A developer-focused platform for debugging AI agent executions through structured execution traces.
The project aims to help developers understand how an AI agent reached a particular result by capturing and visualizing model calls, tool calls, retrieved context, state changes, and other execution steps.
🚧 In Development
The project is currently in the planning and initial development phase.
- Backend: Python, FastAPI
- Database: PostgreSQL, JSONB
- Frontend: React, JavaScript
- Deployment: Docker
Detailed architecture, protocols, and development documentation will be added under the docs/ directory as the project evolves.
Licensed under the Apache License 2.0.