karnavals

Karnavals uses a schema-based approach for orchestrating functions across multiple models, enabling seamless integration with various APIs and services. This architecture allows developers to define function signatures and expected inputs/outputs in a structured format, facilitating easier debugging and maintenance. By leveraging the Model Context Protocol (MCP), it ensures that context is preserved across function calls, enhancing the overall efficiency of multi-step workflows.

Karnavals implements a contextual state management system that tracks the state of interactions across multiple API calls. This is achieved through a centralized context store that retains relevant information, allowing developers to access and manipulate state data easily. The design leverages a reactive programming model to ensure that state changes are propagated to all dependent functions, enhancing responsiveness and reducing errors in multi-step processes.

Karnavals supports integration with multiple AI model providers through a unified API interface. This is accomplished by abstracting the differences between various model APIs and providing a consistent method for developers to interact with them. The system employs an adapter pattern, allowing for easy addition of new providers without modifying existing code, thus promoting extensibility and flexibility.