Telerik and Kendo Meet WebMCP: Turning Your Telerik & Kendo UI Apps into Agent-Ready Workspaces

For some time, AI agents have tried to use the web the way humans do—by looking at pixels, guessing at buttons and clicking around. It mostly works. Sometimes. Until a label changes, a layout shifts or a modal pops up at the wrong moment. 

There is a better way—one where your application tells the agent exactly what it can do, in a language the agent natively understands. That language is WebMCP. 

With our Q2 2026 release, every Blazor, Angular or React application built on Progress Telerik and Progress Kendo UI will be able to expose its business logic as agent tools—with no rewrites, no API gateways and no screen scraping. And to make adoption effortless, we’re shipping a dedicated Browser Extension in the marketplace that lets you talk to any Telerik or Kendo-powered page out of the box. 

Join us for the Telerik and Kendo UI 2026 Q2 release webinar on June 16 to see firsthand how your UI can become directly usable by AI agents. Save your seat now! 

What Is WebMCP? 

WebMCP is a proposed web standard—incubated by the Google Chrome team and currently shipping behind a flag in Chrome 146—that lets a website publish a structured contract of what it can do to AI agents running in the browser. 

Instead of an agent scraping the DOM and guessing how to submit a form, the page itself says: “Here are my tools. Here is what each one does. Here are the inputs each one needs.” The agent then calls those tools the same way it would call any function. 

There are two ways a page can declare its tools: 

• Imperative API—register tools from JavaScript with navigator.modelContext.registerTool() and unregister them when they no longer apply to the page state. 

• Declarative API—annotate an existing HTML <form> with toolname and tooldescription attributes and the browser automatically turns it into a tool the agent can call. 

The result is a clean separation: humans interact with your UI through clicks; agents interact with the same UI through tool calls. One application. Two audiences. One source of truth. 

Why does this matter? Three reasons that show up immediately: 

• Reliability over hallucination. The agent stops inventing what it thinks should happen and starts calling exactly what your app exposes. That is the difference between a demo and a deployment. 

• Zero backend changes to start. The declarative API turns existing forms into tools just by adding two attributes. Your authentication, validation and business logic stay where they are. 

• Dynamic control. Tools register and unregister as your UI state changes. A checkout tool is only available when the user is on the checkout page. That is a security boundary, not just a UX nicety. 

New Benefits for Your App’s Users 

Standards are abstract. End-user outcomes are not. Here is what your customers can do once your Telerik or Kendo-powered application speaks WebMCP: 

• Operate dense enterprise data with one sentence. “Filter to all overdue invoices in Q1, group by client and export to Excel.” What used to take seven clicks and a column chooser becomes a single prompt against your Kendo Grid. 

• Stop teaching new hires the UI. Onboarding compresses dramatically because the agent already knows how to use the application. New users describe outcomes; the agent uses the same components they would have used. 

• Move work between systems without integrations. Drag-and-drop has a new sibling: tell-and-confirm. Move a task from one Scheduler to another or replicate a row of a TreeList in another grid without writing a single line of glue code. 

• Bring accessibility forward by default. Voice, natural language and keyboard-first interactions all converge on the same underlying tools. Users who could not navigate a complex grid before suddenly can. 

• Keep humans in control. Every agent action focuses on the real component, highlights what it is about to do and waits for the user where it should. The UI stays the source of truth—the agent just gets there faster. 

One component, two audiences — the same Kendo Grid serves human clicks and agent tool calls. 

Agent-Ready Components by Default 

Telerik and Kendo UI components are not just visual chrome. They encapsulate the real work of an enterprise application—filtering, grouping, virtualizing, exporting, scheduling, validating, charting. Decades of behavior sit behind those APIs. 

That makes them the natural surface for WebMCP. The contract an agent needs to interact with your app already exists—it is the public method surface of the component you are already using. We are exposing it. 

The Q2 2026 ships agent-ready behavior across our flagship components in Preview. A Kendo Grid registers tools like filter_data, group_by, sort_by and export. A Scheduler registers create_event, reschedule and find_availability. A Form registers submit with a fully typed input schema derived from your field definitions. You do not write the tool plumbing—the componentdoes. 

Then the browser-based AI Agent calls the tools and executes the defined logic. Here’s an example of a tool call in our Browser Extension: 

For richer interactions, our imperative wrappers let a component register and tear down tools as its state changes—so a chart only offers drill_down while data is loaded and an editor only offers save_draft while a draft exists. Tools come and go with your UI, which is the trust model enterprises ask for. 

“Flip a property, not rewrite your app. The Telerik and Kendo components you already ship become the agent surface of your application.” 

Of course, you can plug in and customize the tools, adapting their descriptions to suit best your business workflows and scenarios. This helps create an even stronger bridge between the agent and your application. 

The Telerik & Kendo UI Browser Extension 

WebMCP is a powerful primitive but most users today do not have an agentic browser configured. We did not want that to be standing between your application and its first agent-driven workflow. 

That is why, alongside the Q2 2026 release, we are publishing the Telerik & Kendo UI Browser Extension. It is a lightweight companion that brings an agent directly into any page that exposes WebMCP tools—Telerik- and Kendo-powered pages get full first-class treatment.

The Telerik & Kendo UI Browser Extension—available in the marketplace alongside the Q2 2026 release. Image is illustrative; the published version may differ. 

What you get out of the box: 

• Tool discovery for the active tab. Open the extension and see every WebMCP tool the page has registered, with its description and input schema. 

• A chat surface that runs against those tools. Ask the page to do something in plain language; the extension picks the right tool, fills the inputs and calls it. 

• The Blueprint mode enables the user to go to a page, scan the DOM tree and based on the elements the AI sees, it will attempt to map each UI component on the page to a relevant Telerik/Kendo widget. 

• Local-first by design. Your API keys stay in local extension storage. Nothing leaks server-side without your consent. 

• Tools, Usage and Settings tabs. Manage which sites are allowed to register tools, watch how often each tool is called and pick the model that fits the workload. 

If you are building with Telerik and Kendo UI, the extension is the fastest way to demonstrate WebMCP value to your team without setting up a Chrome flag, a Gemini key or a custom agent runtime. Install it, open your existing app and watch the agent use the components you already shipped. 

Availability: Q2 2026 in Preview 

WebMCP itself is still an early-preview standard. Chrome ships it behind a flag, the API surface is still settling and headless agent execution—calling tools without a visible browser tab—is explicitly out of scope today. To enable WebMCP in Chromе, visit chrome://flags/#enable-webmcp-testing and enable the WebMCP for testing. 

We are matching that posture. Our Q2 2026 release ships WebMCP support in Preview across the Telerik and Kendo UI products: 
 
- Telerik UI for Blazor 
- Kendo UI for Angular 
- KendoReact 

We also ship a new Browser Extension in the marketplace: Telerik & KendoUI.  

“Preview” means production-quality engineering with an explicit signal that the underlying contract may evolve as the standard matures. We will move with it. 

If your team has an agent strategy on the 2026 roadmap, this is the moment to wire up a pilot. Pick one workflow—a complex form, a reporting grid, a scheduling view—turn on WebMCP, install the extension and watch how your users interact with it within a week. 

Explore the WebMCP Value 

WebMCP is not a UI revolution. It is a contract revolution—between the apps you build and the agents your users are bringing to them. The applications that publish that contract first will be the ones that feel obviously modern in 2026 and 2027. 

The Telerik and Kendo UI components are now designed to put you on the right side of that line without rewriting what already works. Your components remain the same. Your customers gain a new way to use them. Your team ships once and serves two audiences. 

Download and test the Q2 2026, install the Browser Extension from the marketplace and tell us what your users do with it. We are designing this with you, not for you. 

For further information about the setup, check the Product documentation: 
- Telerik UI for Blazor 
- Kendo UI for Angular 
- KendoReact 

Video: The Power of Kendo and Telerik Apps Through the WebMCP Lens 

Operations Hub: End-user can cut a multi-step logistics workflow—filter delayed shipments, map them, hand the data to the warehouse team and total the exposure—from minutes of clicking down to one sentence. 
 
Telerik and Kendo UI Workflows Powered by WebMCP 

Finance Performance: End-user can turn a quarterly review from question to scheduled meeting in seconds—surface under-performers, generate PDFs and book the follow-up without leaving the dashboard. 

AI Agent-Ready Dashboard powered by Telerik and Kendo UI with WebMCP 

Don’t forget to leave us a comment below, share your view or ask any question.