Baicai Base: AI Agent Framework Foundation

Baicai Base is the foundational framework for building intelligent AI agents for education using the LangGraph architecture. It provides a extensible foundation for creating multi-agent workflows that can collaborate, reason, and execute complex tasks.

🚀 Key Features

🤖 Multi-Agent Architecture

• ReAct Pattern Implementation: Built-in ReAct (Reasoning + Acting) agent framework
• Modular Node System: Extensible node-based architecture for custom agent behaviors
• State Management: Sophisticated state handling with memory persistence
• Conditional Routing: Intelligent routing logic with fail-fast and retry mechanisms

🔧 Core Components

• Base Graph Framework: Abstract base classes for building custom agent graphs
• Code Execution Engine: Safe code execution with debugging capabilities
• LLM Integration: Support for OpenAI-compatible APIs and Groq
• Configuration Management: Flexible configuration system with environment-based settings

🛠️ Developer Tools

• Rich Logging: Comprehensive logging with color-coded output
• Graph Visualization: Built-in Mermaid.js graph visualization
• Code Interpreter: Integrated code execution and debugging
• Memory Management: Persistent state and conversation memory

📋 Requirements

• Python: 3.10 or higher (tested up to 3.11)
• Dependencies: Managed via Poetry
• LLM Access: OpenAI API key or Groq API key

🛠️ Installation

1. Clone the Repository

git clone https://github.com/yourusername/baicai.git
cd baicai/baicai_base

2. Install Poetry (if not already installed)

# macOS/Linux
curl -sSL https://install.python-poetry.org | python3 -

# Windows (PowerShell)
(Invoke-WebRequest -Uri https://install.python-poetry.org -UseBasicParsing).Content | python -

3. Install Dependencies

poetry install

4. Set Up Environment

# Set up your API keys
export OPENAI_API_KEY="your-openai-api-key"
# or
export GROQ_API_KEY="your-groq-api-key"

🏗️ Architecture Overview

Core Components

baicai_base/
├── agents/
│   ├── graphs/           # Graph implementations
│   │   ├── base_graph.py # Abstract base class
│   │   ├── react_coder_graph.py # ReAct pattern implementation
│   │   └── nodes/        # Node implementations
│   └── roles/            # Agent role definitions
├── configs/              # Configuration management
├── services/             # Core services (LLM, etc.)
└── utils/                # Utility functions and helpers

🚀 Quick Start

Basic ReAct Agent Usage

from baicai_base.agents.graphs import ReActCoder
from baicai_base.services import LLM

# Initialize LLM
llm_service = LLM()

# Create ReAct agent
agent = ReActCoder(
    graph_name="MyAgent",
    llm=llm_service.llm,
    need_helper=True
)

# Define a question for the AI model to predict
question = "code a simple function to calculate the sum of two numbers"

# Invoke the AI model with the question and get the results
results = agent.app.invoke({"messages": [("user", question)]}, agent.config)

See react_coder_builder.ipynb for more

📚 Documentation

Core Concepts

• Agent Graphs: Understanding the graph-based architecture
• Node System: Building custom nodes and behaviors
• State Management: Managing agent state and memory
• Configuration: Setting up and managing configurations

API Reference

• BaseGraph: Abstract base class for all agent graphs
• ReActCoder: ReAct pattern implementation
• LLM Service: Language model integration
• Node Classes: Available node types and interfaces

📄 License

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

🙏 Acknowledgments

Special thanks to:

• LangGraph Team: For the foundational graph framework
• LangChain Community: For the excellent LLM integration tools
• OpenAI & Groq: For providing powerful language models
• All Contributors: For helping build and improve this framework