Documentation
AgentNexus Python
Python Library for AI/LLM Agent Development with UI-Driven Workflows and Actions
Get Started
AgentNexus enables developers to build powerful AI/LLM agents with interactive UIs and structured workflows. Get started quickly with our step-by-step guides.
Learn what AgentNexus is and how it can help you
Install and build your first agent in minutes
Understand the fundamental concepts of AgentNexus
See AgentNexus in action with practical examples
Key Features
AgentNexus provides a comprehensive framework for developing AI/LLM agents:
- Declarative Development: Create complex agents using intuitive Python decorators
- Multi-Action Support: Define multiple agent actions endpoints within a single agent
- Workflow-Driven Design: Build multi-step UI workflows with state management
- Rich Component Library: Tables, forms, code editors, and markdown displays
- Automatic Event Handling: Simplified component event management
- Context-Aware State: Preserve state across workflow steps with sessions
Quick Example
from agentnexus.base_types import AgentConfig, Capability, ActionType
from agentnexus.action_manager import agent_action
# Define agent configuration
my_agent = AgentConfig(
name="Quick Example Agent",
version="1.0.0",
description="A simple example agent",
capabilities=[
Capability(
skill_path=["Example", "Demonstration"],
metadata={"example": True}
)
]
)
# Create a simple action
@agent_action(
agent_config=my_agent,
action_type=ActionType.GENERATE,
name="Simple Action",
description="A simple example action"
)
async def simple_action(input_data):
"""Handle a simple action."""
return {"result": f"Processed: {input_data.message}"}