agentweave

AgentWeave 🧵

AgentWeave is a powerful CLI tool for initializing and managing AI agent projects based on LangGraph, making it easier to build, customize, and deploy sophisticated AI agents.

Think of AgentWeave as a "create-react-app" but for AI agents - it helps you bootstrap a complete project with all the necessary components, configurations, and integrations.

Features ✨

• Quick Project Setup: Initialize complete AI agent projects with a single command
• Component Management: Easily add, remove, and update agent components
• Tool Integration: Add and configure agent tools from a growing library of predefined tools
• LangGraph Integration: Built on top of LangGraph for powerful agent orchestration
• Monitoring & Visualization: Built-in components for monitoring, debugging, and visualizing agent behavior
• Memory Management: Easy integration of different memory solutions
• Service Integrations: Connect to various external services and APIs
• Extensible Architecture: Design focused on extensibility and customization
• FastAPI Backend: Well-structured FastAPI backend for serving your agents

Installation 🚀

pip install agentweave

Quick Start 🏁

# Initialize a new agent project
agentweave init {{project_name}}

# Change to the project directory
cd {{project_name}}

# Edit .env to add OpenAI key

# Install the environment (sets up dependencies)
agentweave install_env

# Run your agent in development mode
agentweave run

Environment Setup 🔐

After initializing your project, make sure to:

1. Find the .env.example file in your project
2. Copy it to .env
3. Update the values in .env with your API keys and configuration

This step is essential for your agent to function properly with external services.

Project Commands 📝

Core Commands

# Initialize a new agent project
agentweave init [project-name]

# Run your agent in development mode
agentweave run

# Run the agent in development mode with hot reloading
agentweave dev

# Install environment dependencies
agentweave install_env

Utility Commands

# Display version information
agentweave version

# Display general information about AgentWeave
agentweave info

Project Structure 📂

{{project_name}}/
├── agents/              # Agent definitions
├── tools/               # Custom tools
├── templates/           # Templates for agents and tools
├── memory/              # Memory configurations
├── monitoring/          # Monitoring and visualization
├── frontend/            # UI for interacting with agents
├── config/              # Configuration files
├── README.md            # Project documentation
└── agentweave.yaml      # AgentWeave configuration

Extensibility 🔌

AgentWeave is designed to be highly extensible:

• Create custom templates for new project types
• Develop and share custom tools
• Create plugins for new functionalities
• Integrate with various AI services and platforms

Dependencies

AgentWeave is built using:

• LangGraph v0.0.30+ for agent orchestration
• FastAPI for the backend API
• Typer and Click for the CLI interface
• Rich for beautiful terminal output
• Jinja2 for template rendering
• And more!

Contributing 🤝

We welcome contributions! Please see our Contributing Guidelines for more information.

License 📄

AgentWeave is open source software licensed as MIT.