Markdown source

MCP Apps as Agentic App Runtime

Overview

MCP Apps as an agentic app runtime is the idea that MCP servers can return interactive UI, not only structured text or JSON, so agents and humans can operate richer task surfaces inside MCP hosts.

Significance

MCP Apps - Extending the frontier presents MCP Apps as a path from tool calls to interactive software experiences. This matters when a task needs both model-readable data and human-operable UI.

Applied Use

• Keep server-returned UI scoped to the task.
• Separate model data from user interaction state.
• Design for host constraints and permissions.
• Use MCP Apps when text-only tool output would force awkward back-and-forth or when human review is part of the workflow.

Connections

• liad yosef
• ido salomon
• mcp apps
• mcp
• agentic web

Evidence Graph

This evidence graph consolidates scheduled talks, linked videos, transcripts, and slide-derived material connected to this topic.

Linked Sessions

• MCP Apps - Extending the frontier
• Rebuilding the web for agents

Media Signals

• youtube-o-zkvb0iFDQ — 3,969 transcript words; 1 slide-derived text signals
• Transcript signals for youtube-o-zkvb0iFDQ: apps, host, claude, back, chatgpt, look, mcpui, chat.
• Slide-derived themes for youtube-o-zkvb0iFDQ: engineering, future.
• Evidence links for youtube-o-zkvb0iFDQ: youtube o zkvb0iFDQ, youtube o zkvb0iFDQ transcript, youtube o zkvb0iFDQ slides, youtube o zkvb0iFDQ dense slides, youtube o zkvb0iFDQ reconstructed slides

Evidence Boundary

This page is grounded in official schedule context and related MCP UI/MCP Apps transcript evidence. It should be updated as MCP Apps implementation details and SEP-1865 adoption evolve.