AgenticFleet

A powerful multi-agent system for adaptive AI reasoning and automation. AgenticFleet combines Chainlit's interactive interface with AutoGen's multi-agent capabilities to create a flexible, powerful AI assistant platform.

demo.mp4

Table of Contents

1. Introduction
2. System Architecture
3. Quick Start
   • Installation & Environment Setup
   • Starting AgenticFleet
   • Using Docker
4. Installation Guide
   • Prerequisites
   • Installation Steps
   • Troubleshooting Installation
   • Optional Feature Sets
   • Warning About Editable Installations
5. Model Provider Installation
6. Supported Model Providers
7. Key Features
8. Configuration
9. Error Handling
10. Community Contributions
11. Star History

Introduction

AgenticFleet operates through a coordinated team of specialized agents that work together to provide advanced AI capabilities. This project leverages Chainlit's interactive interface and AutoGen's multi-agent system to deliver robust and adaptive solutions.

System Architecture

graph TD
    User[Chainlit UI] -->|HTTP| App[app.py]
    App --> AgentTeam[MagenticOneGroupChat]
    AgentTeam --> WebSurfer
    AgentTeam --> FileSurfer
    AgentTeam --> Coder
    AgentTeam --> Executor
    WebSurfer -->|Selenium| Web[External Websites]
    FileSurfer -->|OS| FileSystem[Local Files]
    Executor -->|Subprocess| Code[Python/Runtime]

Loading

• WebSurfer: Navigates the web, extracts data, and processes screenshots.
• FileSurfer: Manages file operations and extracts information from local files.
• Coder: Generates and reviews code, ensuring quality and efficiency.
• Executor: Executes code safely in an isolated environment and provides feedback.

Quick Start

Installation & Environment Setup

Before starting AgenticFleet, install the package using the uv package manager:

uv pip install agentic-fleet

Then, set up your environment:

1. Copy the Example File

   cp .env.example .env

2. Configure Environment Variables

   Open the .env file and set the required values. At a minimum, configure your Azure OpenAI settings:

   # Required: Azure OpenAI Configuration
   AZURE_OPENAI_API_KEY=your_api_key
   AZURE_OPENAI_ENDPOINT=your_endpoint
   AZURE_OPENAI_DEPLOYMENT=your_deployment
   AZURE_OPENAI_MODEL=your_model

Starting AgenticFleet

After installing the package and configuring your environment, start AgenticFleet using one of the following commands:

# Start with OAuth enabled (default)
agenticfleet start

# Or start without OAuth
agenticfleet start no-oauth

Using Docker

If you prefer using Docker, follow these instructions:

# Pull the latest image
docker pull qredence/agenticfleet:latest

# Run with minimum configuration (replace placeholders with your actual values)
docker run -d -p 8001:8001 \
  -e AZURE_OPENAI_API_KEY=your_key \
  -e AZURE_OPENAI_ENDPOINT=your_endpoint \
  -e AZURE_OPENAI_DEPLOYMENT=your_deployment \
  -e AZURE_OPENAI_MODEL=your_model \
  qredence/agenticfleet:latest

# Alternatively, run with additional configuration including OAuth
docker run -d -p 8001:8001 \
  -e AZURE_OPENAI_API_KEY=your_key \
  -e AZURE_OPENAI_ENDPOINT=your_endpoint \
  -e AZURE_OPENAI_DEPLOYMENT=your_deployment \
  -e AZURE_OPENAI_MODEL=your_model \
  -e USE_OAUTH=true \
  -e OAUTH_GITHUB_CLIENT_ID=your_client_id \
  -e OAUTH_GITHUB_CLIENT_SECRET=your_client_secret \
  qredence/agenticfleet:latest

# To run without OAuth:
docker run -d -p 8001:8001 \
  -e AZURE_OPENAI_API_KEY=your_key \
  -e AZURE_OPENAI_ENDPOINT=your_endpoint \
  -e USE_OAUTH=false \
  qredence/agenticfleet:latest

Installation Guide

Prerequisites

• Python Version: 3.10-3.12
• Operating Systems: macOS, Linux, Windows

Installation Steps

1. Install uv Package Manager

   uv is a fast and efficient package manager. Choose your preferred installation method:

   macOS/Linux

   # Using pip
   pip install uv
   
   # Using Homebrew (macOS)
   brew install uv
   
   # Using curl
   curl -LsSf https://astral.sh/uv/install.sh | sh

   Windows

   # Using pip
   pip install uv
   
   # Using winget
   winget install uv

2. Create and Activate a Virtual Environment

   # Create a new virtual environment
   uv venv
   
   # Activate the virtual environment
   # On macOS/Linux
   source .venv/bin/activate
   
   # On Windows
   .venv\Scripts\activate

3. Install AgenticFleet

   # Install the latest stable version
   uv pip install agentic-fleet
   
   # Install Playwright for web automation and scraping (needed by WebSurfer)
   uv pip install playwright
   playwright install --with-deps chromium

   Playwright Installation Notes:

   • Installs the Chromium browser for web automation.
   • Includes necessary browser dependencies.
   • Required for web scraping and browser-based agents.
   • Supports both headless and headed modes.

4. Verify Installation

   # Check installed version
   uv pip show agentic-fleet
   
   # Run a quick version check
   python -c "import agentic_fleet; print(agentic_fleet.__version__)"

Troubleshooting Installation

• Ensure you're using Python 3.10-3.12.
• Update uv to the latest version: pip install -U uv.
• If issues arise, consult our GitHub Issues.

Optional Feature Sets

# Install with optional telemetry features
uv pip install 'agentic-fleet[telemetry]'

# Install with optional tracing features
uv pip install 'agentic-fleet[tracing]'

Warning About Editable Installations

DO NOT use -e unless you are a core contributor.
Editable installations are not supported in production, may introduce unexpected behaviors, and void package support. They are intended solely for package development. If you make local modifications, please file a GitHub issue and submit a pull request.

Model Provider Installation

Please refer to the existing documentation or the docs/installation.md file for details on installing model providers.

Supported Model Providers

AgenticFleet supports multiple LLM providers including OpenAI, Azure OpenAI, Google Gemini, DeepSeek, Ollama, Azure AI Foundry, and CogCache. For specifics on configuration and usage, please refer to the detailed sections in the documentation.

Key Features

• Advanced multi-agent coordination
• Support for several LLM providers
• GitHub OAuth authentication (optional)
• Configurable agent behaviors and execution isolation
• Comprehensive error handling and automated recovery
• Multi-modal content processing (text, images, etc.)

Configuration

For complete configuration details, review the .env.example file and the docs/usage-guide.md for further instructions.

Error Handling

AgenticFleet includes robust error handling:

• Graceful degradation on failures
• Detailed error logging and reporting
• Automatic cleanup and session recovery
• Execution timeout management

Community Contributions

AgenticFleet welcomes contributions from the community. Please review our CONTRIBUTING.md and CODE_OF_CONDUCT.md for guidelines on submitting issues and pull requests.

Star History