WebMCP Enables Agents To Interact With Webpages
Google's WebMCP proposal now runs client-side in Chrome experimental previews, allowing AI agents to call Declarative and Imperative page APIs. In recent tests (Chrome 146+), the inspector and demo flight app showed agent calls like "listFlights" returning structured JSON and integration with Google's Gemini API. This enables agents to interact directly with DOM elements while keeping human users in the loop for context and clarification.
Key Points
- 1Demonstrates WebMCP client-side APIs enabling agent calls like 'listFlights' to return structured JSON.
- 2Highlights Declarative and Imperative APIs for standard HTML actions and complex JavaScript interactions.
- 3Enables developers to adopt Chrome EPP flags, inspector tooling, and Gemini integration for agent workflows.
Scoring Rationale
Hands-on Chrome preview and structured WebMCP APIs increase practicality, but experimental status and limited rollout constrain near-term impact.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
