World Wide Web Consortium (W3C) participants, including Google and Microsoft, launched the WebMCP API, a JavaScript interface that allows web applications to provide client-side “tools” to AI agents. The API allows agents to interact directly with web pages and participate in collaborative workflows with human users within the same web interface.
WebMCP is in early preview at Google, Andre Cipriani Bandarra, Google’s developer relations engineer for Chrome and Web, said in a Feb. 10 blog post. “WebMCP aims to provide a standard way to expose structured tools to enable AI agents to take actions on your site faster, more reliably, and more accurately,” said Bandara.
A draft community group report on WebMCP was published by the W3C Web Machine Learning Community Group on February 12th. The report describes the WebMCP API as a JavaScript interface that allows web developers to expose web application functionality as a “tool.” This refers to JavaScript functions with natural language descriptions and structured schemas that can be called by agents, browser agents, and assistive technologies. According to the report, web pages that use WebMCP appear as Model Context Protocol (MCP) servers that implement tools in client-side scripts rather than backends, enabling collaborative workflows where users and agents collaborate within the same web interface. Editors of the report include Khusal Sagar and Dominic Farolino from Google and Brandon Walderman from Microsoft. The specification is not a W3C standard or listed on the W3C Standards Track, the report said.
