AutoGen Studio

Rapidly prototype AI agents with a visual interface

AutoGen Studio is an AutoGen-powered AI application (user interface) to help you rapidly prototype AI agents, enhance them with skills, compose them into workflows, and interact with them to accomplish tasks. It is built on top of the AutoGen framework.

AutoGen Studio is under active development and is currently not meant to be a production-ready app. Expect breaking changes in upcoming releases.

Installation

There are two ways to install AutoGen Studio: from PyPI or from source. We recommend installing from PyPI unless you plan to modify the source code.

Install from PyPI (Recommended)

We recommend using a virtual environment (e.g., venv) to avoid conflicts with existing Python packages. With Python 3.10 or newer active in your virtual environment, use pip to install AutoGen Studio:

1. Create a virtual environment

   python -m venv autogen-env
   source autogen-env/bin/activate  # On Windows: autogen-env\Scripts\activate
   
2. Install AutoGen Studio

   pip install -U autogenstudio
   

Install from Source

This approach requires familiarity with building interfaces in React and Git LFS.

1. Install Git LFS

   AutoGen Studio uses Git Large File Storage (LFS) for managing image and other large files.

   # On Debian/Ubuntu
   apt-get install git-lfs
   
   # On macOS with Homebrew
   brew install git-lfs
   
   # On Windows with Chocolatey
   choco install git-lfs
   

   Then set up git-lfs:

   git lfs install
   
2. Clone and install dependencies

   git clone https://github.com/microsoft/autogen.git
   cd autogen/python/packages/autogen-studio
   pip install -e .
   
3. Build the UI

   npm install -g gatsby-cli
   npm install --global yarn
   cd frontend
   yarn install
   yarn build
   

   Windows users may need alternative commands to build the frontend. See the README for details.

Running the Application

Once installed, run the web UI by entering the following in your terminal:

autogenstudio ui --port 8081

This command will start the application on the specified port. Open your web browser and go to http://localhost:8081/ to use AutoGen Studio.

CLI Options

AutoGen Studio accepts several parameters to customize the application:

--host · string · default: "localhost"

Specify the host address

--port · number · default: "8080"

Specify the port number

--appdir · string · default: "~/.autogenstudio"

Directory where app files (database, generated files) are stored

--database-uri · string

Database URI for backend storage. Examples:

• SQLite: sqlite:///database.sqlite
• PostgreSQL: postgresql+psycopg://user:password@localhost/dbname

If not specified, defaults to database.sqlite in the --appdir directory.

--reload · boolean · default: "false"

Enable auto-reloading when code changes are made

--upgrade-database · boolean · default: "false"

Upgrade the database schema to the latest version

Example Commands

# Run on custom port
autogenstudio ui --port 8081

# Use PostgreSQL backend
autogenstudio ui --database-uri postgresql+psycopg://user:password@localhost/dbname

# Specify custom app directory
autogenstudio ui --appdir /path/to/folder

# Enable development mode with auto-reload
autogenstudio ui --reload

AutoGen Studio Lite

AutoGen Studio Lite provides a lightweight way to quickly prototype and experiment with AI agent teams without the full database setup. It’s designed for rapid experimentation.

CLI Usage

Launch Studio Lite from the command line:

# Quick start with default team
autogenstudio lite

# Use custom team file
autogenstudio lite --team ./my_team.json --port 8080

# Custom session name with auto-open browser
autogenstudio lite --session-name "My Experiment" --auto-open

Programmatic Usage

Use Studio Lite directly in your Python code:

from autogenstudio.lite import LiteStudio
from autogen_agentchat.teams import RoundRobinGroupChat

# Quick start with default team
studio = LiteStudio()

# Use with custom AutoGen team objects
team = RoundRobinGroupChat([agent1, agent2], termination_condition=...)

# Context manager usage
with LiteStudio(team=team) as studio:
    # Studio runs in background
    # Do other work here
    pass

Database Configuration

AutoGen Studio uses SQLModel (Pydantic + SQLAlchemy) for its database layer. This provides:

• Entity linking (skills, models, agents, and workflows are linked via association tables)
• Support for multiple database backends via SQLAlchemy dialects

Supported Databases

SQLite

Default option, file-based database

autogenstudio ui --database-uri sqlite:///database.sqlite

PostgreSQL

Production-ready relational database

autogenstudio ui --database-uri postgresql+psycopg://user:password@localhost/dbname

MySQL

Popular open-source database

autogenstudio ui --database-uri mysql://user:password@localhost/dbname

Oracle / SQL Server

Enterprise database support

Environment Variables

Place a .env file in your --appdir directory to set environment variables for the app. This file will be automatically loaded when the application starts.

.env

OPENAI_API_KEY=your-api-key-here
AZURE_API_KEY=your-azure-key-here
# Add other environment variables as needed

Project Structure

The AutoGen Studio codebase is organized as follows:

• autogenstudio/ - Backend classes and web API (FastAPI)
• frontend/ - Web UI built with Gatsby and TailwindCSS

Key Features

Visual Agent Builder

Define and configure AI agents through an intuitive interface

Workflow Composition

Compose multiple agents into workflows to accomplish complex tasks

Skills Enhancement

Enhance agents with custom skills and capabilities

Interactive Sessions

Interact with agents in real-time to test and refine behavior

Troubleshooting

Git LFS Issues

If you encounter build errors related to image formats when installing from source:

# Install git-lfs if not already installed
git lfs install

# Fetch all LFS files
git lfs fetch --all

# Checkout LFS files to working directory
git lfs checkout

Database Migration

To upgrade your database schema to the latest version:

autogenstudio ui --upgrade-database

Contributing

We welcome contributions to AutoGen Studio! To contribute:

1. Review the AutoGen contribution guide
2. Check the roadmap for current priorities
3. Use the tag proj-studio for issues, questions, and PRs related to Studio
4. Submit a pull request with your contribution

Resources

GitHub Repository

View source code and contribute

AutoGen Framework

Learn about the underlying AutoGen framework

FAQs

Find answers to common questions

PyPI Package

View package on PyPI