nomos 0.3.7

Configurable multi-step agent framework for building advanced LLM-powered assistants

[!NOTE] Looking for client-side integration? Check out our TypeScript/JavaScript SDK.

NOMOS is a framework for building advanced LLM-powered assistants with structured, multi-step workflows. It helps you create sophisticated AI agents through configurable flows, tools, and integrations — making complex agent development accessible from no-code to full-code approaches.

pip install nomos[cli]

To learn more about NOMOS, check out the documentation. If you're looking for quick prototyping, try our Playground for drag-and-drop agent creation.

Table of Contents

• Why use NOMOS?
• NOMOS Ecosystem
• Key Features
• Documentation
• Additional Resources

Complete Documentation | Try Playground | Quick Start Guide

---

Why use NOMOS?

NOMOS helps developers build sophisticated AI agents through structured workflows and configurable components, making complex agent development accessible to teams of all skill levels.

Use Case	Description
Multi-step Workflows	Complex, stateful interactions with specialized tools and intelligent routing
Rapid Prototyping	No-code Playground → YAML config → Full Python implementation
Tool Integration	Python functions, CrewAI, LangChain tools, External APIs with auto-documentation
Production Deployment	Built-in session management, error handling, and monitoring

NOMOS Ecosystem

Component	Description	Link
Playground	Drag-and-drop flow designer for rapid prototyping	Try it live →
TypeScript SDK	Full-featured client library for web and Node.js	Documentation →
Docker Images	Pre-configured containers with Redis, PostgreSQL support (GitHub Packages)	Deployment Guide →
CLI Tools	Complete toolkit: init, run, serve, test, validate, schema	CLI Reference →

Key Features

Category	Feature	Description
Architecture	Step-based Flows	Define agent behavior as sequences of steps with tools and transitions
	Advanced Flow Management	Organize steps into flows with shared context and components
	Flow Memory	Each flow maintains context with intelligent cross-flow summarization
Development	Multiple Config Options	Python API or declarative YAML configuration
	Playground	Drag-and-drop interface for designing flows Try it live →
	Interactive CLI	Bootstrap agents with nomos init, run with nomos run
Tools & Integration	Tool Integration	Register Python functions, packages, CrewAI, External APIs or LangChain tools
	Auto Documentation	Tool descriptions generated from docstrings
	External Packages	Reference any Python package function via configuration
	External API Tools	Use any REST API with ease (No wrapping needed.)
LLM Support	Multiple Providers	OpenAI, Mistral, Gemini, Ollama, and HuggingFace
	Structured Responses	Step-level answer models for JSON/object responses
	Persona-driven	Consistent, branded agent responses
	Decision Examples	Retrieve relevant examples to guide step decisions
Production Ready	Session Management	Redis/PostgreSQL storage for conversation persistence
	Error Handling	Built-in recovery with configurable retry limits
	API Integration	FastAPI endpoints for web and WebSocket interaction
	Monitoring	Elastic APM tracing and distributed monitoring
	Docker Deployment	Pre-built base image for rapid deployment
Extensibility	Custom Components	Build your own tools, steps, and integrations
	Scalable Design	Horizontal scaling with stateless architecture

Documentation

For detailed information, check out our comprehensive guides:

Guide	Description
Getting Started	Installation, setup, and your first agent
CLI Usage	Complete command-line interface guide
Configuration	Python API and YAML configuration
Flow Management	Advanced workflow organization
LLM Support	Supported models and providers
Examples	Real-world implementation examples
Deployment	Production deployment strategies
Community	Contributing, support, and project information

Additional Resources

• Tutorials: Step-by-step guides for getting started with NOMOS, from installation to your first agent.
• How-to Guides: Quick, actionable code snippets for common tasks like tool integration, flow management, and deployment.
• Examples: Real-world implementations including a barista ordering system, financial advisor, and travel planner.
• API Reference: Detailed documentation on Python API and YAML configuration options.
• CLI Reference: Complete command-line interface documentation for development and deployment.

Join the NOMOS community! For roadmap, support, contributing guidelines, and more, see our Community Guide.