- Step1: Clone the GitHub repository and install dependencies.
- Step2: Configure your .env with LLM API keys and vector store credentials.
- Step3: Run the development server (npm run dev) and open the UI in your browser.
- Step4: Create a new project and add agents on the canvas.
- Step5: Configure each agent’s prompt, memory, and integrations.
- Step6: Link agents with workflow edges to define task sequences.
- Step7: Test conversations in the built-in simulator.
- Step8: Monitor logs and tweak settings in real time.
- Step9: Deploy to production or export configuration files.
- Step10: Scale by adding more agents or external APIs as needed.