WebMCP · Maker and Shipaholic plans

Your AI agent can do the directory rounds.

Launch to Win registers your launch packages as WebMCP tools in the browser. An agent that supports document.modelContext can list your products, read every field and asset, and fill any directory's submission form from that data. You review, you submit.

See plans with WebMCP Read the documentation

// tools registered on launchto.win when you're signed in

list-packages() → your launch packages
get-package(slug) → name, tagline, descriptions, pricing, links, logo & screenshot URLs
create-package(name, url, tagline, …) → a new package; the URL triggers background enrichment

Results come back as MCP tool results. Session-authenticated: the agent acts as you. Full reference

Why this matters

Building got cheap. Launching is still manual.

01

Agents build the product

Cursor, Claude Code and friends can ship an app in an afternoon. More makers now have five products than one.

02

Agents can't read your launch copy

Until now, an agent asked to submit your product had to guess the tagline, find a logo and invent a description. WebMCP gives it the real package.

03

Directories still want a human

Most directories require your own account and a review step. WebMCP keeps you in the loop: the agent fills, you press submit.

How it works

From package to submitted, agent-driven

  1. 01

    Sign in on a WebMCP plan

    When you're signed in on Maker or Shipaholic, launchto.win registers its tools with document.modelContext in your browser.

  2. 02

    Your agent discovers the tools

    Any WebMCP-aware agent in that browser sees list-packages, get-package and create-package.

  3. 03

    The agent reads the package

    It pulls the exact name, tagline, descriptions at every length, pricing, links and asset URLs — no guessing.

  4. 04

    It fills the form, you submit

    Point the agent at a directory's submit page. It fills the fields from your package. You check it and press submit.

Prefer to click yourself? Use the extension.

The Chrome extension fills any directory form from your package with one click, including logo and screenshot uploads. Included on every plan.

Chrome extension

Several products? That is the point.

One package per product, one agent, every directory. Read the playbook for launching a whole portfolio without redoing the work.

Launch multiple products

For developers and agent builders

The reference: enable, verify, every schema and error.

Browser support table, a live check that the tools are registered, the input schema and result of each tool, the package document, and a changelog.

Read the documentation

FAQ

Frequently asked questions

What is WebMCP?
WebMCP is a browser proposal (webmachinelearning/webmcp) that lets a web page register tools with document.modelContext. An AI agent running in the browser can discover those tools and call them, with the user's own session and permissions.
Which agents can use it?
Any in-browser agent built against the WebMCP proposal. Chrome 149+ and Edge 150+ expose it through an origin trial that launchto.win has joined; Brave Leo and ChatGPT desktop are experimental. The documentation keeps the browser table current.
What can the agent do with my package?
List your launch packages, read a package with every field and asset URL, and create a new package from a name and URL (the rest is enriched in the background). It acts as you, in your browser, on your account.
Does it submit to directories automatically?
The agent reads your package through WebMCP and then fills the directory's form itself. You review and submit. Fully automated server-side submission is a separate, semi-automated feature on the Shipaholic plan.
Which plan includes WebMCP?
Maker ($19/month) and Shipaholic ($49/month). The Free plan includes the Chrome extension but not WebMCP.

Give your agent the real package.

Build a launch package, upgrade to Maker or Shipaholic, and your agent can read it through WebMCP.

See pricing Build a package first

Free plan available. No credit card required.