace
• Enterprise SSO & RBAC

🤝 Contributing

AgentOS is Apache 2.0 licensed. We welcome contributions!

git clone https://github.com/agentos-ai/agentos.git
cd agentos
pip install -e ".[dev]"
pytest

📄 License

Apache 2.0 — free for personal, academic, and commercial use.

Built with ❤️ by the AgentOS team. The OS for your AI workforce.