Practical steps to make your website agent-ready

on YouTube ↗30m

" summary

Describes a concrete path to make sites agent-ready by adding WebMCP tools that agents call directly: shows a booking demo and a maze game using WebMCP Imperative API (document.modelContext.registerTool with JSON Schema inputs and execute blocks) and form-based Declarative API (data attributes, tool autosubmit, agent-invoked flag). Covers debugging in Chrome DevTools (breakpoints, inputs/outputs), running the experimental Lighthouse Agentic Browsing audit (milestone 150), and reports live demos that work.

AI-written summary · about

" chapters (11)

1. 0:00Introduction
2. 0:36The Agentic Web
3. 1:54How AI agents see websites and complex journeys
4. 5:28Introduction to WebMCP and tool demonstrations
5. 10:37Defining WebMCP tools (Imperative & Declarative APIs)
6. 15:26Workshop: Making your website agent-ready
7. 17:21Implementing to-do app tools
8. 20:11Testing and debugging WebMCP tools
9. 25:13Declarative API demo
10. 26:26Lighthouse Agentic Browsing audit
11. 27:50Summary and next steps

" related talks

• 22mThe agent-ready web: Simplify user actions with WebMCPTara Agyemang · AI Engineer Europe 2026
• 11mDon't let AI agents push your buttons - use webMCP instead!Khushal Sagar · Web AI Summit 2025
• 23mBuilding Agent Interfaces: Lessons from Chrome DevTools (MCP) for AgentsMichael Hablich · AI Engineer Europe 2026
• 23mIntroducing WebMCP: Agents in the BrowserRachel Nabors · AI Engineer Europe 2026
• 25mArchitecting AI in Angular: WebMCP, A2UI, and the Agent ExperienceAnkita Sood · Commit Your Code 2026
• 53mAI Didn’t Kill the Web, It Moved in!Olivier Leplus, Yohan Lasorsa · AI Engineer Europe 2026
• 20mWebMCP: Stop scraping the Web, start calling itPatrick Brosset · dotAI 2026
• 23mBeyond Scraping: The Future of Web Data Access with WebMCPYanai Edri · ReactNext 2026