This project is a client-side data management library and query orchestrator designed to synchronize remote server state with local client state. It functions as a type-safe state manager and cache orchestrator that coordinates data loading across diverse backends, including REST, GraphQL, and WebSockets.
The system distinguishes itself through a durable workflow engine for executing asynchronous functions with persisted state and deterministic replay. It also provides a standardized AI integration adapter to connect large language models to application data, supporting real-time response streaming and tool-based skill execution.
The library covers a broad capability surface including data normalization, optimistic UI updates, and reactive data querying using differential dataflow. It further provides headless state management for forms and chat logic, parallel route data fetching, and a set of automation tools for task queue management and rate limiting.
The project includes a command-line interface for project template installation and ecosystem tool discovery.
