Aihu is a framework designed for creating durable Web Components that can be interacted with by AI agents, moving beyond the limitations of disposable user interfaces. Through Aihu, agents can directly interact with and manipulate real components using llms.txt and the Model Context Protocol (MCP). This ensures that the UI remains consistent and reusable, as the agent interacts with the actual component displayed on the screen rather than a transient interface that is regenerated constantly.

Key Features

• Durable Components: Aihu enables the creation of custom elements that can be efficiently managed and interacted with by AI, ensuring that components are both trustworthy and reusable.
• Author SFCs: With Aihu, single-file components (.aihu files) encapsulate markup, state, styles, and agent interfaces, allowing for streamlined development. A built-in Rust compiler emits standards-based custom elements alongside the agent manifest, eliminating the need for a separate API layer.
• High Performance: The framework boasts a reactive runtime that is sub-2kB in size, with vanilla custom elements output and zero runtime dependencies. Targeted updates to the DOM are executed 122 times faster than traditional methods, providing an efficient user experience.

npx create-aihu my-app --template agent   # Generate a durable component with AI capabilities in a single command

Built-In Meta-Framework

Aihu includes a complete meta-framework that encompasses routing, server-side rendering, authentication, data loading, and cloud adapters. It has been built with the goal of creating applications where every component is callable by an AI agent. Here are some core functionalities:

• Reactive Core: Built on tiny reactive primitives that support signals, computed values, and effects, facilitating efficient updates and rendering.
• Integrated Routing: File-based routing, server-side rendering, and data loading form a cohesive application structure, allowing developers to focus on building applications rather than integrating multiple libraries.
• AI-Agent Interface: Components can declare their exposed state and actions directly, with the framework auto-generating necessary metadata and protocol definitions.

Developer Experience

The Aihu CLI enhances the developer experience by providing commands for scaffolding, building, and managing applications. A VS Code extension offers support for syntax highlighting and future enhancements through a language server. Additionally, an extensive portfolio of examples showcases the framework’s capabilities, making it easier for developers to get started.

Compliance and Accessibility

Aihu prioritizes compliance with industry standards, ensuring that every application created is discoverable by AI tools. The framework adheres to the Model Context Protocol (MCP) specifications and incorporates accessibility-oriented UI primitives, making it suitable for a wide range of applications.

For a deeper look into its capabilities and to run examples, Aihu provides a range of real-world applications including functionalities such as live updates, data-driven UIs, and integrated agent interactions.

Explore the full potential of Aihu, a pioneering framework that redefines the interaction between AI and web components, enabling the creation of rich, durable, and responsive applications.