KNOWLEDGE DOSSIER · Agents

WebMCP: JavaScript tools exposed by a page

Study the experimental API and the exact status of the community draft.

Edition: 13 November 2026 Extended dossier Scientific review pending
AT A GLANCE
TypeWebMCP: JavaScript tools exposed by a page
AudienceLearners, creators, developers and digital decision-makers
01

What WebMCP proposes

WebMCP proposes a JavaScript interface for pages to expose functions as callable tools for agents. Natural language descriptions and structured input schemas help discover features normally activated by users.

02

Status as of 13 November 2026

The Web Machine Learning Community Group report remains a community draft rather than a W3C Recommendation. The imperative API is documented; a declarative approach built around HTML forms is explored separately and must not be described as universally implemented.

03

User-interface control

Exposed tools should describe expected effects and honor the page’s authorization context. A function available in JavaScript is not automatically authorized to run without human consent.

04

Experimental prototype

Start with a reversible operation such as a search filter, then verify browser support. Do not imply that a live agent is available when the example only demonstrates a schema or simulation.

05

Community status and exposed functions

The proposed imperative WebMCP API explores how pages declare functions callable by agents. The Community Group report is not a W3C Recommendation and browser availability needs checking. Input schemas help describe parameters but do not replace execution authorization or application safeguards.

06

Exercise: side-effect-free navigation function

Describe a function opening a course page from an allowed identifier. Check malformed parameters produce understandable errors and the HTML UI remains keyboard accessible. In a demo, distinguish conceptual descriptions from actually supported APIs; do not claim an external agent is connected when no real invocation occurs.

HOW IT FITS TOGETHER

Core steps

  1. 01Read the community draft
  2. 02Model a harmless action
  3. 03Check actual browser support
  4. 04Review permission boundaries
CRITICAL THINKING

Three questions to ask

  • Is this an adopted W3C Recommendation?

  • Is the declarative form API part of the current core?

  • Is the exposed action actually authorized?

PRIMARY SOURCES

Read the source documents

Versions, reuse terms and document status can change. Check them on the original sites.