sasefied 0.1.2

Sasefied is an agentic systems platform for building multi-agent workflows across industries

Sasefied Agent Library

A comprehensive, production-ready library for building intelligent AI agents and complete agentic systems with industry-specific capabilities. Sasefied provides a robust framework for creating specialized agents, orchestrating multi-agent collaborations, complete with pre-built industry solutions, advanced tools, and a sophisticated prompt management system.

🚀 Features

Core Agent Framework

• BaseAgent Class: Extensible foundation for all agent types
• ReAct Architecture: Built on LangGraph for reasoning and acting
• Tool Integration: Seamless integration with LangChain tools
• Multi-LLM Support: Compatible with any LangChain-compatible chat model

Multi-Agent Orchestration

• Agentic Systems: Complete multi-agent ecosystems for complex workflows
• Agent Orchestration: Coordinate multiple specialized agents working together
• Industry Orchestrators: Pre-built systems for airlines, EV batteries, and agriculture
• Communication Protocols: Structured agent-to-agent interactions
• Task Distribution: Intelligent workload management across agents
• Workflow Coordination: Complex multi-step process automation

Industry-Specific Agents

• Airlines: Complete suite of 8 specialized agents for airline operations
  • Passenger Service, Revenue Management, Flight Operations
  • Safety Compliance, Maintenance Engineering, Network Planning
  • Crew Management, Ancillary Revenue
• EV Batteries: Specialized agents for electric vehicle battery lifecycle
  • Battery Technology, Manufacturing & Supply, Safety & Compliance
  • Lifecycle Management, System Orchestrator
• Fruits & Agriculture: Supply chain and quality management agents
  • Customer Service, Market Control, Quality Control
  • Supply Chain, Sustainability Management

Advanced Tools

• DeepSearchAgent: Multi-step web research with DuckDuckGo integration
• HTTP Tool: Generic HTTP request tool for API interactions
• Web Scraping: Static and dynamic content extraction (Selenium support)

Prompt Management Hub

• Agent Prompt Hub: Comprehensive library of pre-built prompts
• CLI Interface: Command-line tools for prompt exploration and management
• Web Interface: Browser-based prompt management system
• Search & Filter: Advanced search across industries and categories
• Export Capabilities: Multiple export formats for prompts

📦 Installation

pip install sasefied

Optional Dependencies

For enhanced web scraping capabilities:

pip install sasefied[scraping]

For web interface:

pip install sasefied[web]

🎯 Quick Start

Basic Agent Usage

from sasefied.agents import DeepSearchAgent
from langchain_openai import ChatOpenAI

# Initialize LLM
llm = ChatOpenAI(model="gpt-4")

# Create a deep search agent
search_agent = DeepSearchAgent(llm=llm)

# Use the agent
result = search_agent.invoke([
    {"role": "user", "content": "Research the latest developments in quantum computing"}
])
print(result["messages"][-1].content)

Industry-Specific Agents

from sasefied.industry.airlines import create_passenger_service_agent
from langchain_openai import ChatOpenAI

llm = ChatOpenAI(model="gpt-4")

# Create airline passenger service agent
agent = create_passenger_service_agent(llm=llm)

# Handle passenger inquiry
response = agent.invoke([
    {"role": "user", "content": "What are the baggage policies for international flights?"}
])

Multi-Agent Agentic Systems

from sasefied.industry.airlines import create_airline_orchestrator
from langchain_openai import ChatOpenAI

llm = ChatOpenAI(model="gpt-4")

# Create complete airline management system
orchestrator = create_airline_orchestrator(llm)

# Coordinate multiple agents for complex operations
result = orchestrator.handle_flight_disruption(
    flight_id="AA123",
    issue="weather_delay",
    passengers=150
)
# Automatically coordinates: Operations, Crew, Passenger Service, Revenue Management

Using the Prompt Hub

from sasefied.hub import AgentPromptExplorerHub

# Initialize the hub
hub = AgentPromptExplorerHub()

# Search for prompts
prompts = hub.search_prompts("customer service", industry="retail")

# Export prompts
hub.export_prompts(prompts, format="json", output_file="customer_prompts.json")

CLI Usage

# Explore available prompts
sasefied-hub explore

# Search for specific prompts
sasefied-hub search "revenue management" --industry airlines

# Export prompts
sasefied-hub export --industry healthcare --format yaml

🏗️ Architecture

sasefied/
├── agents/                 # Core agent framework
│   ├── base.py            # BaseAgent class
│   └── deep_search.py     # DeepSearchAgent implementation
├── industry/              # Industry-specific agents
│   ├── airlines/          # Airline industry agents
│   ├── ev_batteries/      # EV battery industry agents
│   └── fruits/            # Agriculture industry agents
├── hub/                   # Prompt management system
│   ├── core/              # Core models and repository
│   ├── cli.py             # Command-line interface
│   ├── web.py             # Web interface
│   └── hub.py             # Main hub functionality
├── tools/                 # Utility tools
│   └── http.py            # HTTP request tool
└── agentic_systems/       # Multi-agent orchestration

🔧 Configuration

Environment Variables

# OpenAI API (if using OpenAI models)
OPENAI_API_KEY=your_api_key_here

# Optional: Custom model configurations
DEFAULT_MODEL=gpt-4
DEFAULT_TEMPERATURE=0.7

Custom Agent Development

from sasefied.agents.base import BaseAgent
from langchain_core.tools import BaseTool
from langchain_openai import ChatOpenAI

class CustomAgent(BaseAgent):
    def __init__(self, llm: ChatOpenAI, tools: List[BaseTool] = None):
        super().__init__(
            name="CustomAgent",
            description="Your custom agent description",
            tools=tools or [],
            llm=llm
        )
    
    def get_system_prompt(self) -> str:
        return "You are a specialized agent for..."

📚 Documentation

• API Reference
• Agent Development Guide
• Industry Solutions
• Prompt Hub Guide

🤝 Contributing

We welcome contributions! Please see our Contributing Guide for details.

Development Setup

git clone https://github.com/your-org/sasefied.git
cd sasefied
pip install -e ".[dev]"

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🆘 Support

• 📖 Documentation
• 🐛 Issue Tracker
• 💬 Discussions

🌟 Roadmap

• Additional industry modules (Healthcare, Finance, Manufacturing)
• Advanced orchestration patterns
• Performance monitoring and analytics
• Integration with more LLM providers
• Enhanced web scraping capabilities
• Agent marketplace and sharing platform

🏆 Acknowledgments

Built with:

• LangChain - LLM framework
• LangGraph - Agent orchestration
• DuckDuckGo - Search integration

---

Sasefied - Empowering the next generation of intelligent agents.