A comprehensive web-based management system for designing, configuring, and executing CrewAI projects with a modern UI/UX.
- Visual Project Management: Create and manage CrewAI projects through an intuitive web interface
- Agent Configuration: Design and configure AI agents with custom roles, goals, and backstories
- Task Orchestration: Define tasks, set dependencies, and assign agents
- Universal LLM Support: Integrate with any LLM provider (OpenAI, Anthropic, Google, Ollama, etc.) via LiteLLM
- Tool Library: Access built-in tools and create custom tools with Python
- Project Templates: Quick-start with pre-built templates for common use cases
- Real-time Execution Logs: Monitor crew execution with live WebSocket updates
- Background Task Processing: Asynchronous execution using Celery for better performance
- LangChain Tool Integration: Full integration with DuckDuckGo Search, Wikipedia, and custom tools
- Mem0 Memory System: Agent memory management with search and context retrieval
- Token Usage Dashboard: Comprehensive analytics with charts (usage over time, by agent, by model)
- Visual Flow Designer: Interactive workflow visualization using React Flow
- Multi-Format Export: Export execution results to Excel, Word, and PDF
- Monaco Code Editor: Professional code editing for custom tools
- Execution Detail View: Comprehensive execution logs, status, and results viewer
- FastAPI - Modern Python web framework with WebSocket support
- PostgreSQL - Robust relational database
- SQLAlchemy - ORM for database operations
- CrewAI 0.22.5 - AI agent orchestration framework
- LiteLLM - Universal LLM provider integration (OpenAI, Anthropic, Google, etc.)
- Celery + Redis - Background task processing and message queue
- Mem0 - AI memory management with vector storage (ChromaDB)
- Alembic - Database migrations
- React 18 - Modern UI framework
- TypeScript - Type-safe development
- Vite - Fast build tool and HMR
- Tailwind CSS - Utility-first styling
- React Query (TanStack Query) - Data fetching, caching, and synchronization
- Zustand - Lightweight state management
- Monaco Editor - VS Code-powered code editor
- Recharts - Composable charting library
- React Flow - Interactive node-based UI
- Docker & Docker Compose (recommended)
- OR Python 3.11+, Node.js 18+, PostgreSQL 16, Redis 7
- Clone the repository:
git clone <repository-url>
cd CrewAI- Start all services:
docker-compose up -d- Access the application:
- Frontend: http://localhost:5173
- Backend API: http://localhost:8000
- API Documentation: http://localhost:8000/docs
- Initialize the system:
- Navigate to http://localhost:5173/initialize
- Click "Initialize System"
- Wait for initialization to complete
See SETUP.md for detailed installation instructions.
crewai-manager/
├── backend/ # FastAPI backend
│ ├── app/
│ │ ├── api/ # API endpoints
│ │ ├── models/ # Database models
│ │ ├── schemas/ # Pydantic schemas
│ │ ├── services/ # Business logic
│ │ └── utils/ # Utilities
│ ├── alembic/ # Database migrations
│ └── requirements.txt
│
├── frontend/ # React frontend
│ ├── src/
│ │ ├── components/ # React components
│ │ ├── pages/ # Page components
│ │ ├── services/ # API services
│ │ ├── types/ # TypeScript types
│ │ └── stores/ # State management
│ └── package.json
│
└── docker-compose.yml # Docker orchestration
- Navigate to the Projects page
- Click "Create Project"
- Enter project details
- Add agents with roles and goals
- Define tasks and assign agents
- Execute the project
- Go to LLM Providers
- Click "Add Provider"
- Select provider (OpenAI, Anthropic, etc.)
- Enter API key and configuration
- Test the connection
- Open a project
- Click "Execute Project"
- Provide input data
- Monitor real-time execution logs
- View results and token usage
Interactive API documentation is available at:
- Swagger UI: http://localhost:8000/docs
- ReDoc: http://localhost:8000/redoc
Core APIs:
POST /api/v1/initialize- Initialize systemGET /api/v1/projects- List projectsPOST /api/v1/projects- Create projectGET /api/v1/llm-providers- List LLM providersPOST /api/v1/projects/{id}/execute- Execute project
Advanced APIs:
GET /api/v1/agents/{id}/memory- Get agent memoriesPOST /api/v1/agents/{id}/memory/search- Semantic memory searchGET /api/v1/token-usage?range=7d- Token usage statisticsGET /api/v1/executions/{id}/export?format=excel- Export results (excel/word/pdf/json/html/markdown)WS /ws/execution/{id}- WebSocket for real-time execution updates
cd backend
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
pip install -r requirements.txt
uvicorn app.main:app --reloadcd frontend
npm install
npm run devcd backend
# Create migration
alembic revision --autogenerate -m "Description"
# Apply migrations
alembic upgrade headDATABASE_URL=postgresql://crewai:crewai123@localhost:5432/crewai_db
REDIS_URL=redis://localhost:6379
SECRET_KEY=your-secret-key
ENCRYPTION_KEY=your-encryption-key
VITE_API_URL=http://localhost:8000
VITE_WS_URL=ws://localhost:8000
- API keys are encrypted using Fernet encryption
- Environment variables for sensitive data
- CORS configuration for production
- Input validation on all endpoints
- Fork the repository
- Create a feature branch
- Make your changes
- Submit a pull request
MIT License - see LICENSE file for details
For issues and questions:
- GitHub Issues: [Create an issue]
- Documentation: See SETUP.md for detailed setup instructions
✅ Completed Features:
- Real-time execution monitoring with WebSocket
- Celery background task processing
- LangChain tools integration (DuckDuckGo, Wikipedia)
- Mem0 memory management system
- Token usage analytics dashboard
- React Flow visual designer
- Export to Excel/Word/PDF
- Comprehensive execution logs viewer
- Multi-user authentication
- Custom tool marketplace
- Export project to standalone Python code
- Scheduled/recurring executions
- Team collaboration features
- Advanced workflow conditions
- Email notifications for execution completion
Current Version: 1.0.0
CrewAI Manager Development Team