Skip to main content
Glama

WebMCP Vision Tool

WebMCP Vision Tool is an agent-ready shopping experience for visual product decisions. It combines fast, structured commerce actions with focused product imagery, so an AI agent can answer questions about appearance without relying on full-page screenshots or guesswork.

Live release

The client and server are deployed from the tested release on main. The server uses Neon PostgreSQL for production data.

Related MCP server: Vision MCP Server

The experience

Shoppers can search the catalog, inspect product details, compare specifications, evaluate appearance, match products to a room photo, and add products to a cart. The application uses two complementary tool layers:

  • Structured tools handle discovery, filtering, product details, compatibility, cart actions, coupons, and checkout.

  • Vision tools provide bounded product-image references when a request needs a visual judgment about finish, color, shape, or style.

The application supplies product data and image references. A multimodal agent performs the visual interpretation; this project does not pretend to be an image model.

Vision tools

Tool

Purpose

getProductImage

Return one product image and its metadata for visual evaluation.

compareProductAesthetics

Return an ordered set of product images for side-by-side comparison.

highlightVisualDifference

Return the assets needed to explain visual differences between products.

matchToUserPhoto

Return catalog candidates that can be compared with a shopper photo.

pickBestFit

Provide candidate images and preference context for a visual recommendation.

Vision results are bounded and deterministic. Product IDs are validated, duplicate candidates are removed, and missing images are reported instead of silently substituted. Image bytes, credentials, and uploaded photos are never placed in WebMCP tool schemas.

Shopper flow

  1. Use structured search to narrow the catalog.

  2. Request product images only when the question requires visual judgment.

  3. Optionally select a local JPG, PNG, or WebP room photo. The browser keeps a temporary object URL until the shopper removes it or leaves the page.

  4. Compare the returned references and explain the trade-offs.

  5. Add an item to the cart only when the shopper asks. Checkout is never invoked implicitly.

Technology

  • Wasp 0.25 full-stack application

  • React, TypeScript, and Vite client

  • Node server with Prisma and PostgreSQL

  • use-webmcp-tool for browser tool registration

  • Vitest for contract and behavior tests

This is one application repository with a React/Vite browser client and an Express/TypeScript API. The two deploy independently without maintaining separate codebases.

Local development

Run the project with Node 24.14.1 or newer and a PostgreSQL database.

npm install
npx prisma generate
npx prisma migrate deploy
npm run build:server
npm run build:client
npm run start:server

The client runs on port 3000 and the server on port 3001. In Codespaces, the development proxy keeps browser requests same-origin. Keep .env files, database URLs, API keys, and uploaded photos out of Git.

Verification

From the repository root:

npm run build:server
npm run build:client
npm test

License

WebMCP Vision Tool is released under the MIT License.

A
license - permissive license
Not graded
quality - not tested
B
maintenance

Maintenance

Maintainers
Response time
Release cycle
Releases (12mo)
Commit activity

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Servers

  • F
    license
    A
    quality
    Not graded
    maintenance
    Enables AI agents to analyze images through vision AI providers (Gemini, OpenAI, Claude), performing tasks like image description, object detection with bounding boxes, region-specific analysis, and precise color extraction without consuming context window with raw pixels.
    4
  • A
    license
    Not graded
    quality
    C
    maintenance
    Enables AI agents to perform product discovery from natural language shopping intents, returning ranked products with merchant links without completing checkout.
    MIT

View all related MCP servers

Related MCP Connectors

  • Agent-native product catalog: 300M+ products, 150,000+ stores, deliver_to ranking.

  • Product discovery for AI agents: ranked products and bundles from the open merchant web.

  • Policy review and purchase discovery for AI-agent commerce actions.

View all MCP Connectors

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/onerandomd3v/Web-MCP-Vision-Tool'

If you have feedback or need assistance with the MCP directory API, please join our Discord server