Web FAQs

Question 1

How does MCP-Web achieve human-AI parity?

Accepted Answer

MCP-Web makes the frontend the primary control surface for both humans and AI. Humans interact via the UI, while AI agents control the same underlying state and actions through MCP tools, ensuring identical capabilities and smooth transitions between human and AI control.

Question 2

What are 'visual tools' or 'MCP Apps'?

Accepted Answer

MCP Apps are visual tools built from components that render interactive UI directly inline within AI chat interfaces. This feature allows AI agents to not only invoke actions but also interact with and present visual elements, enhancing complex task execution.

Question 3

How do I get started with integrating MCP-Web into my application?

Accepted Answer

To get started, install `@mcp-web/core` in your frontend, then use it to add custom tools or expose your application's state. Finally, run the `@mcp-web/bridge` and connect your MCP-compatible AI agent via its URL or a local client.

Question 4

What is MCP-Web and what problem does it solve?

Accepted Answer

MCP-Web enables AI agents to directly control frontend web applications by exposing the app's state, actions, and user interfaces as type-safe Model Context Protocol (MCP) tools. This allows for powerful automation and seamless AI interaction with web UIs.

Question 5

Is MCP-Web compatible with different frontend frameworks?

Accepted Answer

Yes, MCP-Web is framework-agnostic. It works seamlessly with any frontend framework, including React, Vue, Svelte, and vanilla JavaScript, providing flexible integration options for developers.

Web

Web

Key Features

Use Cases

Key Features

Use Cases