https://github.com/bonk1t/agent-os
Self-Improving AI Agent that creates and refines its own tools, running them in a secure sandbox to expand its capabilities.
https://github.com/bonk1t/agent-os
ai ai-agents ai-agents-framework genai workflow-automation
Last synced: 10 days ago
JSON representation
Self-Improving AI Agent that creates and refines its own tools, running them in a secure sandbox to expand its capabilities.
- Host: GitHub
- URL: https://github.com/bonk1t/agent-os
- Owner: bonk1t
- License: agpl-3.0
- Created: 2023-11-16T21:28:36.000Z (almost 2 years ago)
- Default Branch: main
- Last Pushed: 2025-01-19T01:37:17.000Z (9 months ago)
- Last Synced: 2025-01-19T02:26:17.197Z (9 months ago)
- Topics: ai, ai-agents, ai-agents-framework, genai, workflow-automation
- Language: Python
- Homepage:
- Size: 2.8 MB
- Stars: 33
- Watchers: 6
- Forks: 8
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Agent OS Platform
Agent OS Platform is an open-source API and web application for managing LLM-driven multi-agent workflows.
Building on OpenAI's Assistants API, it offers a collaborative environment for developing, testing, and deploying AI teams.
## Architecture
### Backend
- FastAPI application for API and WebSocket endpoints
- Firebase Authentication and Firestore for data persistence
- E2B for secure sandbox execution
- Redis for message bus and state management
### Frontend
- Gatsby-based web application
- TailwindCSS for styling
- Ant Design for UI components
- Real-time updates via WebSocket
## Key Features
- **Configuration Management**: Centrally manage configurations for agencies, agents, and skills
- **Custom Skills**: Extend AI agents with specialized skills
- **Secure Execution**: Isolated sandbox environments for running agent code
- **Real-time Communication**: WebSocket support for live updates
- **Modern UI**: Beautiful and responsive interface with best UX practices
## Getting Started
### Quick Start with Docker
1. Create `.env.docker` from `.env.docker.testing` template
2. Run:
```bash
source .env.docker
FIREBASE_CONFIG=$FIREBASE_CONFIG docker-compose up --build
```
### Local Development
1. Backend (FastAPI):
```bash
cd backend
pip install -r requirements.txt
uvicorn main:app --reload
```
2. Frontend (Gatsby):
```bash
cd frontend
npm install
yarn start
```
For detailed setup instructions, refer to:
- [Backend README](backend/README.md)
- [Frontend README](frontend/README.md)
## Status
> [!NOTE]
> Agent OS Platform is a research project exploring multi-agent workflows.